Page 1
Nature Climate Change
| Volume 13 | February 2023 | 174–181
174
nature climate change
Article
https://doi.org/10.1038/s41558-022-01562-8
Temperature effect on erosion-induced
disturbances to soil organic carbon cycling
Zhengang Wang
1,2,8
, Yizhe Zhang
2,8
, Gerard Govers
3
, Guoping Tang
2
,
Timothy A. Quine
4
, Jianxiu Qiu
2
, Ana Navas
5
, Haiyan Fang
6
, Qian Tan
1
& Kristof Van Oost
7
Erosion exerts control on soil organic carbon (SOC) and both erosion and
SOC are affected by climate. To what extent temperature controls the
coupling between these erosion–C interactions remains unclear. Using
137
Cs
and SOC inventories from catchments spanning different climates, we find
that increasing decomposition rates with temperature result in the efficient
replacement of SOC laterally lost by erosion in eroding areas but lower
preservation of deposited SOC in depositional areas. When combined at the
landscape level, the erosion-induced C sink strength per unit lateral SOC flux
increases with temperature from 0.19 g C (g C)
−1
at 0 °C to 0.24 g C (g C)
−1
at 25 °C. We estimated that the global C sink of 0.050 Pg C yr
−1
induced by
water erosion on croplands increases by 7% because of climate change.
Our results reveal a negative feedback loop between climate change and
erosion-induced disturbance to SOC cycling.
Soils store more carbon than the atmosphere and vegetation com-
bined, thereby offering a large potential to mitigate climate change via
C sequestration
1
,
2
; at the same time, soils also pose a threat as global
warming and agricultural expansion might release parts of this large
stock
3
–
7
. Soil organic C (SOC) storage is controlled by the balance
between C inputs and decomposition, both of which are mediated by
climate
8
,
9
. SOC stocks and turnover times are known to covary with cli-
mate globally
10
,
11
but the relationship is complex: the sensitivity of SOC
cycling to temperature change is higher in cold regions than in warm
regions
11
and different ecosystems have different degrees of sensitivity
and coping mechanisms adapting to climate warming
4
,
5
,
12
,
13
. It is there-
fore not surprising that the response of SOC to climate change is among
the largest uncertainties under future climate change projections
3
,
8
,
14
,
15
.
Erosion causes a large flux of SOC across the land-to-ocean aquatic
continuum (LOAC)
16
–
18
, especially when natural land cover is converted
to croplands
19
,
20
. These lateral SOC fluxes exert substantial impacts on
C cycling processes along the LOAC and result in notable amounts of C
exchange between land and atmosphere
21
,
22
. Previous studies focusing
on the aquatic system showed that these disturbances of erosion to the
C cycling are also modulated by temperature. For instance, organic C
in the riverine water and sediments was found to decompose more
rapidly with increasing temperature resulting in higher CO
2
evasion
from rivers in warm regions
23
,
24
. Similarly, the decomposition rate of
buried organic C in lakes was found to increase with temperature
25
,
26
.
As is the case of mobilized sediments entering aquatic envi-
ronments, temperature is expected to affect C dynamics in eroding
uplands as well, where most C mobilized by erosion is redeposited
27
–
29
.
In eroding areas, the reduced SOC stocks (due to SOC erosion) will
decrease soil heterotrophic respiration, which, in combination with
continued photosynthesis, results in a net uptake of atmospheric
C by the soil. This C uptake replaces parts of the SOC lost laterally
by erosion, a process that is referred to as ‘dynamic replacement’
30
–
33
.
Received: 30 January 2022
Accepted: 21 November 2022
Published online: 30 January 2023
Check for updates
1
Guangdong Provincial Key Laboratory of Water Quality Improvement and Ecological Restoration for Watersheds, Institute of Environmental and
Ecological Engineering, Guangdong University of Technology, Guangzhou, China.
2
Key Laboratory of Urbanization and Geo-simulation of Guangdong
Province, School of Geography and Planning, Sun Yat-sen University, Guangzhou, China.
3
Division of Geography, Department of Earth and Environmental
Sciences, KU Leuven, Leuven, Belgium.
4
Geography, College of Life and Environmental Sciences, University of Exeter, Exeter, UK.
5
Soil and Water
Department, Estación Experimental de Aula Dei (EEAD-CSIC), Zaragoza, Spain.
6
Key Laboratory of Water Cycle and Related Land Surface Processes,
Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China.
7
Georges Lemaître Center for Earth and
Climate Research (TECLIM), Earth and Life Institute, Université Catholique de Louvain, Louvain-la-Neuve, Belgium.
8
These authors contributed equally:
Zhengang Wang, Yizhe Zhang. e-mail:
wangzhg33@mail.sysu.edu.cn
Page 2
Nature Climate Change
| Volume 13 | February 2023 | 174–181
175
Article
https://doi.org/10.1038/s41558-022-01562-8
spatial variation of normalized SOC stocks was independent of varia-
tions of SOC inputs (Extended Data Fig. 3b). In contrast, an increase
of SOC decomposition rates not only decreased total SOC stocks
(Extended Data Fig. 4a) but also decreased the spatial variation of SOC
stocks (Extended Data Fig. 4b). When both SOC input and decomposi-
tion rates were allowed to vary with temperature, the spatial variability
of SOC stocks would decrease with temperature (Fig.
1b
and Extended
Data Fig. 5b). This is because in eroding areas the C replacement ratio
of the laterally exported SOC increased with temperature, while in
depositional areas the burial efficiency of deposited SOC decreased
with temperature (Extended Data Fig. 5c).
In line with model predictions, field observations showed that the
average C replacement ratio of the eroding profiles increased with tem-
perature from ~65.0% at 0 °C to ~87.5% at 25 °C (Fig.
1c
). This indicates
that the C sink effect in the eroding area increases with temperature.
A linear regression between soil redistribution (independent vari-
able) and the burial efficiency of deposited SOC (dependent variable)
showed that the slope of regression lines decreased with increasing
temperature (Fig.
1d
). This indicates that CBE decreased more rapidly
with depositional rates in warm regions than in cold regions. Thus, the
difference between areas subjected to soil redistribution and refer-
ence areas becomes smaller with increasing temperature, as eroded
SOC is more rapidly replaced and deposited SOC is less efficiently
preserved in warmer environments. Consequently, reduced SOC
variability with soil redistribution at the landscape scale is observed
in warm regions (Fig.
1b
).
The overall and integrated influence of temperature on
erosion-induced C exchange at the scale of upland catchment is deter-
mined by the balance between the increase of C replacement with
temperature in eroding areas (more C replaced in warm regions) and
the decrease of CBE with temperature in depositional areas (more C
lost in warm regions). Model simulations suggest that the combina-
tion of these two counteracting processes will result in an increase
of the efficiency of the C sink per unit lateral SOC flux (the ratio of
erosion-induced land–atmosphere C exchanges to lateral SOC fluxes
due to erosion) with increasing temperature (Extended Data Fig. 5d).
