Jukuri, open repository of the Natural Resources Institute Finland (Luke) 
   
 
   
All material supplied via Jukuri is protected by copyright and other intellectual property rights. Duplication 
or sale, in electronic or print form, of any part of the repository collections is prohibited. Making electronic 
or print copies of the material is permitted only for your own personal use or for educational purposes.  For 
other purposes, this article may be used in accordance with the publisher’s terms. There may be 
differences between this version and the publisher’s version. You are advised to cite the publisher’s 
version. 
 
This is an electronic reprint of the original article.  
This reprint may differ from the original in pagination and typographic detail. 
 
Author(s): Tommaso Tadiello; Marco Acutis; Alessia Perego; Calogero Schillaci and Elena Valkama 
Title: Soil organic carbon under conservation agriculture in Mediterranean and humid 
subtropical climates: Global meta-analysis 
Year:  2023 
Version: Publisher’s version 
Copyright:    The author(s) 2023   
Rights:  CC BY-NC-ND 4.0  
Rights url: https://creativecommons.org/licenses/by-nc-nd/4.0/  
 
Please cite the original version: 
Tommaso Tadiello; Marco Acutis; Alessia Perego; Calogero Schillaci and Elena Valkama, 2023. Soil organic 
carbon under conservation agriculture in Mediterranean and humid subtropical climates: Global meta-
analysis. 74, e13338. https://doi.org/10.1111/ejss.13338  
S P E C I A L I S S U E P A P E R
Soil organic carbon under conservation agriculture
in Mediterranean and humid subtropical climates:
Global meta-analysis
Tommaso Tadiello1 | Marco Acutis1 | Alessia Perego1 |
Calogero Schillaci2 | Elena Valkama3
1DiSAA, Department of Agricultural and
Environmental Sciences, University of
Milan, Milan, Italy
2European Commission, Joint Research
Centre (JRC), Ispra, Italy
3Natural Resources Institute Finland
(Luke), Bioeconomy and Environment,
Sustainability Science and Indicators,
Jokioinen, Finland
Correspondence
Tommaso Tadiello, DiSAA, Department
of Agricultural and Environmental
Sciences, University of Milan, via
Celoria 2, Milan, Italy.
Email: tommaso.tadiello@unimi.it
Funding information
European Joint Programme SOIL,
SOMMIT, Grant/Award Number: 862695;
H2020 LANDSUPPORT, Grant/Award
Number: 774234; Natural Resources
Institute Finland (Luke)
Abstract
Conservation agriculture (CA) is an agronomic system based on minimum soil
disturbance (no-tillage, NT), permanent soil cover, and species diversification.
The effects of NT on soil organic carbon (SOC) changes have been widely stud-
ied, showing somewhat inconsistent conclusions, especially in relation to the
Mediterranean and humid subtropical climates. These areas are highly vulnera-
ble and predicted climate change is expected to accentuate desertification and,
for these reasons, there is a need for clear agricultural guidelines to preserve or
increment SOC. We quantitively summarized the results of 47 studies all around
the world in these climates investigating the sources of variation in SOC
responses to CA, such as soil characteristics, agricultural management, climate,
and geography. Within the climatic area considered, the overall effect of CA on
SOC accumulation in the plough layer (0–0.3 m) was 12% greater in comparison
to conventional agriculture. On average, this result corresponds to a carbon
increase of 0.48 Mg C ha
1 year
1. However, the effect was variable depending
on the SOC content under conventional agriculture: it was 20% in soils which
had ≤ 40 Mg C ha
1, while it was only 7% in soils that had > 40 Mg C ha
1. We
proved that 10 years of CA impact the most on soil with SOC ≤ 40 Mg C ha
1.
For soils with less than 40 Mg C ha
1, increasing the proportion of crops with
bigger residue biomasses in a CA rotation was a solution to increase SOC. The
effect of CA on SOC depended on clay content only in soils with more than
40 Mg C ha
1 and become null with a SOC/clay index of 3.2. Annual rainfall
(that ranged between 331–1850 mm y
1) and geography had specific effects on
SOC depending on its content under conventional agriculture. In conclusion,
SOC increments due to CA application can be achieved especially in agricultural
soils with less than 40 Mg C ha
1 and located in the middle latitudes or in the
dry conditions of Mediterranean and humid subtropical climates.
Received: 25 January 2022 Revised: 7 December 2022 Accepted: 20 December 2022
DOI: 10.1111/ejss.13338
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any
medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2022 The Authors. European Journal of Soil Science published by John Wiley & Sons Ltd on behalf of British Society of Soil Science.
Eur J Soil Sci. 2023;74:e13338. wileyonlinelibrary.com/journal/ejss 1 of 22
https://doi.org/10.1111/ejss.13338
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Highlights:
• The results of 47 studies were quantitively summarized by using a meta-
analysis
• SOC accumulation due to CA was 12% greater compared to conventional
agriculture
• SOC increment due to CA can reach 20% in soils having less than 40 Mg
C ha
1
• The impacts of pedo-climatic factors and agronomic management practices
were studied
KEYWORD S
C sequestration, conservation agriculture, Mediterranean and humid subtropical climates,
meta-analysis, no-till, SOC
1 | INTRODUCTION
Soil organic carbon (SOC), the major component of soil
organic matter, has a great impact in all soil processes.
SOC dynamics are regulated by climatic variables, geo-
graphical characteristics, soil physico-chemical properties,
quantity and quality of C inputs into the soil and manage-
ment practices, as well their interactions (Haddaway
et al., 2017; Lal, 2004; Lorenz & Lal, 2018; Ogle
et al., 2019; Paul et al., 1997; West & Post, 2002). Playing
an important role in modifying the SOC dynamics
(Lorenz & Lal, 2018), conservation agriculture (CA) is
being promoted by the FAO as an approach for achieving
sustainable land management, environmental protection,
and climate change adaptation and mitigation (Pisante
et al., 2015). CA utilizes three agronomic principles:
(1) minimum soil disturbance, avoiding soil inversion, that
is, no-tillage (NT) or minimum tillage or vertical tillage;
(2) permanent soil cover that is guaranteed by retaining
crop residues or by cover crop adoption; and (3) the inte-
gration of crop rotations involving at least three different
crops (FAO, 2017b).
Since the 2000s, the effects of NT on SOC changes
have been summarized in nine meta-analyses conducted
mostly on a global scale, but with a few on regional or
national scales (Table 1). Most of them revealed a range
of SOC sequestration increases attributable to NT ranging
from 8 to 18% in the 0–0.25/0.3 m soil layer. In contrast,
for temperate climates, Mondal et al. (2020) reported a
40% and a 5% of SOC increases in the 0–5 and 5–10 cm
soil depths, respectively, while there was a significant
decrease in deeper soil layers, resulting in a slight overall
increase of 1.1% in the 0–0.6 m soil layer (Table 1). Hadd-
away et al. (2017) found a SOC sequestration almost
twice as large as that reported by Luo et al. (2010) and by
West and Post (2002). Moreover, Li et al. (2020) reported
that conservation tillage practices increased SOC stock in
humid or perhumid (Thornthwaite, 1948) climate condi-
tions, but not in semi-humid. Mondal et al. (2020)
highlighted that in tropical and subtropical climates, the
effect was positive and significant only up to 10-cm
depth, whereas in temperate climates, changes were sig-
nificant but negative further down the profile. However,
Sun et al. (2020) stated that CA adoption is recommended
in arid regions, while other meta-analyses did not report
a significant effect of climate (Haddaway et al., 2017; Luo
et al., 2010; Virto et al., 2012).
In addition, different meta-analyses demonstrated
inconsistent outcomes for the effects of soil texture on
SOC accumulation under NT management (Table 1). For
example, the effect of NT varied from study to study:
increasing effects were only reported in silty and sandy
soils within the 0–0.3 m layer (Li et al., 2020), or in loamy
and clay soils within the 0–0.1 m layer (Mondal
et al., 2020). Sometimes, there were no effects of the tex-
ture class (Haddaway et al., 2017; Sun et al., 2020).
On the other hand, the impact of carbon input into
soil, which has been considered as a key factor involved
in SOC accumulation, was confirmed by many meta-
analyses (Table 1). It is also common to refer to the effect
of experiment duration to explain the variability of SOC
responses across the studies. In this case, five out of nine
meta-analyses demonstrated the importance of the NT
management duration on SOC accumulation, finding
greater effects when the duration of the experiment was
longer. Nevertheless, Luo et al. (2010) and Sun et al.
(2020) found no consistent relationships with the dura-
tion of NT practice.