The empirical data also suggest an increase of the C sink strength per
unit lateral SOC flux but this trend is not statistically significant (Fig.
1e
).
Thus, in colder climates where C turnover is slower, the dominant effect
of soil erosion is to cause a physical redistribution of SOC, thereby
having less impact on the C exchange between soil and atmosphere.
In ecosystems characterized by faster C decomposition rates, higher
C replacement ratios of the laterally lost SOC result in a more efficient
pathway to sequester atmospheric CO
2
from the atmosphere at erod-
ing sites. At depositional sites, the soil–atmosphere C exchange is also
enhanced but in the opposite direction, leading to higher losses of C
from the soil to the atmosphere as temperature increases. Carbon
sequestration in eroding areas increases more rapidly with tempera-
ture than C loss in depositional areas, so that the overall erosion sink
becomes larger with increasing temperature.
Erosion–C coupling at the global scale
The temperature effects on erosion–C interactions at the global scale
were explored using available spatial datasets in a spatially explicit
approach (Supplementary Methods Section 5). The net primary
productivity (NPP) was regarded as an estimate of the vegetation C
inputs to soils based on the assumption that the vegetation biomass
is at steady state
11
. We acknowledge that some regions are subjected to
disturbances such as land use change and fires and are at a transition
stage. Nevertheless, since the remote sensing-derived NPP (resolution
30 arcsec) and Harmonized World Soil Database derived from SOC
profiles (resolution 30 arcsec) represent the average condition of the
raster grid, they can be regarded as an approximation of the equilibrium
condition of the area. At the global scale, both NPP and SOC decom-
position rates (estimated by dividing NPP by SOC stocks) increased
In depositional areas, the lateral influx of SOC leads to a disequilibrium
between the SOC stock and local vegetation inputs. The buried sedi-
ments will therefore lose more SOC by microbial decomposition than
they gain SOC through vegetation inputs and SOC stored in deposited
sediments declines after burial
34
–
36
. The ratio between the preserved
organic C in buried sediments and the originally deposited SOC is
termed C burial efficiency (CBE)
37
,
38
. Thus, erosion brings fresh soil
minerals from deeper horizons to the surface where they are mixed
with and adsorb some of the C derived from vegetation
39
. During the
geomorphic cascade of erosion, transport and deposition, a fraction
of the fixed C is decomposed, while the remaining is preserved in burial
sites often characterized by low decomposition rates, resulting in a net
sink of atmospheric C at the scale of eroding landscapes
21
,
40
. While both
C inputs and SOC decomposition depend on temperature, the effects of
temperature on the replacement of SOC lost from the eroding areas and
the preservation of buried SOC in the depositional areas have hitherto
not been explored. Therefore, the relationship between temperature
and the efficacy of the C sink resulting from the operation erosion and
subsequent sediment burial can, at present, not be fully assessed and
it remains uncertain how this C sink will respond to climate change.
Here, we investigated the temperature effects on erosion–C cou-
pling via three lines of evidence. First, we compiled a dataset from
the literature of SOC and
137
Cs inventories of agricultural soils in Asia,
Europe and North America spanning a wide range of climatic conditions
(Extended Data Fig. 1 and Supplementary Table 1). In total, 1,562 profiles
from 20 catchments including stable and erosional soils at upland
hillslopes as well as colluvial soils at footslopes were examined. The arti-
ficial fallout radioisotope
137
Cs was used as a tracer to determine lateral
soil and SOC redistribution rates (erosion and deposition rates). The
direction and magnitude of net exchanges of C with the atmosphere
were derived from field observations of SOC stocks and soil redistribu-
tion rates (SRRs) at stable, eroding and depositional sites. The effect
of temperature on erosion–C coupling was assessed by confronting
the SOC cycling characteristics with the mean annual temperature of
each catchment. Second, we used a model that combines a SOC cycling
module with soil redistribution processes to simulate SOC dynamics at
the hillslope scale in eroding landscapes as a function of temperature
(Supplementary Methods Section 6). The model was run in a series
of scenarios with different C input and decomposition conditions to
explore how temperature affects the erosion–C coupling via chang-
ing the C input and decomposition rates. Third, by combining global
estimates of SOC stocks and of agricultural soil erosion rates
21
, we
estimated the global lateral SOC fluxes on croplands in relation to sur-
face erosion (Supplementary Methods Section 5). The same erosion–C
integrated model was applied at the global scale to investigate how the
erosion–C coupling varies with temperature.
Erosion–C coupling in upland catchments
As observed in previous studies
41
, SOC stocks decreased with tempera-
ture (Fig.
1a
), indicating a more rapid increase of SOC decomposition
rate than SOC input with temperature. Within each catchment, SOC
stocks were positively correlated with SRRs with eroding areas depleted
of SOC and depositional areas enriched in SOC (Extended Data Fig. 2).
To avoid the influence of SOC stock differences on the comparison
between catchments, we normalized the SOC stock by dividing them
by the average SOC stock of the stable sites within each catchment.
The slope of the regression line between the SRRs (derived from
137
Cs
inventories) and the normalized SOC stock represents the sensitivity of
the (normalized) SOC stock to the variation of SRRs for each catchment.
As shown in Fig.
1b
, the slope of the regression lines decreased with
increasing temperature, indicating that SOC stocks are more sensitive
to soil redistribution in cold regions than in warm regions.
Numerical simulations of SOC dynamics at the hillslope scale
(Supplementary Methods Section 6) demonstrated that SOC stocks
changed linearly with SOC inputs (Extended Data Fig. 3a) and that the
Page 3
Nature Climate Change
| Volume 13 | February 2023 | 174–181
176
Article
https://doi.org/10.1038/s41558-022-01562-8
with temperature, while SOC stocks decreased with temperature
(Fig.
2
). To assess the temperature sensitivity of SOC cycling processes,
we used the concept of Q
10
values, that is the factor by which a given
SOC cycling metric varies for a 10 °C increase in temperature
42
. We
obtained a globally averaged Q
10
value of 1.43 for the SOC decomposi-
tion rate, which is in line with earlier estimates
11
,
43
. Due to this positive
relationship between SOC decomposition rates and temperature, the C
replacement ratio of laterally lost SOC by erosion was found to increase
with temperature while the CBE of deposited SOC decreased with tem-
perature (Fig.
1
). In line with the field data interpretation and hillslope
scale model simulations, the magnitude of the erosion-induced C sink
per unit of lateral SOC flux was positively correlated with temperature
increasing from ~0.19 g C (g C)
−1
at 0 °C to ~0.24 g C (g C)
−1
at 25 °C at
the global scale (Fig.
2d
).
To assess the C sink strength per unit cropland area, it is also neces-
sary to consider the absolute magnitude of the lateral SOC fluxes. As
reported in previous studies
21
,
44
, we estimated global soil and C ero-
sion to be 17.89 and 0.26 Pg yr
−1
, respectively. We obtained a global C
sink of 0.050 Pg C yr
−1
induced by water erosion on croplands (Fig.
3
).
This estimate is lower than the one obtained by ref.
21
because the
latter included both water and tillage erosion on croplands as well as
agricultural erosion on pasture and rangelands. The water erosion rate
of croplands was found to be positively related to the mean annual
precipitation at the global scale
21
(Extended Data Fig. 6a). Since mean
annual precipitation was positively related to mean annual temperature
(Extended Data Fig. 6b), the water erosion rate on croplands was also
positively correlated with mean annual temperature globally (Extended
Data Fig. 6c). This explains why, despite a negative correlation between
10
0
10
1
10
2
SOC stock (kg C m
–2
)
a
y
= 17.51 ×
e
–0.079
×
x
(R
2
= 0.26,
P
= 0.022)
y
= 10.45 × e
–0.0091
×
x
(R
2
= 1.00,
P
< 0.001)
y
= 10.45 × e
–0.0091
×
x
(R
2
= 0.042,
P
< 0.001)
10
–1
10
0
10
1
Sensitivity of SOC to SRR variation
(yr cm
–1
)
b
y
= 0.89 × e
–0.033
×
x
(R
2
= 0.26,
P
= 0.022)
y
= 0.57 × e
–0.034
×
x
(R
2
= 0.99,
P
< 0.001)
y
= 0.63 × e
–0.039
×
x
(R
2
= 0.32,
P
< 0.001)
20
60
100
140
Mean C replacement ratio (%)
c
y
= 0.90 ×
x
+65.04 (R
2
= 0.27,
P
= 0.022)
y
= 0.60 ×
x
+75.46 (R
2
= 1.00,
P
< 0.001)
y
= 0.90 ×
x
+72.02 (R
2
= 0.88,
P
< 0.001)
10
15
20
25
Mean annual temperature (°C)
–1.2
–0.8
–0.4
0
Sensitivity of CBE to SRR variation
(yr cm
–1
)
d
y
= –0.025 ×
x
–0.21 (R
2
= 0.62,
P
= 0.0014)
y
= –0.0057 ×
x
–0.59 (R
2
= 0.99,
P
< 0.001)
y
= –0.016 ×
x
–0.62 (R
2
= 0.64,
P
< 0.001)
0
5
0
5
10
15
20
25
Mean annual temperature (°C)
10
–1
10
0
C sink strength per unit lateral C flux
(g C (g C)
–1
)
e
y
= 0.21 ×
e
0.010
×
x
(R
2
= 0.19,
P
= 0.14)
y
= 0.16 × e
0.014
×
x
(R
2
= 1.00,
P
< 0.001)
y
= 0.22 × e
0.0027
×
x
(R
2
= 0.016,
P
< 0.001)
Field observation
Hillslope simulation
Global data
Fig. 1 | Variation of SOC cycling metrics on eroding croplands with
temperature. a, Variation of SOC stocks (top 1 m) with temperature. b, Variation
of the slope of linear regression lines between the SRR (independent variable)
and the normalized SOC stock (dependent variable) with temperature.
c, Variation of the average replacement ratio of SOC laterally lost by erosion with
temperature.
d, Variation of the slope of linear regression lines between the SRR
(independent variable) and the CBE (dependent variable) with temperature.
e, Variation of the C sink strength per unit lateral SOC flux with temperature.
The red points denote the observations from the cropland sites, the blue points
denote the results of hillslope scale model simulations and the grey points
denote the results of global model simulations. The lines are the regression
curves of the points of the same colour. Texts denote the regression equations
of points of the same colour. Error bars denote the standard deviation. SRR in
b and d denotes soil redistribution rate and CBE in d denotes C burial efficiency.
One-sided F-test was used in the regression analysis. The sample numbers of field
observations of each cropland site are displayed in Supplementary Table 1.
Page 4
Nature Climate Change
| Volume 13 | February 2023 | 174–181
177
Article
https://doi.org/10.1038/s41558-022-01562-8
SOC stocks and temperature, the C erosion rate of croplands and the
erosion-induced C sink per unit cropland area were positively corre-
lated with temperature (Fig.
2e
and Extended Data Fig. 6d). C erosion
rates show an exponential increase with temperature
21
, which explains
why the Q
10
value for the C sink per unit cropland area (3.05) is much
higher than the Q
10
value for the C sink per unit lateral SOC flux (1.03).
The relationships between various SOC cycling metrics and tem-
perature derived from the three independent lines of research were
consistent with each other (Fig.
1
) but the strength of the relationships
varied due to differences in methods, scales and datasets. The SOC
stock variability caused by soil redistribution derived from model
simulations is lower than that obtained from field data (Fig.
1b
), which is
probably related to the fact that the model did not take into account the
relationships between soil redistribution and SOC inputs to soils with
lower C inputs in eroding areas due to the negative effect of erosion on
crop productivity
45
,
46
. An assessment based on
137
Cs and C stock inven-
tory data showed that the spatial variability of crop productivity due to
soil redistribution accounted for 35.5 ± 25.1% of the spatial variability
of SOC stocks, confirming that the erosion–C input relation is indeed
important (Supplementary Methods Section 12). In addition, C burial
efficiencies were closely related to the SRR (Extended Data Fig. 5c)
and tests showed that the difference in the SRRs among the datasets
used in different methods accounted for the main source of differ-
ences between model simulations at the hillslope and global scales
(Supplementary Figs. 8 and 9).
Erosion–C coupling under climate change
We applied a space-for-time approach to assess the potential varia-
tion of temperature effects on SOC cycling in croplands due to future
climate change (Supplementary Methods Section 10). The SOC cycling
metrics (for example, SOC stock, NPP and erosion) under future cli-
mate projections generated by eight global climate models in four
scenarios (Supplementary Tables 2 and 3) were estimated using the
exponential relationships between these metrics and climatic factors
(temperature and precipitation). As predicted in previous studies
3
,
47
, we
estimated a SOC stock decline of 4.1 ± 1.6% in croplands in 2100 (Fig.
4
).