The reason for such inconsistencies mainly stems
from the different methodologies or even erroneous
approaches applied in the nine meta-analyses. For
instance, three meta-analyses included studies in which
2 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
T
A
B
L
E
1
D
es
cr
ip
ti
on
of
m
et
a-
an
al
ys
es
st
ud
yi
n
g
th
e
ef
fe
ct
s
of
n
o-
ti
lla
ge
(N
T
)
on
so
il
or
ga
n
ic
ca
rb
on
(S
O
C
)
st
oc
k
co
m
pa
re
d
to
co
n
ve
n
ti
on
al
ti
lla
ge
pu
bl
is
h
ed
du
ri
n
g
th
e
la
st
20
ye
ar
s
L
i
et
al
.(
20
20
)
Su
n
et
al
.(
20
20
)
M
on
d
al
et
al
.(
20
20
)
A
gu
il
er
a
et
al
.(
20
13
)
H
ad
d
aw
ay
et
al
.(
20
17
)
L
u
o
et
al
.(
20
10
)
W
es
t
an
d
P
os
t
(2
00
2)
V
ir
to
et
al
.(
20
12
)
A
n
ge
rs
an
d
E
ri
k
se
n
-
H
am
el
(2
00
8)
O
ri
gi
n
gl
ob
al
gl
ob
al
gl
ob
al
M
ed
it
er
ra
n
ea
n
gl
ob
al
m
os
tl
y
U
SA
an
d
C
an
ad
a
gl
ob
al
gl
ob
al
gl
ob
al
T
yp
e
of
ef
fe
ct
si
ze
lo
g
re
sp
on
se
ra
ti
o
lo
g
re
sp
on
se
ra
ti
o
lo
g
re
sp
on
se
ra
ti
o
lo
g
re
sp
on
se
ra
ti
o
m
ea
n
di
ff
er
en
ce
m
ea
n
di
ff
er
en
ce
m
ea
n
di
ff
er
en
ce
n
on
-s
ta
n
da
rd
m
et
ri
cs
n
on
-s
ta
n
da
rd
m
et
ri
cs
N
o
ti
ll
ov
er
al
le
ff
ec
ta
+
13
%
(§
3.
1)
b
+
8%
(F
ig
u
re
4)
+
1.
1%
(F
ig
ur
e
S6
)
+
18
.2
%
(§
3)
+
16
%
(F
ig
ur
e
35
)b
+
2.
8%
(p
.2
28
)
+
16
%
(p
.1
94
3)
+
7%
(p
.2
1)
+
5%
(p
.1
37
3)
So
il
de
pt
h
ra
n
ge
(m
)
0–
0.
3
0–
0.
3
0–
0.
6
0–
0.
25
0–
0.
3
0–
1
0–
0.
3
0–
0.
3
0–
0.
9
C
ro
ps
h
er
ba
ce
ou
s
cr
op
h
er
ba
ce
ou
s
cr
op
ce
re
al
an
d
le
gu
m
e
ce
re
al
s,
h
or
ti
cu
lt
ur
e
h
er
ba
ce
ou
s
cr
op
h
er
ba
ce
ou
s
cr
op
h
er
ba
ce
ou
s
cr
op
h
er
ba
ce
ou
s
cr
op
m
ai
n
ly
ce
re
al
B
u
lk
de
n
si
ty
m
ea
su
re
d
pe
do
tr
an
sf
er
m
ea
su
re
d
pe
do
tr
an
sf
er
m
ea
su
re
d
pe
do
tr
an
sf
er
m
ea
su
re
d
m
ea
su
re
d
m
ea
su
re
d
E
ff
ec
t
of
m
od
er
at
or
s
C
lim
at
e
n
o
ef
fe
ct
in
se
m
i-
h
um
id
cl
im
at
e
(F
ig
u
re
7)
po
si
ti
ve
ef
fe
ct
in
dr
y
re
gi
on
(F
ig
ur
e
4)
th
e
ef
fe
ct
de
pe
n
de
n
t
on
cl
im
at
e
an
d
so
il
de
pt
h
in
te
ra
ct
io
n
(T
ab
le
3)
-
n
o
ef
fe
ct
(p
.2
2)
n
o
ef
fe
ct
(F
ig
ur
e
5a
,b
)
-
n
o
ef
fe
ct
(F
ig
ur
e
1a
–c
)
-
So
il
po
si
ti
ve
ef
fe
ct
w
it
h
sa
n
dy
an
d
si
lt
y
so
ils
(F
ig
ur
e
6)
n
o
ef
fe
ct
of
cl
ay
an
d
sa
n
d
(F
ig
u
re
1)
lo
am
y
an
d
cl
ay
te
xt
ur
e
po
si
ti
ve
ly
af
fe
ct
s
on
ly
0–
0.
1
m
la
ye
r
(T
ab
le
3)
-
n
o
ef
fe
ct
(p
.2
2)
-
-
n
o
ef
fe
ct
(F
ig
ur
e
1d
)
-
C
in
pu
t
po
si
ti
ve
ef
fe
ct
re
ga
rd
le
ss
of
ti
ll
ag
e
in
te
n
si
ty
(§
3.
1)
th
e
m
os
t
im
po
rt
an
t
m
an
ag
em
en
t
fa
ct
or
(d
ec
is
io
n
tr
ee
,F
ig
u
re
1)
ef
fe
ct
on
ly
in
up
pe
r
la
ye
rs
(§
4)
po
si
ti
ve
(§
4)
-
-
-
po
si
ti
ve
(F
ig
ur
e
2)
-
E
xp
er
im
en
t
du
ra
ti
on
po
si
ti
ve
(§
3.
4)
n
o
ef
fe
ct
(F
ig
ur
e
1)
po
si
ti
ve
ef
fe
ct
on
ly
0–
0.
1
m
la
ye
r
(T
ab
le
3)
-
po
si
ti
ve
(p
.2
2)
n
o
ef
fe
ct
(F
ig
ur
e
5c
)
pe
ak se
qu
es
tr
at
io
n
ra
te
s
in
5
to
10
yr
(F
ig
ur
e
5)
-
po
si
ti
ve
(F
ig
ur
e
2)
N

of
st
u
di
es
c
n
ot
de
cl
ar
ed
fo
r
SO
C
un
de
r
n
o
ti
ll
(2
43
in
to
ta
l)
11
5
n
ot
de
cl
ar
ed
fo
r
SO
C
un
de
r
n
o
ti
ll
(5
22
in
to
ta
l)
33
29
69
74
37
24
a R
el
at
ed
to
th
e
se
le
ct
ed
so
il
de
pt
h
ra
n
ge
.
b
E
ff
ec
t
si
ze
is
w
ei
gh
te
d
by
in
ve
rs
e
va
ri
an
ce
.
c N

of
st
u
di
es
th
at
ev
al
ua
te
d
th
e
ef
fe
ct
s
of
n
o-
ti
ll
ag
e
on
SO
C
st
oc
k
co
m
pa
re
d
to
co
n
ve
n
ti
on
al
ti
lla
ge
.
TADIELLO ET AL. 3 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
bulk densities were not originally measured but were
estimated from pedotransfer functions to compute SOC
stocks from SOC concentrations. The potential uncer-
tainty which can arise by applying a pedotransfer func-
tion developed in a particular area and which is then
applied on different sites can seriously impact the final
results. In fact, in recent works, Schillaci et al. (2021) and
Nasta et al. (2020) tested the performance in predicting
bulk density (BD) of new and previous developed pedo-
transfer functions, respectively. They found that the asso-
ciated error (RMSE %, Root Mean Square Error of
prediction) of the pedotransfer function ranged on aver-
age between 19% and 26%, respectively.
This review of the existing meta-analyses highlighted that
the climate effect is highly variable or influences the SOC
sequestration in interaction/addition with other factors. In
addition, it has been highlighted a knowledge gap for a spe-
cific climate, in particular, for areas characterized by mild
winters and hot summers (Bouma, 2005; Hernandez-Ochoa &
Asseng, 2018), since only one meta-analysis summarising
33 studies on herbaceous crops was conducted in similar cli-
matic conditions (Aguilera et al., 2013). This type of climate
is found in areas mainly characterized by temperate and
Mediterranean climates and it can be identified with the Cfa,
Csa, Csb Köppen sub-types areas (Bouma, 2005; Kottek
et al., 2006). These areas belong to the warm temperate cli-
mates and they are all univocally identified as areas with the
temperature of the coldest month (Tmin) ranging between 
3
C < Tmin < +18 C, and a hot (where Tmax, the monthly
mean temperatures of the warmest month is Tmax > +22C)
or warm (where Tmon, the mean monthly temperature is
Tmon ≥ +10C) summer (Kottek et al., 2006). This characteri-
zation identifies areas with high SOC mineralization rates
due to the high temperature during the summer season
(A´lvaro-Fuentes & Paustian, 2011; Pravalie et al., 2021), and
with soils having quite low (usually between 0.7% and 1%
SOC content depending on the textural class) SOC content
(FAO, 2017a; Jones et al., 2005). Thus, these areas are highly
vulnerable and predicted climate change is expected to accen-
tuate human-induced desertification processes like intensive
use of agricultural lands, poor irrigation practices, and defor-
estation (Ruiz et al., 2020; Spinoni et al., 2015; Underwood
et al., 2009). In addition, an increase of extreme events and
especially drought is expected in parts of the Mediterranean
area and even in some humid areas (Mihailescu & Bruno
Soares, 2020) where agricultural practices are also affecting
soil fertility since the production system (based on winter
wheat and maize, soybean and sunflower during summer) is
based on intensive traditional plough-based crop production
systems (Mazzoncini et al., 2016).
Due to the similarities of these climate regions and
their common problems, it is important to treat these
threatened areas as a whole for a broader comprehension
of the climate change impact on SOC stocks. Thus, to
update the findings and define clear guidelines under this
specific climate condition, the present work aims to con-
duct a robust meta-analytic approach that highlights the
possibility of CA adoption to mitigate soil C depletion. A
strong scientific comprehension of CA practices is needed
to support the new policies that will be applied in future.