–10
0
10
20
30
Mean annual temperature (°C)
10
–4
10
0
10
4
C sink strength per unit cropland
area (g C m
–2
yr
–1
)
e
y = 0.48 × e
0.11
× x
(R
2
= 0.38, P < 0.001)
Q
10
= 3.05
–10
0
10
20
30
Mean annual temperature (°C)
10
–1
10
0
C sink strength per unit lateral C flux
(g C (g C)
–1
)
d
y = 0.22 × e
0.0027
× x
(R
2
= 0.016, P < 0.001)
Q
10
= 1.03
10
–3
10
–1
10
1
SOC decomposition rate (yr
–1
)
c
y = 0.038 × e
0.036
× x
(R
2
= 0.28, P < 0.001)
Q
10
= 1.43
10
–2
10
–1
10
0
10
1
NPP (kg C m
–2
yr
–1
)
b
y = 0.40 × e
0.027
× x
(R
2
= 0.29, P < 0.001)
Q
10
= 1.31
10
0
10
1
10
2
SOC stock (kg C m
–2
)
a
y = 10.45 × e
–0.0091
× x
(R
2
= 0.042, P < 0.001)
Q
10
= 0.91
Fig. 2 | Temperature effects on SOC cycling metrics on eroding croplands
at the global scale. a, Variation of SOC stocks (top 1 m) with temperature.
b, Variation of NPP with temperature. c, Variation of SOC decomposition rates
with temperature.
d, Variation of the C sink strength per unit lateral C flux.
e, Variation of the C sink strength per unit cropland area with temperature. Lines
are the regression curves of corresponding SOC cycling metrics and texts denote
the regression equations. One-sided F-test was used in the regression analysis.
Page 5
Nature Climate Change
| Volume 13 | February 2023 | 174–181
178
Article
https://doi.org/10.1038/s41558-022-01562-8
Increasing precipitation under future climate projections could result
in an increase of 17.9 ± 8.6% in water erosion on croplands, which is
lower than a previous estimate that suggested an increase of well over
30% (ref.
48
). The integrated effects of SOC stock and erosion variation
cause the lateral SOC flux by water erosion on croplands to increase by
6.3 ± 7.2%. The C sink efficiency per unit lateral SOC flux would increase
by 1.1 ± 1.4% due to global warming and the total C sink by water erosion
on croplands was predicted to increase by 7.4 ± 6.2%.
Numerical simulations of SOC cycling at the global scale under dif-
ferent soil moisture conditions demonstrated that the SOC decomposi-
tion rate peaked at an intermediate soil moisture of 0.29 m
3
m
−3
under
present climatic conditions. This rate was 50% higher than the SOC
decomposition rates under either dry (0.16 m
3
m
−3
) or wet (0.40 m
3
m
−3
)
conditions (Supplementary Fig. 4 and Extended Data Fig. 7b). While
important to recognize, this soil moisture effect on SOC decomposi-
tion was significantly smaller than the effect of temperature, where the
SOC decomposition rate at 25 °C was ~2.5 times that at 0 °C (Fig.
2c
).
Because of the positive correlation between the C sink per unit lateral
SOC flux and SOC decomposition rates, the variation of the C sink per
unit lateral SOC flux with soil moisture was similar to the variation of
the SOC decomposition rate with soil moisture (Supplementary Fig. 5
and Extended Data Fig. 7c). The C sink strength per unit cropland area
is largely controlled by the SOC erosion rate. Since lateral SOC fluxes
were higher in areas where soil moisture is higher as both the SOC flux
and soil moisture are positively related to mean annual precipitation
(Supplementary Fig. 3 and Extended Data Fig. 7a), the C sink strength
per unit cropland area also increased with increasing soil moisture
(Supplementary Fig. 6 and Extended Data Fig. 7d). Soil moisture effects
under projected future climates were similar to those for the present
climate (Extended Data Fig. 7), with the limited differences being due
to differences between present and future lateral soil fluxes rather than
variations in soil moisture per se.
In summary, our analysis, based on multiple lines of evidence,
shows strong support for the proposition that temperature is an impor-
tant control on the magnitude of erosion-induced disturbances to the
SOC cycling in croplands. We found that the C replacement ratio of
60° N
30° N
0°
30° S
120° W
0
6
8
10
12
SOC stock (kg C m
–2
)
NPP (kg C m
–2
yr
–1
)
SOC mineralization rate (yr
–1
)
C sink strength per unit cropland area (g C m
–2
yr
–1
)
C sink strength per unit lateral C flux (g C (g C)
−1
)
16
22
30
50
0
0.15
0.30
0.45
0.60
0.75
0.90
1.05
1.35
0
0.03
0.06
0.09
0.12
0.15
0.20
0.30
0.60
0
2
6
10
18
25
36
50
68
0
0.10
0.15
0.18
0.21
0.24
0.26
0.28
0.38
60° W
0°
60° E
120° E
120° W
60° W
0°
60° E
120° E
120° W
60° W
0°
60° E
120° E
120° W
60° W
0°
60° E
120° E
120° W
60° W
0°
60° E
120° E
60° N
a
b
c
e
d
30° N
0°
30° S
60° N
30° N
0°
30° S
60° N
30° N
0°
30° S
60° N
30° N
0°
30° S
Fig. 3 | Spatial variation of SOC cycling metrics on eroding croplands. a, SOC stocks (top 1 m). b, NPP. c, SOC decomposition rates. d, The C sink strength per unit
lateral C flux.
e, The C sink strength per unit cropland area.
Page 6
Nature Climate Change
| Volume 13 | February 2023 | 174–181
179
Article
https://doi.org/10.1038/s41558-022-01562-8
SOC that is laterally lost by erosion increases with temperature, while
the CBE of deposited SOC decreases with temperature. This results
in a larger landscape-scale spatial variation of SOC stocks due to ero-
sion in cold regions than in warm regions. Model simulations revealed
that increasing C decomposition rates with temperature are the main
factor controlling the spatial heterogeneity of SOC cycling in eroding
landscapes, while increasing C inputs with temperature exert the same
impacts in both eroding and depositional areas. Our results also show
that the erosion-induced C sink per unit lateral SOC flux increases
with temperature. The positive relation between sink strength and
temperature is further enhanced by the fact that C erosion rates are
also positively related to temperature. Together these mechanisms
cause a significant increase of the C sink per unit cropland area with
temperature. Previous studies had revealed that lateral SOC fluxes due
to erosion exert an important impact on global C cycling
16
,
21
,
22
and this
impact is mediated by the erosion intensity
35
. Our results improve our
understanding of the complexity of geomorphology–biogeochem-
istry interactions by adding the effects of temperature. Variations
in temperature not only affect C erosion directly but also change the
in situ SOC cycling responses to the erosion-induced lateral SOC fluxes.
The variability of erosion–C coupling with temperature presented in
our analysis implies that a more detailed assessment of the spatial
variability of erosion-induced SOC cycling disturbances is required to
fully understand the feedback loop between SOC cycling and climate
changes (Fig.
4
). In the context of exacerbated erosion by climatic
extremes and future global warming
49
–
51
, soil redistribution would play
a more important role in global C cycling. Substantial C emission from
soils is expected in the future global warming
3
–
5
,
52
; however, this net
emission would cease when SOC stocks reach a new equilibrium. In con-
trast, the C uptake caused by erosion and subsequent burial functions
as long as erosion occurs. Our results show that increasing temperature
also enhances the erosion-induced C sink that could potentially offset a
small fraction of the C emission from soils due to global warming. This
effect is analogous to the widely recognized negative feedback loop
between silicate weathering and climate change
53
,
54
. Nevertheless, in
the immediate future, efforts to replace lost SOC in eroded landscapes
60° N
30° N
0°
30° S
120° W
–12
–10
–8
–6
–4
–2
0
2
4
–60
–40
–20
0
20
∆ SOC stock (relative variation, %)
∆ Soil erosion (relative variation, %)
∆ Lateral C flux (relative variation, %)
∆ C sink strength per unit cropland area (relative variation, %)
∆ C sink strength per unit lateral C flux (relative variation, %)
40
60
80
100
–60
–40
–20
0
20
40
60
80
100
–60
–40
–20
0
20
40
60
80
100
–60
–40
–20
0
20
40
60
80
100
60° W
0°
60° E
120° E
120° W
60° W
0°
60° E
120° E
120° W
60° W
0°
60° E
120° E
120° W
60° W
0°
60° E
120° E
Relativ
e v
ariation (%)
120° W
60° W
0°
60° E
120° E
60° N
a
b
c
d
e
f
30° N
0°
30° S
60° N
30° N
0°
30° S
60° N
30° N
0°
30° S
60° N
30° N
0°
30° S
40
20
0
–20
SOC
stock
Soil
erosion
Lateral
C flux
C sink
efficiency
C sink
Fig. 4 | Response of SOC cycling in eroding croplands to climate change. a–f, Relative variation of SOC stocks (a), soil erosion (b), the lateral SOC flux due to erosion
(
c), the C sink strength per unit lateral SOC flux (d), the C sink strength per unit cropland area (e) caused by erosion in 2100 and the statistics of them (f). Error bars in f
denote the standard deviation.
Page 7
Nature Climate Change
| Volume 13 | February 2023 | 174–181
180
Article
https://doi.org/10.1038/s41558-022-01562-8
through erosion control may yield net C-sequestration benefits in the
pursuit of global net-zero objectives.
Online content
Any methods, additional references, Nature Portfolio reporting sum-
maries, source data, extended data, supplementary information,
acknowledgements, peer review information; details of author contri-
butions and competing interests; and statements of data and code avail-
ability are available at
https://doi.org/10.1038/s41558-022-01562-8
.
References
1. Lal, R. Soil carbon sequestration impacts on global climate
change and food security. Science
304
, 1623–1627 (2004).
2. Amundson, R. et al. Soil and human security in the 21st century.
Science
348
, 1261071 (2015).
3. Crowther, T. W. et al. Quantifying global soil carbon losses in
response to warming. Nature
540
, 104–108 (2016).
4. Nottingham, A. T., Meir, P., Velasquez, E. & Turner, B. L. Soil carbon
loss by experimental warming in a tropical forest. Nature
584
,
234–237 (2020).
5. Xue, K. et al. Tundra soil carbon is vulnerable to rapid microbial
decomposition under climate warming. Nat. Clim. Change
6
,
595–600 (2016).
6. Houghton, R. A. & Nassikas, A. A. Global and regional fluxes of
carbon from land use and land cover change 1850–2015. Glob.
Biogeochem. Cycles
31
, 456–472 (2017).
7. Hicks Pries, C. E., Castanha, C., Porras, R. C. & Torn, M. S. The
whole-soil carbon flux in response to warming. Science
355
,
1420–1423 (2017).
8. Davidson, E. A. & Janssens, I. A. Temperature sensitivity of soil
carbon decomposition and feedbacks to climate change. Nature
440
, 165–173 (2006).
9. Mahecha, M. D. et al. Global convergence in the temperature
sensitivity of respiration at ecosystem level. Science
329
,
838–840 (2010).
10. Carvalhais, N. et al. Global covariation of carbon turnover times
with climate in terrestrial ecosystems. Nature
514
, 213–217 (2014).
11. Koven, C. D., Hugelius, G., Lawrence, D. M. & Wieder, W. R. Higher
climatological temperature sensitivity of soil carbon in cold than
warm climates. Nat. Clim. Change
7
, 817–822 (2017).
12. Sistla, S. A. et al. Long-term warming restructures Arctic tundra
without changing net soil carbon storage. Nature
497
, 615–618
(2013).
13. Dorrepaal, E. et al. Carbon respiration from subsurface peat
accelerated by climate warming in the subarctic. Nature
460
,
616–619 (2009).
14. Karhu, K. et al. Temperature sensitivity of soil respiration rates
enhanced by microbial community response. Nature
513
, 81–84
(2014).
15. Knorr, W., Prentice, I. C., House, J. I. & Holland, E. A. Long-term
sensitivity of soil carbon turnover to warming. Nature
433
,
298–301 (2005).
16. Galy, V., Peucker-Ehrenbrink, B. & Eglinton, T. Global carbon
export from the terrestrial biosphere controlled by erosion.
Nature
521
, 204–207 (2015).
17. Stallard, R. F. Terrestrial sedimentation and the carbon cycle:
coupling weathering and erosion to carbon burial. Glob.
Biogeochem. Cycles
12
, 231–257 (1998).
18. Regnier, P. et al. Anthropogenic perturbation of the carbon fluxes
from land to ocean. Nat. Geosci.
6
, 597–607 (2013).
19. Montgomery, D. R. Soil erosion and agricultural sustainability.
Proc. Natl Acad. Sci. USA
104
, 13268–13272 (2007).
20. Cerdan, O. et al. Rates and spatial variations of soil erosion in
Europe: a study based on erosion plot data. Geomorphology
122
,
167–177 (2010).
21. Van Oost, K. et al. The impact of agricultural soil erosion on the
global carbon cycle. Science
318
, 626–629 (2007).
22. Chappell, A., Baldock, J. & Sanderman, J. The global significance
of omitting soil erosion from soil organic carbon cycling
schemes. Nat. Clim. Change
6
, 187–191 (2016).
23. Raymond, P. A. et al. Global carbon dioxide emissions from inland
waters. Nature
503
, 355–359 (2013).
24. Lauerwald, R., Laruelle, G. G., Hartmann, J., Ciais, P. & Regnier,
P. A. G. Spatial patterns in CO
2
evasion from the global river
network. Glob. Biogeochem. Cycles
29
, 534–554 (2015).
25. Gudasz, C. et al. Temperature-controlled organic carbon
mineralization in lake sediments. Nature
466
, 478–481 (2010).