For instance, the new EU common agricultural policy
(2023–2027) will consider CA as a potential agricultural
practice to increase carbon (European Commission, 2021).
Therefore, the present study aims to summarize stud-
ies on the effects of CA on SOC sequestration capability
in the plough layer (0–0.3 m) in Mediterranean and
humid subtropical climates (from all over the world) by
using a weighted meta-analysis. We used a rigorous
approach that relies on including studies with measured
bulk density (BD) (when carbon stock was not already
reported) and utilising no pedotransfer functions to com-
pute the carbon stock. We examined the sources of varia-
tion in SOC responses to CA across the studies, such as
climate, geography, soil characteristics (clay and sand
content, pH, SOC/clay index, and SOC stock amount
under conventional tillage), and agricultural manage-
ment (nitrogen (N) fertilization levels, duration of CA
practice, crop diversification, proportion of high-residue
crops in rotation, and legume presence in the rotation).
2 | MATERIALS AND METHODS
2.1 | Systematic search and data
extraction from literature
The database creation for the final statistical analysis
involved three different steps: (1) primary studies collec-
tion from different online database resources, (2) selection
of studies with several inclusion criteria to match the
research purpose, and (3) data extraction.
2.1.1 | Studies collection
All the screening process followed the PRISMA checklists
(Page et al., 2021) for evaluating research synthesis in sys-
tematic reviews and meta-analysis. Figure S1 reports the
flow diagram of literature search and screening adapted
from the PRISMA checklists (Page et al., 2021).
We found the articles by searching for keywords with
a nested query (Supplementary file 1) in Web of Science
and Scopus databases. The query was based on four differ-
ent parts related to CA, conventional tillage, SOC, and the
list of the 67 countries that belong to Mediterranean and
humid subtropical climates (i.e. that have at least one
4 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
correspondence with the Cfa, Csa, and Csb sub-types) of
the Köppen classification (Chen & Chen, 2013; Peel
et al., 2007). Finally, the Bibliometrix R package (Aria &
Cuccurullo, 2017) was utilized to merge all the results in a
unique dataset (n = 911), excluding the duplicated studies.
After this first step, the resulting 911 studies were evalu-
ated with a first screening that allowed to exclude a large
part of the database. There were many reasons why this
first screening was needed, for instance, because studies
were not peer-reviewed or belonged to modelling studies.
The outcome of this first screening ended up with 186 stud-
ies that were further screened based on the inclusion cri-
teria listed in the next paragraph. The final database used
for the meta-analysis was composed of 47 studies.
2.1.2 | Inclusion criteria
To be included in the final database, a study had to meet
the following criteria:
1. the study was conducted on herbaceous field crops;
2. the study coordinates belong to Cfa, Csa, or Csb;
3. the study had an appropriate control group (conven-
tional agriculture): inversion/mixing tillage (mould-
board/disk ploughing, disk harrow or chisel ploughing)
in spring, autumn or in both, residues incorporated and
no cover crop utilization. Within the single study, the
rotation with the least number of crops was selected;
4. the study had an appropriate treatment group (CA):
no tillage management, residues retained on the top
of the soil (chopped or not), and with or without cover
crops. Within the single study, the rotation with the
largest number of crops was selected;
5. the study assessed the effect of CA on SOC stock or
concentration in the plough layer reported either for a
single soil layer (e.g. 0–0.30 or 0–0.2 m) or for multiple
soil layers (e.g. 0–0.15 and 0.15–0.30 m);
6. along with SOC concentration, the study reported BD
measured separately for control and treatment;
7. at the end of experiment, SOC was recorded as means for
treatment (CA) and control (conventional agriculture),
with sample sizes and standard deviations (SD) or stan-
dard errors (SE), or statistical analysis references (e.g. P
[F] or LSD value from the ANOVA table) to compute SD.
2.1.3 | Data extraction
The data extraction method is crucial to deal with the
non-independence of the observations that can lead to
underestimates of the standard error of the mean effect
and, therefore, liberal evaluations of the statistical
significance of effects (Gurevitch & Hedges, 1999;
Nakagawa et al., 2017).
To avoid problems with non-independence of the effect
sizes, only one pair comparison corresponding to the last
sampling date was extracted from a study. If an article
reported results from different experimental sites with differ-
ent pedo-climatic characteristics, those sites were considered
as independent studies and were included in the database.
However, if several articles referred to the same experimental
site with the same pedological characteristics, the article with
the longest experimental durationwas chosen.
Several articles treated factorial experiments, in
which tillage treatments were studied in combination
with different fertilization or cover crops. In the case of
different fertilization levels, the second one from the top
was chosen for both control and treatment. Legume cover
crops were selected as a first choice when available.
Data were extracted from tables and digitized from fig-
ures using WebPlotDigitizer software (Rohatgi, 2020). SE
were converted to SD (SE¼ SDffiffi
n
p where n is the number of
replicates) where necessary. When no measure of variability
was provided,we extracted the SD from the ANOVA table using
the EX-TRACT tool (Acutis et al., 2021; Acutis et al., 2022).
This tool allows the estimation of the experimental error
(i.e. standard deviation and standard error of treatments
mean) associated to statistical analysis results of published
articles (i.e. estimated from the LSD, P[F] values, or even
from the letters assignment indicating differences among
means based on the results of amultiple comparison test).
2.2 | Database creation
The final database used for this study consisted of 47 stud-
ies published in 41 articles in peer-reviewed scientific
journals (Table S1, Figure S2). Entire database for meta-
analysis is available in Zenodo (Database 1, https://doi.
org/10.5281/zenodo.7404592).
No restrictions were set about the articles’ publication
date: those selected were published between 1998 and
2020. The final number of studies was only 5% of that
obtained by searching for keywords.
Studies were located in North America (n = 19),
South America (n = 9), Europe (n = 10), Asia (n = 8),
and Africa (n = 1) between 23 and 36 S and 19 to 45
N of latitude (Table S1, Figure S2). The soils mainly
belonged to the clay, loam, and silt loam texture classes,
and annual precipitation ranged from 331 to 1850 mm.
The major climate sub-types were Cfa (n = 33) and Csa
(n = 13), while only one study referred to Csb.
The soil management of the controls included differ-
ent soil inversion techniques of which the fall/spring
moldboard ploughing was the most frequent (71% of the
TADIELLO ET AL. 5 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
studies). In all cases, control and treatments included
nitrogen fertilization, weed control (without soil mechan-
ical disturbance for the treatments), and no grazing.
These main agronomic features were kept the same dur-
ing the entire duration of experiments, ranging from 2 to
51 years. At least three different crops in the rotation of
the treatment were reported in 11 studies, and monocrop-
ping in nine studies. Four studies did not report any
information on crop rotation.
Most of the studies (39) report the BD while the stan-
dard deviation was not always reported: 27 studies did
not report SD or SE. For the remaining articles (20), no
additional computation was required to obtain a measure
of variability.
2.3 | SOC computation
The results for SOC changes in the plough layer for con-
trols and treatments were reported as stock (Mg ha
1 or
kg m
2) in 37 studies or as concentration (g kg
1 or %)
and BD that were converted to stock in 10 studies. Stud-
ies that reported C concentrations with no measured BD
were excluded from the final database. Moreover, to
avoid false computation due to ignoring the differences
in soil BD between treatments (Du et al., 2017; Toledo
et al., 2013), only studies with BD measured separately
for treatment and control were considered.
SOC was reported for a single topsoil layer (e.g. 0–0.30
or 0–0.2 m) in 29 studies and for multiple soil layers
(e.g. 0–0.15 and 0.15–0.30 m) in 18 studies (Database
2, https://doi.org/10.5281/zenodo.7404592).
Nowadays, to assess different agronomic practices, it is
usually required to report SOC as mass per unit area
(Mg ha
1 or kg m
2) for a single soil layer (e.g. 0–0.3) and
the associated standard deviation. Nevertheless, sometimes,
data were not directly reported as mass per unit area, or
multiple layers were included. To deal with these kinds of
data, we developed a specific methodology to compute a
mean of SOC for a single soil layer and its SD from a prod-
uct of two normally distributed variables (C concentration
and BD, being correlated variables) or from a multiple cor-
related layer sum, when SOC results for several layers were
reported (Tadiello et al., 2022). This method allows consid-
ering the correlation coefficient between C concentration
and BD or/and between multiple sub-soil layers for a better
estimation of the SD of the single stock soil layer.
Within the countries considered in this study, the
total CA area under the Cfa, Csa, and Csb Köppen cli-
mate zones has been retrieved from Chen and Chen
(2013). Then, within this area, the difference between the
total agricultural land and the agricultural land under
CA areas has been used to compute the potential increase
of C sequestration if all the agricultural land were under
CA. The complete dataset and statistical approach are
reported in Table S2.
2.4 | Explanatory variables
To explain the variation in SOC stock due to the CA
application in the plough layer, we included pedo-
climatic and management-related explanatory variables
(moderators) listed in Table 2. Latitude and longitude
were expressed as decimal degrees; latitude moderator
was expressed as the absolute value (e.g. 30 and 
 30
refer to the same latitude, indicating an equal distance
from the equator).