26. Marotta, H. et al. Greenhouse gas production in low-latitude lake
sediments responds strongly to warming. Nat. Clim. Change
4
,
467–470 (2014).
27. Rommens, T. et al. Soil erosion and sediment deposition in the
Belgian loess belt during the Holocene: establishing a sediment
budget for a small agricultural catchment. Holocene
15
, 1032–
1043 (2005).
28. Phillips, J. D. Fluvial sediment budgets in the North Carolina
Piedmont. Geomorphology
4
, 231–241 (1991).
29. Trimble, S. W. & Crosson, P. U.S. Soil erosion rates—myth and
reality. Science
289
, 248–250 (2000).
30. Harden, J. W. et al. Dynamic replacement and loss of soil carbon on
eroding cropland. Glob. Biogeochem. Cycles
13
, 885–901 (1999).
31. Billings, S. A., Buddemeier, R. W., Richter, D. D., Van Oost, K. &
Bohling, G. A simple method for estimating the influence of
eroding soil profiles on atmospheric CO
2
. Glob. Biogeochem.
Cycles
24
, GB2001 (2010).
32. Berhe, A. A., Harden, J. W., Torn, M. S. & Harte, J. Linking soil
organic matter dynamics and erosion-induced terrestrial carbon
sequestration at different landform positions. J. Geophys. Res.
Biogeosci.
113
, G04039 (2008).
33. Nadeu, E., Berhe, A. A., de Vente, J. & Boix-Fayos, C. Erosion,
deposition and replacement of soil organic carbon in
Mediterranean catchments: a geomorphological, isotopic and
land use change approach. Biogeosciences
9
, 1099–1111 (2012).
34. Wang, Z. et al. The fate of buried organic carbon in colluvial soils:
a long-term perspective. Biogeosciences
11
, 873–883 (2014).
35. Wang, Z., Van Oost, K. & Govers, G. Predicting the long-term fate
of buried organic carbon in colluvial soils. Glob. Biogeochem.
Cycles
29
, 65–79 (2015).
36. VandenBygaart, A. J., Kroetsch, D., Gregorich, E. G. & Lobb, D. Soil
C erosion and burial in cropland. Glob. Change Biol.
18
, 1441–1452
(2012).
37. Burdige, D. J. Burial of terrestrial organic matter in marine
sediments: a re-assessment. Glob. Biogeochem. Cycles
19
,
GB4011 (2005).
38. Sobek, S. et al. Organic carbon burial efficiency in lake sediments
controlled by oxygen exposure time and sediment source.
Limnol. Oceanogr.
54
, 2243–2254 (2009).
39. Doetterl, S. et al. Erosion, deposition and soil carbon: a review of
process-level controls, experimental tools and models to address
C cycling in dynamic landscapes. Earth Sci. Rev.
154
, 102–122
(2016).
40. Wang, Z. et al. Human-induced erosion has offset one-third of
carbon emissions from land cover change. Nat. Clim. Change
7
,
345–349 (2017).
41. Jobbágy, E. G. & Jackson, R. B. The vertical distribution of soil
organic carbon and its relation to climate and vegetation. Ecol.
Appl.
10
, 423–436 (2000).
42. Meyer, N., Welp, G. & Amelung, W. The temperature sensitivity
(Q10) of soil respiration: controlling factors and spatial prediction
at regional scale based on environmental soil classes. Glob.
Biogeochem. Cycles
32
, 306–323 (2018).
Page 8
Nature Climate Change
| Volume 13 | February 2023 | 174–181
181
Article
https://doi.org/10.1038/s41558-022-01562-8
43. Zhou, T., Shi, P., Hui, D. & Luo, Y. Global pattern of temperature
sensitivity of soil heterotrophic respiration (Q10) and its
implications for carbon–climate feedback. J. Geophys. Res.
Biogeosci.
114
, G02016 (2009).
44. Doetterl, S., Van Oost, K. & Six, J. Towards constraining the
magnitude of global agricultural sediment and soil organic
carbon fluxes. Earth Surf. Process. Landf.
37
, 642–655 (2012).
45. Kosmas, C. et al. The effects of tillage displaced soil on soil
properties and wheat biomass. Soil Tillage Res.
58
, 31–44
(2001).
46. Dercon, G. et al. Spatial variability in crop response under contour
hedgerow systems in the Andes region of Ecuador. Soil Tillage
Res.
86
, 15–26 (2006).
47. Lugato, E., Lavallee, J. M., Haddix, M. L., Panagos, P. &
Cotrufo, M. F. Different climate sensitivity of particulate and
mineral-associated soil organic matter. Nat. Geosci.
14
,
95–300 (2021).
48. Borrelli, P. et al. Land use and climate change impacts on global
soil erosion by water (2015–2070). Proc. Natl Acad. Sci. USA
117
,
21994–22001 (2020).
49. Min, S.-K., Zhang, X., Zwiers, F. W. & Hegerl, G. C. Human
contribution to more-intense precipitation extremes. Nature
470
,
378–381 (2011).
50. Li, Z. & Fang, H. Impacts of climate change on water erosion: a
review. Earth Sci. Rev.
163
, 94–117 (2016).
51. Sutton, R. T. & Hawkins, E. ESD ideas: global climate response
scenarios for IPCC assessments. Earth Syst. Dynam.
11
, 751–754
(2020).
52. Walker, T. W. N. et al. Microbial temperature sensitivity and
biomass change explain soil carbon loss with warming. Nat. Clim.
Change
8
, 885–889 (2018).
53. Berner, R. A., Lasaga, A. C. & Garrels, R. M. The carbonate–
silicate geochemical cycle and its effect on atmospheric
carbon dioxide over the past 100 million years. Am. J. Sci.
283
,
641–683 (1983).
54. Gaillardet, J., Dupré, B., Louvat, P. & Allègre, C. J. Global silicate
weathering and CO
2
consumption rates deduced from the
chemistry of large rivers. Chem. Geol.
159
, 3–30 (1999).
Publisher’s note
Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds
exclusive rights to this article under a publishing agreement with
the author(s) or other rightsholder(s); author self-archiving of the
accepted manuscript version of this article is solely governed by the
terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to Springer Nature Limited
2023
Page 9
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Methods
Soil profile database analysis
Soil profiles with SOC and
137
Cs inventories from 20 agricultural catch-
ments in Asia, Europe and North America were compiled from the
literature (Extended Data Fig. 1). In total, 1,562 profiles including sta-
ble and erosional soils at upland hillslopes as well as colluvial soils at
footslopes were examined. The SRRs at the decadal scale were derived
from
137
Cs inventories using a conversion model that takes into account
factors such as plough layer depth and sampling year
21
. The effects of
soil redistribution on the SOC cycling were assessed via the following
set of metrics.
First, the C replacement ratio (CRR, %), which is the ratio of the
replaced C by the new C supplemented from photosynthesis to the
continuous C loss from the top soil layers by erosion, was calculated:
CRR
=
ΔVC
e
C
e
×
100
%
(1)
Where ∆VC
e
(kg C m
−2
) is the cumulative erosion-induced C uptake
from the atmosphere for the eroding profiles and C
e
(kg C m
−2
) is the
cumulative lost SOC by erosion.
Second, the CBE (%), which is the ratio between the preserved
organic C in buried sediments and the originally deposited SOC, was
calculated:
CBE
=
C
1
C
0
×
100
%
(2)
where C
1
(kg C m
−2
) is the preserved organic C in buried sediments and
C
0
(kg C m
−2
) is the originally deposited SOC.
Third, to assess the sensitivity of SOC stocks to soil erosion rate
variations, a linear regression between normalized SOC stocks and
SRRs was performed for each agricultural catchment:
SOC
n
−
1
=
a
×
R
(3)
Where SOC
n
is the normalized SOC stock being divided by the SOC
stock of reference positions not subjected to erosion or deposition, R
(cm yr
−1
) is the SRR with negative values indicating erosion and posi-
tive values indicating deposition and a (yr cm
−1
) is a coefficient. Steep
regression lines indicate high sensitivities of SOC stocks to soil redis-
tribution variations.
Fourth, to assess the sensitivity of CBE to deposition rate varia-
tions, a linear regression between CBE and
137
Cs-derived soil deposition
rate was performed for each site:
CBE
−
100
% =
b
×
D
(4)
Where CBE (%) is the CBE of the bulk deposited SOC, D (cm yr
−1
) is the
deposition rate and b (yr cm
−1
) is a coefficient.
A simple box model was used to estimate erosion-induced C
exchange for agricultural profiles
40
,
55
,
56
, which is based on the concept
that the SOC stock variation for a soil profile during a period is the result
of cumulative soil–atmosphere C exchanges and the cumulative lateral
SOC fluxes by erosion. The effects of temperature on the erosion–C
coupling were assessed by confronting various metrics with the mean
annual temperature of each catchment (Supplementary Methods).
Model simulations at the hillslope scale
Since temperature mediates both C inputs and decomposition, we used
a model that combines SOC cycling with soil redistribution processes
to disassemble the effects of C input and decomposition on erosion–C
coupling. The SOC cycling processes were represented using a one-pool
C model, which included SOC fluxes such as C input and decomposi-
tion. SOC decomposition was assumed to follow first-order kinetics.
Soil profiles in the model were represented as discrete layers and SOC
dynamics were simulated in each layer independently. Carbon inputs to
soils and SOC decomposition rates were set to decrease exponentially
with soil depth
57
,
58
. The model tracks the C depth profile evolution
caused by the soil removal in eroding areas and the soil accumulation
in depositional areas. The SOC content of each layer was updated by
mixing the soils from different layers of the previous time step. Details
of the model are provided in the Supplementary Methods. The model
was run in a series of scenarios with different C input and decomposi-
tion conditions at an ‘ideal’ hillslope with a standard soil redistribution
pattern to test how these temperature-mediated factors affect SOC
cycling in eroding landscapes.
Model estimations at the global scale
The variability of erosion-induced C exchanges between atmosphere
and land under different climatic conditions at the global scale was
estimated using existing datasets. Global estimates of agricultural
erosion
21
were combined with the SOC stock to estimate the lateral
SOC flux by erosion. The model tracked the mobilization and burial of
SOC from sources to sinks and the in situ SOC dynamics were simulated
using the above one-pool SOC cycling model under the condition of
soil redistribution. NPP was used as an approximate of SOC inputs.
Similarly, the sensitivity of SOC stock to soil redistribution variations,
C replacement ratio of laterally lost SOC by erosion in eroding areas,
CBE of deposited SOC in depositional areas and erosion-induced C
exchanges between land and atmosphere were assessed. Details of the
estimation are provided in the Supplementary Methods.
Model predictions under climate change projections
In this study, we used future climate projections generated by eight
global climate models (Supplementary Table 2) in four socioeconomic
and representative concentration pathway scenarios (Supplementary
Table 3). A space-for-time approach was applied to assess the variation
of the effects of erosion on SOC cycling in eroding croplands. Using
the exponential relationship between C cycling metrics (for exam-
ple, SOC stock, NPP and erosion) and climatic factors (temperature
and precipitation), we calculated the C cycling metrics under future
climatic conditions based on the C cycling metrics of present day and
the variation of climatic factors. With the derived SOC stocks, NPP
and erosion rate above, the effects of erosion on the SOC cycling on
croplands in various climate scenarios were assessed as described in
Supplementary Methods.
Data availability
The SOC stock of the top 1 m generated by the Harmonized World Soil
Database is available at
https://www.fao.org/soils-portal/data-hub/
soil-maps-and-databases/harmonized-world-soil-database-v12/en/
.
The SOC stock of the top 1 m generated by the Northern Circumpolar
Soil Carbon Database is available at
https://bolin.su.se/data/ncscd/
tiff.php
. The MODIS NPP dataset is available at
http://files.ntsg.umt.
edu/data/NTSG_Products/MOD17/GeoTIFF/MOD17A3/
. The land use
data generated by the History database of the Global Environment
(HYDE) is available at
https://dataportaal.pbl.nl/downloads/HYDE/
.
The mean annual temperature provided by the Climatic Research
Unit is available at
https://www.uea.ac.uk/web/groups-and-centres/
climatic-research-unit/data
. The mean annual precipitation provided
by the Global Precipitation Climatology Centre is available at
https://
opendata.dwd.de/climate_environment/GPCC/html/fulldata-monthly_
v2018_doi_download.html
. The MODIS potential evapotranspiration
(PET) data are available at
http://files.ntsg.umt.edu/data/NTSG_Prod-
ucts/MOD16/MOD16A3.105_MERRAGMAO/
. The mean annual tempera-
ture and mean annual precipitation as well as the surface soil moisture
(0–10 cm) of historical simulations and future projections generated
by global climate models of coupled model intercomparison project
(CMIP6) are available at
https://esgf-node.llnl.gov/projects/cmip6/
.
Page 10
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
The basemap data used to create figures were downloaded at
http://
naturalearthdata.com/downloads/
. The
137
Cs inventory and SOC stock
data are available from the corresponding author on request. Source
data are provided with this paper.
Code availability
The codes of C–erosion coupling model programmed in MATLAB are
available at
https://doi.org/10.5281/zenodo.7224539
(ref.
59
).
References
55. Quine, T. A. & Van Oost, K. Quantifying carbon sequestration as
a result of soil erosion and deposition: retrospective assessment
using caesium-137 and carbon inventories. Glob. Change Biol.
13
,
2610–2625 (2007).