If annual average annual rainfall or temperature
were missing in a study, we used the value available
from World Bank Group (World Bank Group, Climate
change knowledge Portal, 2021). When only the tex-
tural class of the soil was available, we used the central
values of clay and sand of the given textural class as
continuous moderators. The pH was extracted from the
articles in 24 out of 47 articles. When not reported, the
World Soil Information Service WoSIS (https://www.
isric.org/explore/wosis) was used to retrieve data. The
SOC/clay index was computed between the SOC stock
under control and the clay percentage. This kind of
index has been widely adopted (as a ratio between SOC
and clay concentrations in Prout et al., 2022; Prout
et al., 2021) to describe soil physical properties such as
bulk density, water retention characteristics or clay dis-
persibility rather than using the SOC or clay total con-
tents (Dexter et al., 2008).
Based on the SOC in control (SOCctrl) in the plough
layer (0–0.3 m) measured at the end of the experiment,
we created two different study groups (i.e. “SOC ≤ 40 Mg
ha
1” and “SOC > 40 Mg ha
1” groups). A single study
(x) was then assigned to one of the groups following this
formula:
if SOCctrl=layer depthð Þ 0:3 ≤ 40Mgha
1 Then x ” SOC ≤ 40” group
if SOCctrl=layer depthð Þ0:3 > 40Mgha
1 Then x ”SOC> 40” group
(
6 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
where layer depth is expressed in metres.
The threshold was selected based on the paper by
Lugato et al. (2014), who reported that, in the Mediterra-
nean area, frequently the topsoil SOC stock values were
below 40 Mg C ha
1.
We included the number of crops in a treatment rota-
tion as an explanatory variable because the presence of at
least three different crops is one of the three CA princi-
ples defined by FAO (2016). Moreover, we included the
proportion of crops with high residue biomass (grain
maize, sorghum, cotton, and rice) in a rotation as an indi-
cator of C input to the soils.
2.5 | Meta-analysis
Meta-analyses were conducted in R (R Core Team, 2018)
with the “metafor” package (Viechtbauer, 2010, see Sup-
plementary file 2 for the code used).
Quantitative meta-analysis involves calculating an
effect size (i.e. the magnitude of the treatment effect) that
can be averaged across independent studies. Since two
experimental groups have been compared, the response
ratio (r) was computed for the response variables as an
index of the effect size (Gurevitch et al., 2018):
r¼XCA=XC ð1Þ
where XCA and XC represent the means for treatments
(CA) and for control (conventional agriculture), respec-
tively, averaged for experimental replicates or samples.
Since the distribution of r is skewed, performing sta-
tistical analyses in the metric of the natural logarithm of
r is usually preferred due to its much more normal dis-
tribution in small samples than that of r (Hedges
et al., 1999):
ln rð Þ¼ ln XCA=XC
 ¼ ln XCA 
 ln XC  ð2Þ
We calculated the variance of ln(r) as:
V ln rð Þ ¼ SDCAð Þ
2
nCA XCA
 2þ SDCð Þ2
nC XC
 2 ð3Þ
where SDCA and SDC are the corresponding standard
deviations, and n is the sample size.
We assumed that studies do not share the same effect
sizes and consequently, we used a random effects model to
combine estimates across the studies. The application of this
kind of model accounts for experimental method
TABLE 2 Categorical and continuous explanatory variables (moderators) included in the meta-analysis
Type Moderator Group or range
Climate Köppen classification Cfa, Csa, and Csb
Rainfall (mm year
1) 331–1850
Average annual temperature (C) 14–25
Geography Continent Asia, Africa, Europe, North
America, and South America
Longitude (degree) 
121–139
Latitude (degree absolute value) 19–45
Soil SOC stock under conventional agriculture (Mg ha
1) 18–102
SOC stock under conventional agriculture (Mg ha
1) SOC ≤ 40, SOC > 40a
Clay (%) 7–76
Sand (%) 3–78
pH (H2O) 4.5–8.5
SOC/clay index 0.3–7.1
Agronomic management N fertilization level (kg N year
1) 0–390
Experiment duration (years) 2–51
Number of crops in treatment rotation < 3, ≥ 3
Proportion of crops with high residue biomass (%)b 0–100
Legumes presence yes, no
aBased on Lugato et al. (2014).
bGrain maize, sorghum, cotton, and rice.
TADIELLO ET AL. 7 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
differences between studies (that are considered only a ran-
dom sample of possible effect sizes) which may introduce
variability (“heterogeneity”, τ2) among the true effects.
We calculated the weighted mean of the log response
ratio for all studies as:
ln rð Þ¼
Pn
i¼1
wi lnriPn
i¼1
wi
ð4Þ
where lnri is the log response ratio for study i, n is the
number of studies, and wi is the weight for study i, defined
as (Koricheva et al., 2013; Borenstein et al., 2009):
wi¼ 1Viþ τ2 ð5Þ
where Vi is the variance of the study i and τ2 denotes the
amount of residual heterogeneity (between-study vari-
ance). Because the variance of the effect sizes is a func-
tion of the sample size (Equation 3), studies with a larger
sample size had lower variances and received heavier
weights.
The τ2 parameter is considered the variance of the
true effect size. Since is it not possible to compute it from
the entire population of the effect size, the τ2 is an esti-
mation from the observed effect:
τ2¼ Q
dfð Þ
C
ð6Þ
where
Q¼
Xk
i¼1
wi Y i
Mð Þ2,df ¼ n
1,C¼
X
wi

P
W 2iP
Wi
where wi is the study weight, Yi is the study effect size, M
is the summary effect, and n is the number of studies. The
τ2 coefficient is in the same metric (squared) as the effect
size itself and reflects the absolute amount of variation in
that scale (Borenstein et al., 2009). To describe the distri-
bution of the effect size, it is more useful to use its “stan-
dard deviation” measurement expressed as:
τ¼
ffiffiffiffi
τ2
p
ð7Þ
that is on the same scale as the effect size itself but, while
τ2 is a squared value, τ is not.
The rma function has been used to compute the ran-
dom model and the maximum-likelihood estimator
(“ML”) to estimate the amount of heterogeneity
(Raudenbush, 2009). When a moderator was taken into
account in the model to explain at least part of the total
heterogeneity, a mixed-effect model was fitted.
The Cochran's Q-test (Hedges & Olkin, 1985) was
used to test the null hypothesis H0 : τ2= 0 that examines
the between-group heterogeneity, while an omnibus test
(that excludes the intercept β0Þ of all the model coeffi-
cients was conducted when moderators were included in the
model (test of moderator, model [QM] heterogeneities).
Weighted meta-regressions were run to study the effect of
continuous explanatory variables, with ln(r) as the dependent
variable and the continuous variables as independent ones.
For the outliers' identification, we used the backward
search algorithm specifically developed for meta-analysis
(Mavridis et al., 2017). Backward search algorithms start
with the full dataset and remove sequentially outlying
observations until all outliers have been removed. This
method can be useful when there are a few outlying stud-
ies (Mavridis et al., 2017).
Moreover, the descriptive I2 statistic (%) was reported
for the overall effect size. This coefficient is useful to
explain the estimated amount of heterogeneity as the
inconsistency across the studies (Borenstein et al., 2009),
and it is expressed as the ratio between the true heteroge-
neity and the total variance:
I2¼ τ
2
τ2þVY 100 ð8Þ
where VY is the within study variance.
As an additional parameter, the metafor package also
computed the R2 statistic (Raudenbush, 2009) as the ratio:
R2¼bτ2REþbτ2MEbτ2RE ð9Þ
where bτ2RE refers to the random model τ2 (total amount of
heterogeneity) while bτ2ME to the estimated value of τ2
based on the mixed-effect model. The R2 coefficient
defines the amount of heterogeneity accounted for by the
moderator inclusion in the model. This coefficient does
not take into account the within-study variance and, for
this reason, it cannot be compared to the classical R2
referred to OLS (ordinary least square) regression.
Results were back-transformed, except for meta-
regression, and reported in the text and figures as per-
centage changes from the controls:
Response %ð Þ¼ EXP ln rð Þð Þ
1½ 100 ð10Þ
The percentage difference between the control and the
treatment is a straightforward way to show the increment/
decrement of SOC due to the CA technique.
8 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
The p-value and the 95%CI were used to identify the
significant effect of continuous and categorical modera-
tors, respectively (Hedges et al., 1999).
To detect possible publication bias in the meta-analy-
sis, we first used a graphical method based on two funnel
plots (Nakagawa et al., 2017; Nakagawa et al., 2021;
Sterne & Egger, 2001). The x-axis displays the ln(r), and
the y-axis is the sample size and the standard error,
respectively, in the two funnel plots. When the standard
error (SE) was used as the vertical axis, it had the zero
FIGURE 1 Forest plot showing the results of 47 studies examining the effect of conservation agriculture (CA) on SOC sequestration.
The diamonds are centred on the summary effect, which was estimated for the “SOC ≤ 40 Mg ha
1” and “SOC > 40 Mg ha
1” groups
separately. The overall effect diamond (n = 47) is also displayed at the bottom of the plot. Lateral tips of diamonds represent the 95%
confidence intervals. Numbers in the right-hand column are summary effect estimates for each study [lower 95% CI, upper 95% CI]. Dotted
vertical line indicates conventional agriculture (control)
TADIELLO ET AL. 9 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
placed at the top (i.e. standard error 0 at the top). In this
way, the largest studies have the smallest SE, and they
are placed at the top of the graph. When present, the
diagonal lines show the expected 95% confidence inter-
vals around the summary effect. Moreover, we checked
the funnel plots asymmetry with the Egger's regression
test for the mixed-effects model reported by Viechtbauer
(2010) and implemented in the “regtest” function.
To assess the robustness of the observed effects, the
fail-safe number (Nfs) has been computed for both “SOC
≤ 40 Mg ha
1” and “SOC > 40 Mg ha
1” groups to esti-
mate the number of non-significant, unpublished, or miss-
ing studies need to be added to a meta-analysis to change
its results from significant to non-significant. Specifically,
we used the Rosenthal method (Rosenthal, 1979) that esti-
mates how many missing studies we would need to
retrieve and incorporate in the analysis before the p-value
became non-significant (Borenstein et al., 2009).
3 | RESULTS
3.1 | Overall effect and SOC stock
amount
For the entire database, the CA effects on SOC was
highly variable, ranging from 
9% to 99% compared to
the controls (Median of 43 Mg C ha
1), with a weighted
summarized effect of 12% (95% CI 8%–17%, n = 47,
Figure 1). The summarized effect can be translated to an
annual carbon stock increase of 0.48 Mg C ha
1 y
1,
based on the weighted average duration of experiments
of 11 years.
Across 47 studies, the SOC stock under conventional
agriculture (control) were highly variable, ranging
between 18 and 102 Mg C ha
1 in the plough layer.
Unequal controls may cause a “noise” in meta-analysis,
confounding the effects of explanatory variables, such as
pedo-climatic factors and management practices. There-
fore, we ran a weighted meta-regression between SOC
stock and the response to CA. The meta-regression indi-
cated that increasing SOC stock reduced linearly the
response to CA, and an increase of 1 Mg ha
1 in SOC
stock was associated with a 0.22% decrease in the
response (R2 = 13.4; QM = 5.98, p = 0.014, n = 47;
Figure 2). For example, when soils had 30 Mg C ha
1,
the application of CA increased the SOC amount by 16%
(ln(r) = 0.15). However, on more fertile soils that had
60 Mg C ha
1, the SOC increase due to CA was only 9%
(ln(r) = 0.08) and no response can be expected on soils
reaching 100 Mg C ha
1.
To eliminate the effect of unequal controls, we subdi-
vided the database into two groups with a threshold value
of 40 Mg C ha
1. Compared to the controls, the weighted
summarized effect was a 20% (12%–28%, n = 22) of SOC
increase in the “ SOC ≤ 40 Mg ha
1” group, and only a 7%
(95% CI 3%–11%, n = 25) in the “SOC > 40 Mg ha
1” group
(Figure 1). However, the first group had a larger variability
of responses (τ = 0.14), compared to the latter (τ = 0.08).
Since within each group, no associations between
SOC stock and the response to CA were found (p = 0.93
and p = 0.70 for the “SOC ≤ 40 Mg ha
1” and “SOC
FIGURE 2 Weighted meta-
regression between SOC
changes due to CA expressed as
ln(r) and the SOC stock under
conventional agriculture
(control) for the entire database.
Point size represents study
weight in the analysis as
expressed in the Equation (5).
The dotted line represents the
control. For back-
transformation of ln(r), see
Equation 10
10 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
T
A
B
L
E
3
T
h
e
ef
fe
ct
of
pe
do
-c
lim
at
ic
an
d
ge
og
ra
ph
ic
al
fa
ct
or
s
an
d
m
an
ag
em
en
t
pr
ac
ti
ce
s
on
SO
C
ch
an
ge
s
du
e
to
co
n
se
rv
at
io
n
ag
ri
cu
lt
ur
e
in
so
ils
w
it
h
SO
C
≤
40
M
g
h
a

1
an
d
>
40
M
g
h
a

1
u
n
de
r
co
n
ve
n
ti
on
al
ag
ri
cu
lt
ur
e
V
ar
ia
bl
e
C
at
eg
or
y
E
xp
la
n
at
or
y
va
ri
ab
le
s
SO
C
≤
40
M
g
h
a

1
SO
C
>
40
M
g
h
a

1
n
R
2
(%
)
p
(Q
M
)
D
f
(Q
M
)
F
ig
u
re
O
u
tl
ie
rs
(s
tu
d
y
ID
)
n
R
2
(%
)
p
(Q
M
)
D
f
(Q
M
)
F
ig
u
re
O
u
tl
ie
rs
(s
tu
d
y
ID
)
Pe
do
-c
lim
at
ic
an
d
ge
og
ra
ph
ic
al
C
la
y
(%
)
21
0.
00
0.
98
0
1
3A
-
23
16
.3
5
0.
04
7
1
3A
14
Sa
n
d
(%
)
21
0.
00
0.
82
9
1
-
-
24
4.
86
0.
31
2
1
-
-
SO
C
/c
la
y
in
de
x
21
5.
76
0.
45
2
1
3B
-
23
29
.3
5
0.
00
6
1
3B
14
pH
(H
2O
)
22
1.
03
0.
77
6
1
3C
29
,4
2
22
37
.3
2
0.
02
1
1
3C
10
,4
6
R
ai
n
fa
ll
(m
m
ye
ar


1 )
21
43
.0
7
0.
00
7
1
4A
42
23
33
.2
8
0.
02
1
1
4A
10
,4
0
T
em
pe
ra
tu
re
22
6.
28
0.
09
2
1
-
-
25
3.
10
0.
43
1
1
-
-
L
at
it
ud
e
(d
eg
re
e
ab
so
lu
te
va
lu
e)
21
24
.9
1
0.
01
8
1
4B
29
23
39
.3
1
0.
00
7
1
4B
10
,4
0
L
on
gi
tu
de
(d
eg
re
e)
22
5.
32
0.
31
0
1
-
-
25
0.
01
0.
99
1
1
-
-
C
on
ti
n
en
t
22
3.
28
0.
83
9
3
5A
-
24
4.
88
0.
80
7
3
5B
-
K
öp
pe
n
cl
im
at
e
21
0.
00
0.
75
1
5C
-
25
0.
63
0.
79
3
1
5D
-
A
gr
on
om
ic
m
an
ag
em
en
t
E
xp
er
im
en
t
du
ra
ti
on
(y
ea
rs
)
22
17
.7
6
0.
02
0
1
6A
24
17
.4
6
0.
04
7
1
6A
47
Pr
op
or
ti
on
of
cr
op
s
w
it
h
h
ig
h
re
si
du
e
bi
om
as
s
(%
)
19
35
.0
6
0.
01
3
1
6B
16
,4
2
22
0.
00
0.
99
9
1
6B
-
N
u
m
be
r
of
cr
op
s
in
ro
ta
ti
on
(t
re
at
m
en
t)
22
8.
25
0.
17
5
1
5
E
-
25
0.
56
0.
74
6
1
5F
-
N
it
ro
ge
n
fe
rt
il
iz
at
io
n
le
ve
l
(k
g
N
ye
ar

1
)
12
4.
45
0.
47
6
1
-
-
15
4.
13
0.
41
1
1
-
-
L
eg
um
es
pr
es
en
ce
22
4.
61
0.
36
21
1
5G
-
25
1.
03
0.
76
46
1
5H
-
N
ot
e:
“n
”
is
th
e
n
um
be
r
of
st
ud
ie
s;
“R
2 ”
es
ti
m
at
es
th
e
am
ou
n
t
of
h
et
er
og
en
ei
ty
ac
co
un
te
d
fo
r
by
th
e
m
od
er
at
or
s
in
cl
ud
ed
in
th
e
m
od
el
;“
p(
Q
M
)”
is
th
e
p-
va
lu
e
of
th
e
h
et
er
og
en
ei
ty
te
st
;a
n
d
“d
f(
Q
M
)”
is
th
e
de
gr
ee
s
of
fr
ee
do
m
of
th
e
re
si
du
al
h
et
er
og
en
ei
ty
te
st
.p
-v
al
ue
s
h
ig
h
lig
h
te
d
in
bo
ld
ar
e
si
gn
if
ic
an
t
(p
<
0.
05
).
O
ut
li
er
s
w
er
e
id
en
ti
fi
ed
by
ba
ck
w
ar
d
se
ar
ch
al
go
ri
th
m
(M
av
ri
di
s
et
al
.,
20
17
).
TADIELLO ET AL. 11 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
> 40 Mg ha
1” groups, respectively), this allowed us to
study the effects of pedo-climatic factors and manage-
ment practices within each group.
3.2 | The source of variation across
studies
The effects of five different moderators were significant
in at least one of the two SOC stock groups (Table 3).
The R2 coefficient varied significantly from one moder-
ator to another, indicating that the SOC response vari-
ability is strictly dependent on specific moderators. For
instance, within the continuous moderator group, the
largest R2 was recorded for rainfall (43%) and latitude
(39%), while the smallest was the pH (1%).
3.2.1 | Pedo-climatic and geographical
factors
In the “SOC ≤ 40 Mg ha
1” group, a change in clay or
sand percentage (within the range of this study) did not
lead to a different response to the CA adoption, while in
the “SOC > 40 Mg ha
1” group, an increased clay con-
tent was positively correlated with the magnitude of the
response (Table 3 and Figure 3a). For example, at 60% of
clay, the SOC increase due to CA adoption was 12% (ln
(r) = 0.10) compared to conventional agriculture, while
no response was found (ln(r) = 0.0) at 8% of clay. The
results obtained with the clay moderator reflected what
was found considering the SOC/clay index (Table 3 and
Figure 3b). In fact, if the “SOC ≤ 40 Mg ha
1” group did
not report a significant regression, while in the
“SOC > 40 Mg ha
1” group, the result was significant.