56. Van Oost, K. et al. Legacy of human-induced C erosion and burial
on soil–atmosphere C exchange. Proc. Natl Acad. Sci. USA
109
,
19492–19497 (2012).
57. Van Oost, K. et al. Landscape-scale modeling of carbon cycling
under the impact of soil redistribution: the role of tillage erosion.
Glob. Biogeochem. Cycles
19
, GB4014 (2005).
58. Gerwitz, A. & Page, E. R. An empirical mathematical
model to describe plant root systems. J. Appl. Ecol.
11
,
773–781 (1974).
59. Wang, Z. Erosion-C model. Zenodo
https://doi.org/10.5281/
zenodo.7224539
(2022).
Acknowledgements
Z.W., G.T. and J.Q. are funded by the Natural Science Foundation of
China (grant no. 42171025 and 41971031). A.N. is funded by the MCIN/
AEI/10.13039/501100011033/ (grant no. PID2019-104857RB-I00 and
PID2019-103946RJ-100).
Author contributions
Z.W. and Y.Z. conceived the research, performed the analysis and
co-wrote the paper. G.G., G.T., T.A.Q., J.Q., A.N., H.F., Q.T. and K.V.O.
assisted in the interpretation of the results and commented on
the manuscript.
Competing interests
The authors declare no competing interests.
Additional information
Extended data
is available for this paper at
https://doi.org/10.1038/s41558-022-01562-8
.
Supplementary information
The online version contains supplementary
material available at
https://doi.org/10.1038/s41558-022-01562-8
.
Correspondence and requests for materials
should be addressed to
Zhengang Wang.
Peer review information
Nature Climate Change thanks Asmeret
Berhe, Julian Campo, Amaury Frankl, Priyanka Singh and the other,
anonymous, reviewer(s) for their contribution to the peer review of
this work.
Reprints and permissions information
is available at
www.nature.com/reprints
.
Page 11
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Extended Data Fig. 1 | Locations of the cropland sites in this study. The numbers in squares denote the identification of the catchment. Details of the study sites are
provided in Supplementary Table 1.
Page 12
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Extended Data Fig. 2 | Relationships between normalized SOC stocks
and soil redistribution rates for the cropland sites. Negative values of
soil redistribution rates indicate erosion, while the positive values indicate
deposition. The SOC stock of each soil profile is normalized by being divided by
the SOC stock of the reference sites (as denoted with red asterisks) not subjected
to erosion or deposition. The regression lines are set to pass through the
reference value so that the sensitivity of normalized SOC stocks to the variation
of soil distribution rates can be assessed by comparing the slopes of these linear
regression lines. The number in parentheses denote the identification of the
catchment, which is shown in Supplementary Table 1.
Page 13
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Extended Data Fig. 3 | SOC cycling in eroding croplands in scenarios of
various C input rates. a, Variation of SOC stocks with soil redistribution rates.
b, Variation of the normalized SOC stocks with soil redistribution rates. c,
Variation of replace ratios of lost SOC due to erosion in the eroding area with
soil redistribution rates and variation of burial efficiencies of deposited SOC in
the depositional area with soil redistribution rates.
d, Variation of the erosion-
induced C sink per unit lateral C flux with C input rates. In
a, b and c, the negative
values of soil redistribution rates indicate erosion, while the positive values
indicate deposition. In
b, the SOC stock of each soil profile is normalized by
being divided by the SOC stock of the reference sites not subjected to erosion
or deposition. For a hillslope with a given soil redistribution pattern, variations
in normalized SOC stocks, C replacement and burial efficiency and the C sink
strength per unit of lateral SOC flux are independent of the C input rate.
Page 14
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Extended Data Fig. 4 | SOC cycling in eroding croplands in scenarios
of various C decomposition rates. a, Variation of SOC stocks with soil
redistribution rates.
b, Variation of the normalized SOC stocks with soil
redistribution rates.
c, Variation of replace ratios of lost SOC due to erosion in the
eroding area with soil redistribution rates and variation of burial efficiencies of
deposited SOC in the depositional area with soil redistribution rates.
d, Variation
of the erosion-induced total C sink per unit lateral C flux with C decomposition
rates. In
a, b and c, the negative values of soil redistribution rate indicate erosion,
while the positive values indicate deposition. In
b, the SOC stock of each soil
profile is normalized by being divided by the SOC stock of the reference sites not
subjected to erosion or deposition.
Page 15
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Extended Data Fig. 5 | SOC cycling in eroding croplands in scenarios of
varying temperature. a, Variation of SOC stocks with soil redistribution rates.
b, Variation of the normalized SOC stocks with soil redistribution rates. c,
Variation of replacement ratio of lost SOC due to erosion in the eroding area
with soil redistribution rates and variation of burial efficiency of deposited
SOC in the depositional area with soil redistribution rates.
d, Variation of the
erosion-induced C sink per unit lateral C flux with temperature. In
a, b and c, the
negative values of soil redistribution rate indicate erosion, while the positive
values indicate deposition. In
b, the SOC stock of each soil profile is normalized
by being divided by the SOC stock of the reference sites not subjected to erosion
or deposition.
Page 16
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Extended Data Fig. 6 | Relationships between erosion and climatic factors in croplands derived from a global estimation of agricultural erosion
21
. a, Soil
erosion rate versus mean annual precipitation.
b, Mean annual precipitation versus mean annual temperature. c, Soil erosion rate versus mean annual temperature. d,
C erosion rate versus mean annual temperature.
Page 17
Nature Climate Change
Article
https://doi.org/10.1038/s41558-022-01562-8
Extended Data Fig. 7 | The effects of soil moisture on the relationships
between temperature and various SOC cycling metrics. a, The relationship
between soil moisture and the coefficient denoting the sensitivity of lateral SOC
fluxes caused by erosion to the variation of mean annual temperature. Details
of the regression between lateral SOC fluxes by erosion and temperature under
various soil moisture conditions are displayed in Supplementary Fig. 3.
b, The
relationship between soil moisture and the coefficient denoting the sensitivity of
SOC decomposition rates to the variation of mean annual temperature. Details
of the regression between SOC decomposition rates and temperature under
various soil moisture conditions are displayed in Supplementary Fig. 4.
c, The
relationship between soil moisture and the coefficient denoting the sensitivity
of the C sink strength per unit lateral SOC flux to the variation of mean annual
temperature. Details of the regression between the C sink strength per unit
lateral SOC flux and temperature under various soil moisture conditions are
displayed in Supplementary Fig. 5.
d, The relationship between soil moisture and
the coefficient denoting the sensitivity of the C sink strength per unit cropland
area to the variation of mean annual temperature. Details of the regression
between the C sink strength per unit cropland area and temperature under
various soil moisture conditions are displayed in Supplementary Fig. 6.
Document Outline