The lower the SOC/clay index, the higher the CA impact
on SOC compared to conventional agriculture. The pH
was also considered in the analysis even though just the
“SOC > 40 Mg ha
1” group was significant (Figure 3c).
The greatest effect size was found with the lowest pH
value (around 5), while the impact of CA becomes null
at pH equal to 7.3.
The meta-regressions indicated that rainfall was an
important factor governing the response to CA in both
groups (R2 = 33% and 43%, Table 3). The relationship
was negative in the “SOC ≤ 40 Mg ha
1” group, and posi-
tive in the “SOC > 40 Mg ha
1” group (Figure 4a). In
both groups, an increase of 100 mm of rainfall was asso-
ciated with a 1% change in the SOC response to CA.
The temperature cannot explain the groups variability
(p > 0.05, Table 3), with a R2 always lower than 10%.
Table 3 shows that the effect of CA strongly depended
on geographical locations as indicated by the large pro-
portion of heterogeneity (25%–39%) accounted for by lati-
tude (degrees absolute value). In “SOC ≤ 40 Mg ha
1”
group, with increasing latitude, the impact of CA sharply
increased, whereas, in “SOC > 40 Mg ha
1” group, CA
had an opposite trend (Figure 4b).
The SOC responses to CA did not reveal any differ-
ences across continents (Figure 5a,b). However, the num-
ber of studies in South America and in Asia for the “SOC
≤ 40 Mg ha
1” group and in Europe for the “SOC
> 40 Mg ha
1” group was limited, and thus, the 95% CIs
were large and overlapping.
No differences in the effect of CA adoption were
found between Csa and Cfa Köppen climate categories,
since the 95% CI were large and overlapping (Table 3;
FIGURE 3 Weighted meta-regressions between SOC changes
due to CA expressed as ln(r) and clay (a), SOC/clay index (b) and
pH (C) in soils with SOC ≤ 40 Mg ha
1 and SOC > 40 Mg ha
1
under conventional agriculture. The point size represents the study
weight in the analysis as expressed in Equation (5). Statistics for
meta-regressions and ID of studies identified as outliers (crosses)
appear in Table 3. For back-transformation of ln(r), see
Equation (10)
12 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FIGURE 4 Weighted meta-regressions between SOC changes due to CA expressed as ln(r) and rainfall (a) and latitude (b) in soils with
SOC ≤ 40 Mg ha
1 and SOC > 40 Mg ha
1 under conventional agriculture. The point size represents the study weight in the analysis as
expressed in Equation (5). Statistics for meta-regressions and ID of studies identified as outliers (crosses) appear in Table 3. For back-
transformation of ln(r), see Equation (10)
FIGURE 5 SOC percentage changes due to CA as effected by continents (a,b), Köppen climate groups (c,d), number of crops in treatment
rotation (e,f), and the legume presence in the treatment rotation (g,h) in soils with SOC ≤ 40 Mg ha
1 and SOC > 40 Mg ha
1 under conventional
agriculture. The symbols indicate weighted average with 95% confidence intervals (CIs) and “n” represents the number of studies. Numbers in the
right columns are summary effect estimates [lower 95% CI, upper 95% CI]. The dashed line indicates conventional agriculture (control). The effect
of CA on SOC was considered significantly different from the control if the 95% CIs do not overlap with zero, and significantly different between
the groups of explanatory variables if their 95% CIs do not overlap. Groups with only one study were excluded from the analysis
TADIELLO ET AL. 13 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Figure 5c,d). The Csb climate was excluded from the
analysis since only one study was found.
3.2.2 | Agronomic management practices
In most of the studies, the duration of experiments ran-
ged from 2 to 30 years. Figure 6a displays the significant
positive relationship between experiment duration and
SOC response due to CA adoption in both SOC stock
groups (Table 3). The regression slope was greater in the
“SOC ≤ 40 Mg ha
1” group; in particular, after ten years
of CA implementation, the percentage change from con-
ventional agriculture was 20% in the “SOC
≤ 40 Mg ha
1” group (ln(r) = 0.17), and 5% in the “SOC
> 40 Mg ha
1” group (ln(r) = 0.04).
The proportion of crops with high residue biomass in
a rotation significantly affected C sequestration under CA
only in the “SOC ≤ 40 Mg ha
1” group (Table 3 and
Figure 6b). If all the crops in a rotation produced a large
amount of residue, the SOC increase was about 29% (ln
(r) = 0.25). In a rotation in which only half of the crops
left a high amount of residue on the soil, the SOC
increase due to CA adoption was about 20% (ln(r) = 0.18)
compared to conventional agriculture. In the “SOC
> 40 Mg ha
1” group, the amount of residues did not
modify the SOC response to CA (Table 3, Figure 6b).
In both groups, the introduction of three or more crops
in a treatment rotation (Figure 5e,f), the presence of
legumes in the rotation (Figure 5g,h), and the nitrogen fer-
tilization level did not have a significant impact (Table 3).
3.3 | Publication bias
Although the funnel plots indicated some asymmetry,
the Egger's regression test did not indicate a significant
asymmetry (p = 0.64) when the sample size is present on
the y-axis (Figure S3A). When the SE appeared in the
y-axis (Figure S3B), some points fell outside the 95% CI,
but the regression test still confirmed a non-significant
asymmetry (p = 0.051). We concluded that our research
does not suffer from publication bias.
In addition, the fail-safe number indicated that the
results are robust for both SOC groups. In fact, the fail-
safe number is 280 and 62 for the “SOC ≤ 40 Mg ha
1”
and “SOC > 40 Mg ha
1” groups, respectively, suggesting
that there would need to be a consistent number of stud-
ies for each group before the cumulative effect would
become statistically non-significant (Supplementary
file 1).
4 | DISCUSSION
This meta-analysis summarizes the results of 47 studies
published over a period of 20 years on the effects of CA
practices on SOC sequestration in the plough layer under
the Mediterranean and humid subtropical climates in
five continents. Since our database included numerous
different crops (n = 23) and rotations (n = 31), this
allowed us to explore different agronomic conditions
(Database 1, https://doi.org/10.5281/zenodo.7404592).
Previousmeta-analysis on similar topic and climatic con-
ditions summarized 33 studies on herbaceous crops
(Aguilera et al., 2013). Other meta-analyses reached
greater number of studies including deeper soil layers
(Luo et al., 2010; Mondal et al., 2020) or they were con-
ducted on a global scale (e.g. Li et al., 2020; Sun
et al., 2020;West & Post, 2002). Although our initial data-
base was larger, many studies were not included to the
database due to (1) the variability of management prac-
tices (e.g. minimum tillage in place of no-till, agronomic
management change during the experiment's duration),
FIGURE 6 Weighted meta-regressions between SOC changes due to CA expressed as ln(r) and experiment duration (a) and proportion
of crop with high residue biomass (b) in soils with SOC ≤ 40 Mg ha
1 and SOC > 40 Mg ha
1 under conventional agriculture. The point size
represents the study weight in the analysis as expressed in Equation (5). Statistics for meta-regressions and ID of studies identified as outliers
(crosses) appear in Table 3. For back-transformation of ln(r), see Equation (10)
14 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
or cover crop inclusion in control treatments; (2) incom-
plete and poor reporting of the results, such as missing
means, SDs or sample sizes for controls and treatments;
(3) utilization of pedotransfer functions or lack of
reported bulk density. Despite such constrains, the num-
ber of studies included in the final database is compara-
ble with other similar meta-analyses (Table 1). This
number was sufficient to perform a robust, weighted
meta-analysis to calculate a summarized effect size
across the studies, as well as the means for the different
categories of explanatory variables, and to determine the
CIs around themeans.
4.1 | Overall effect and the role of SOC
stock
This meta-analysis clearly indicates that under Mediter-
ranean and humid subtropical climates, CA adoption had
an overall positive effect on SOC sequestration amount-
ing to an overall mean increase of 12% in the plough
layer compared to conventional agriculture. This result
can also be translated as an annual C increase of about
0.48 Mg C ha
1 (Table S2). This annual C increase has
been scaled across the 12 countries included in the pre-
sent study, considering the agricultural area (under Cfa,
Csa, and Csb) where CA can be potentially adopted. The
result indicated that, in this area, 0.15 Pg C y
1 can be
stored in the first 30 cm layer due to the CA adoption.
This result shows how the recent literature, combined
with the summarized result of the meta-analysis, can
quantify the real impact of the CA adoption.
This meta-analysis also gives the first evidence that
the magnitude of SOC gain due to CA adoption was
strongly influenced by the SOC stock under conventional
agriculture. In soils with SOC ≤ 40 Mg C ha
1, the
impact of CA adoption was three times larger (20%, 95%
CI 12%–28%, n = 22) than in soils which stored the
higher SOC stock (7%, 95% CI 3%–11%, n = 25). In our
database, the SOC stock under conventional agriculture
varied from 18 to 102 Mg C ha
1 in the plough layer,
which includes the average value of 63.5 Mg ha
1
reported by the FAO (2020) for the warm temperate cli-
mate. The evidence of distinct responses between the two
SOC stock groups (Figure 2) became important to better
understand the variability of CA impacts across Mediter-
ranean and humid subtropical climates. As a rule, the
carbon content achieved under conventional agriculture
must be considered to estimate whether SOC sequestra-
tion can be increased by CA adoption. This can be
explained by the fact that SOC sequestration rates have
been found to be greater in soils that are far from their
potential steady state (Tiefenbacher et al., 2021). On the
contrary, when SOC is already close reaching to a steady
state, the SOC gains are lower (Corsi et al., 2012; Kämpf
et al., 2016).
Previous meta-analyses on this topic which summa-
rized studies in Mediterranean (Aguilera et al., 2013) and
temperate climates (Ogle et al., 2005) did not explicitly
consider the SOC stock under conventional tillage as a
moderator: however, they demonstrated a 17%–18% of
SOC increase as an overall effect, which is somewhat
larger than we found in this meta-analysis (12%) for the
entire database. These larger responses very likely stem
from the use of weighting by sample size or no weighting.
In this meta-analysis, however, the weighting by the
inverse of the variance was used, which usually gives
smaller effect size estimates (Hungate et al., 2009). Our
results of the overall effects agree with the global meta-
analysis by Li et al. (2020), who also used same metrics of
effect size and weighting function (Table 1).
During the last two decades, no-till management has
also been recommended as a practice to mitigate green-
house gas emissions through soil C sequestration (Ogle
et al., 2012) and lower fuel consumption (Aguilera
et al., 2013). The present meta-analysis defines that,
under Mediterranean and humid subtropical climates,
CA produce positive carbon sequestration regardless of
the initial carbon content (i.e. positive effect in both SOC
groups). This finding confirms that CA must be consid-
ered by the lawmakers to mitigate greenhouse gas emis-
sions. In addition, Powlson et al. (2016) suggested to
evaluate the SOC increase considering the management
practices involved (e.g. crop rotation, residue manage-
ment, and soil characteristic).
4.2 | Source of variation across studies
The final distribution of the effects size (Figure 1)
describes a wide range of SOC change in response to CA
(from 
9% to 99% change from the control). The hetero-
geneity of the effect size, as quantified by the high I2
value (94.6%, 95% CI 92.4%–97.2%, n = 47) justifies the
random model utilization (since the different studies do
not share a common effect size). Similar I2 value was
reported by other agronomic (Kim et al., 2020; Tremblay
et al., 2012; Zhao et al., 2020) or ecological meta-analyses
(Senior et al., 2016). The large I2 allows us to investigate
reasons for the variability, applying subgroup analysis or
meta-regression (Nakagawa et al., 2017).
Our results proved that splitting the database up into
two groups based on the amount of SOC under conven-
tional agriculture (i.e. the “SOC ≤ 40 Mg ha
1” and
“SOC > 40 Mg ha
1”) allowed us to detect contrasting
effects of the pedoclimatic and management factors on
TADIELLO ET AL. 15 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SOC sequestration. In fact, the analysis of the whole data-
base would have masked the moderators' impacts on
SOC sequestration, since the effect of the two groups
would have been averaged. Conversely, with this
approach, we can give agronomic explanations of the
moderator impact, separately for the soils with different
SOC stocks under conventional agriculture.
4.2.1 | Pedo-climatic and geographical
factors
Our results suggest a significant positive effect of CA due
to the clay percentage only in soils with SOC
>40 Mg ha
1 (Figure 3a). Several authors acknowledge
the positive effect of the clay percentage on the SOC
adsorption by the mineral fraction and the resulting SOC
accumulation (Du et al., 2017; Haddaway et al., 2017; Xu
et al., 2016). This is related to the fact that clay soils
exhibit strong aggregate formation and stability that pre-
vent SOM decomposition (Lorenz & Lal, 2018). Since C
sequestration is constrained mainly by the availability of
reactive surfaces (Churchman et al., 2020), high C
amount (i.e. “SOC > 40 Mg ha
1” group) still leads to
increase in SOC response, if supported by greater clay
content in the soil (Figure 3a). In contrast, in the soils
with SOC ≤ 40 Mg ha
1, all the C available is likely to be
adsorbed by the clay minerals, making this factor irrele-
vant for further SOC increase. It is interesting to note that
this finding is confirmed by the SOC/clay index: in soil
with already high SOC availability, the CA implementa-
tion does not automatically lead to a greater SOC seques-
tration. An optimal combination of SOC and clay has to
be matched in order to get a positive effect on SOC
sequestration (Figure 3b). In our study, we identify a
maximum SOC/clay index threshold of 3.2: once this
value is overcome the CA adoption became useless to
increase SOC sequestration (i.e. the effect size became
zero). This finding indirectly confirms that soils with
poor clay content are likely to be closer to the soil satura-
tion limit, while soil with greater clay content are more
likely to have a higher SOC sequestration potential.
Worthy of mention is the role of pH. In the “SOC
≤ 40 Mg ha
1” group, the differences in pH did not lead
to a greater impact on carbon sequestration due to CA,
probably because oscillations in the pH values do not
trigger a greater mineralization when the organic matter
is too low and barely available to microbes. On the con-
trary, when the SOC stock is greater (“SOC > 40 Mg
ha
1” group), unfavourable pH values for the mineraliza-
tion process (pH < 5, Aciego Pietri & Brookes, 2008) lead
to lower OM decomposition (Bot & Benites, 2005) thus a
higher SOC sequestration (i.e. greater effect size,
Figure 3c).
In our study, the amount of rainfall showed a signifi-
cant effect on SOC sequestration due to CA in both
groups (Table 3, Figure 4a), while some previous meta-
analyses failed to detect the rainfall effect (Du
et al., 2017; Luo et al., 2010). Sun et al. (2020) indicated
that higher SOC gain with the CA adoption is expected
with a decrement of the humidity index (ratio of annual
mean precipitation to mean temperature). It is likely
that, in our meta-analysis, the enhanced soil water
retention due to CA practices (Lal, 2020) occurred in
soils with SOC ≤ 40 Mg C ha
1, which gave a visible
advantage only in dry conditions (the left side of the
regression shown in Figure 4a), while in geographical
areas where water is not limited, C sequestration
improvement had only a small increment. Another pos-
sible explanation for the good CA performance in C
depleted areas is associated with the irrigation tech-
nique: in some agricultural regions, the irrigation coun-
terbalances the negative effect of scant precipitation on
carbon sequestration (Lorenz & Lal, 2018). In contrast,
we found a positive trend in the soil with SOC
> 40 Mg ha
1, although the low slope indicated a weak
impact of rainfall on CA effect. This finding agrees with
the result by Post et al. (1982), who linked a high rain-
fall regime with SOC accumulation in soils.
Our meta-regressions clearly indicate that the geo-
graphical location of an experiment determined to what
extent CA influenced SOC sequestration. Moving from
the lower latitudes towards the middle latitudes suggests
an increasing advantage of CA in soils with initial SOC
lower than 40 Mg ha
1, as indicated by the positive
meta-regression in the “SOC ≤ 40 Mg ha
1” group
(Figure 4b). In this group, CA showed a small positive
contribution at low latitude where the high temperature
hastens the mineralization of the limited SOC stock.
Conversely, moving to middle latitudes with lower tem-
peratures, the CA effect increased, probably, due to
slower mineralization. From the agronomic point of
view, CA practices are not enough to increase C seques-
tration in conditions with low carbon content and a
warm climate (i.e. low latitude absolute value). In soils
with SOC > 40 Mg ha
1, we found an opposite trend
(Figure 4b). This suggests that the higher effect size is
found at low latitudes which are characterized by high
temperature, where SOC stock is not a limiting driver,
and it can be mineralized without decreasing the SOC
stock accumulation in soil. Conversely, with higher lati-
tudes, probably the introduction of CA practices in soil
with an already high SOC stock is not enough to lead to
an increment in SOC stock accumulation. However, no
significant effect of temperature was found for both
16 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SOC groups (Table 3), indicating that even if tempera-
ture is certainly a driver of the SOC mineralization,
other factors can influence the SOC accumulation.
Therefore, these findings highlight the fact that CA is
not a “standardized solution” to explain the carbon
accumulation problems in agricultural soils, and the
benefits of its application should be evaluated by consid-
ering other agronomic and climatic variables.
The peculiar behaviours of the “SOC ≤ 40 Mg ha
1”
and “SOC > 40 Mg ha
1” groups provide a novel contribu-
tion about the CA effect on SOC sequestration at different
latitudes, due to the lack of previous findings related to
this topic or for results limited to specific soil layers. For
example, Haddaway et al. (2017), who studied SOC
response regardless of the C stock in conventional tillage
(control), found that latitude was positively correlated to C
stocks' mean differences in full profile C stocks.
Other geographical moderators were not useful to
explain heterogeneity of the effect sizes across the stud-
ies. In fact, the continent moderator had no impact to
explain the heterogeneity, probably, because the studies
from different continents had the same climate condi-
tions (i.e. Cfa, Csa). Therefore, in the present meta-analy-
sis, the evidence that Cfa and Csa subgroups of the
Köppen climate classification did not show significant
differences could be partially explained by the similar
average temperature and rainfall during the year. Ogle
et al. (2005) confirmed that large differences occur with
contrasting climates, finding that no-till implementation
led to the largest increases in SOC storage under tropical
moist conditions and the smallest under temperate dry
conditions.
4.2.2 | Agronomic management
Our results support the results of many other meta-
analyses on the same topic (Angers & Eriksen-
Hamel, 2008; Haddaway et al., 2017), which reported that
experiment duration positively influences SOC accumula-
tion in soil (Table 1). Moreover, our results indicate that
in soils with SOC content ≤ 40 Mg ha
1 there was a
quick temporal response to CA (i.e. a greater response
starting from the beginning of the CA implementation).
In addition, the SOC stock accumulation in soils with
low SOC content throughout the years was faster, as con-
firmed by the higher regression slope compared to that of
the “SOC > 40 Mg ha
1” group.
Another critical aspect related to SOC sequestration
is the crop residues management. Differences in the
amounts of yields and, thus, crop residues, directly
influence the amount of C inputs to the soil (Meurer
et al., 2018; Poeplau & Don, 2015). On the other hand,
when with ploughed soil, the residues are expected to
be in contact with deeper soil layers, CA management
leaves them on the soil surface. In the latter case, CA
leads to positive effects, such as soil temperature con-
trol, the limitation of soil erosion, and the reduction of
soil water evaporation, which are all associated with
the reduction of SOC decomposition in soil (Duiker &
Lal, 2000; Luo et al., 2010). In literature, the SOC stock
in CA is known to positively respond to crop residues
retention, as supported by the meta-analysis by Virto
et al. (2012), who found a significant (p = 0.001,
n = 35) relation between SOC accumulation in 0–0.3 m
soil depth and organic input (i.e. crop residue), consid-
ering NT (with residues) as a treatment and inversion
tillage as a control. Our result supported this previous
finding, but with a positive relationship limited to soils
with scarce SOC stock. In fact, we found out that
increasing the proportion of crops with high residue
production in the rotation results in a SOC increase
only in soils with SOC ≤ 40 Mg ha
1, reaching 29%
change from the control when all the crops in the rota-
tion produce high residue amounts.
Even if the positive relationship between SOC
sequestration and the amount of crop residues retained
on the soil has been highlighted in different studies,
the different quality of the crop residue also plays a
role in the C stock accumulation. For example, in the
meta-analysis by Sun et al. (2020), they found that in
most Mediterranean and temperate climates, SOC
sequestration increased when crop residue retention
and crop rotation are applied together. The number of
crops in the rotation indeed plays a role in SOC accu-
mulation, since monoculture produces the worst qual-
ity and quantity of dry matter (Copeland &
Crookston, 1992). The study by Gonzalez-Sanchez et al.
(2012) demonstrated that, in general, the higher C soil
fixation values were found in soils in which crops were
rotated, with on average C sequestration rate 19%
higher in the case of crop rotation and NT rather than
monoculture. In the present meta-analysis, however,
the number of crops in the CA treatment rotation did
not significantly impact the SOC accumulation in the
plough layer (Table 3). This result is likely due to the
unbalanced number of studies between the two levels
considered (i.e. rotations with three or more crops or
with less than three, Figure 5e,f).
4.3 | Perspectives
The current carbon stock data availability under our
studies selection criteria did not allow us to obtain
enough studies to consider the deeper layers (> 0.3 m
TADIELLO ET AL. 17 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
depth). However, other authors highlighted the impor-
tance of also engaging the deeper layers for a complete
evaluation of the soil carbon storage (Kopittke
et al., 2017; Meurer et al., 2018; Piccoli et al., 2016).
Sun et al. (2020) noted that, for the cases when SOC
under no-till relative to conventional tillage increased
in the top 0.3 m, the 0.3–0.6 m layer was also likely to
increase its SOC. However, Du et al. (2017) reported
that no-till management showed slightly lower SOC
storage rates against conventional till in the subsoil
layers (> 0.4 m).
Further research synthesis should address the SOC
response to CA in the deeper soil layers, focus on specific
climatic zones, and different management practices. In
this sense, the present work is useful since a reliable and
replicable procedure is clearly presented.
Lastly, we should note that within this topic, the
information regarding irrigation was rarely reported. The
final low number of studies handling irrigation did not
allow us to include this moderator in the meta-analysis.
5 | CONCLUSIONS
The present meta-analysis evaluated the SOC response
to CA practices in Mediterranean and humid subtropi-
cal climates. Limiting the analysis to specific climates
offered the possibility to detect more precisely the
effects of pedo-climatic and management practices,
which otherwise would have been masked. Therefore,
we have provided a novel contribution to understand-
ing the actual impact of CA in the SOC stock
accumulation.
The meta-analysis showed an overall positive effect
of CA on SOC sequestration (12%). Scaling this result
across the countries and the specific climates consid-
ered in the present study, 0.15 Pg C y
1 can be stored
in the first 30 cm layer due to the CA adoption. By
dividing the whole database into two separate groups
based on the SOC stock (with 40 Mg C ha
1 as the
threshold) under conventional agriculture allowed us
to better explain the variability of SOC responses to CA
management. This meta-analysis highlighted that,
under the climates considered, the effect of CA adop-
tion on SOC accumulation in the plough layer reached
20% in soils with SOC ≤ 40 Mg ha
1, while it only aver-
aged 7% in soils with the SOC > 40 Mg ha
1.
The effect of CA on SOC accumulation depended on
clay content solely in soils with more than 40 Mg C ha
1
under conventional agriculture, while it was not relevant in
soils with less than 40 Mg C ha
1. This result was con-
firmed by the SOC/clay index analysis that revealed that to
get a positive impact of CA on SOC sequestration a specific
range of the SOC/clay index is required. In both soil groups,
experiment duration positively impacted SOC sequestration,
with a greater effect found in the soils with SOC ≤ 40 Mg C
ha
1. In addition, in these soils, the retention of crop resi-
dues enhanced the CA positive contribution.
We conclude that in Mediterranean and humid sub-
tropical climates, the most benefits from CA application in
terms of SOC increase apply to agricultural soils with SOC
content ≤ 40 C Mg ha
1 and located in the middle lati-
tudes and/or in dry areas. With a base annual increment
of 0.48 Mg C ha
1 y
1, we support the idea that a reason-
able carbon gain can be enhanced with a long CA applica-
tion. For instance, to get a reasonable 20% more carbon
stock, it is required ten years of CA application with an
initial carbon stock ≤ 40 Mg ha
1, while more than
30 years are required if the soil already has more than
40 C Mg ha
1. During this period, it is recommended at
least to apply to continue NT management, retain residues
on the top of the soil (chopped or not), and include as
many crops as possible in the rotation.
AUTHOR CONTRIBUTIONS
Tommaso Tadiello: Conceptualization (equal); data
curation (equal); formal analysis (equal); writing – origi-
nal draft (lead). Marco Acutis: Conceptualization (lead);
data curation (equal); formal analysis (equal); supervision
(lead); writing – review and editing (equal). Alessia Per-
ego: Conceptualization (supporting); methodology
(equal); writing – review and editing (equal). Calogero
Schillaci: Conceptualization (equal); data curation
(equal); writing – review and editing (equal). Elena
Valkama: Conceptualization (lead); data curation (lead);
formal analysis (equal); methodology (lead); supervision
(lead); writing – original draft (lead).
ACKNOWLEDGEMENTS
This work is funded by the Agriculture, Environment,
and Bioenergy PhD school of the University of Milan.
This study is also funded by the European Union's
Horizon 2020 Framework Programme for Research
and Innovation (H2020-RUR-2017-2) as part of the
LANDSUPPORT project (grant agreement
No. 774234), which aims at developing a decision sup-
port system for optimising soil management in Europe.
This study is a part of the project “Carbon Market -
Innovative cropping systems for carbon market”
funded by Natural Resources Institute Finland (Luke).
This study is also a part of the project
P
ommit: Sus-
tainable management of soil organic matter to mitigate
trade-offs between C sequestration and nitrous oxide,
methane, and nitrate losses that received funding from
the European Joint Programme SOIL (grant agreement
ID: 862695).
18 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
CONFLICT OF INTEREST
The authors declare that they have no known competing
financial interests or personal relationships that could
have appeared to influence the work reported in this
paper.
DATA AVAILABILITY STATEMENT
Databases are free to download in Zenodo (https://doi.
org/10.5281/zenodo.7404592): Database 1. Meta-analysis
database; Database 2. Original raw data (SOC content
and/or stock, bulk density) for each soil layer as reported
in the articles, and SOC stock computation for a single
soil layer (0–0.3 m).
ORCID
Tommaso Tadiello https://orcid.org/0000-0001-9919-
317X
Marco Acutis https://orcid.org/0000-0002-1576-8261
Alessia Perego https://orcid.org/0000-0002-0601-9699
Calogero Schillaci https://orcid.org/0000-0001-7689-
5697
Elena Valkama https://orcid.org/0000-0002-8337-8070
REFERENCES
Aciego Pietri, J. C., & Brookes, P. C. (2008). Relationships between
soil pH and microbial properties in a UK arable soil. Soil Biol-
ogy and Biochemistry, 40, 1856–1861. https://doi.org/10.1016/j.
soilbio.2008.03.020
Acutis, M., Tadiello, T., Perego, A., Di Guardo, A., Schillaci, C., &
Valkama, E. (2021). Ex-Tract tool_1.2.xlsm. figshare. Software.
https://doi.org/10.6084/m9.figshare.14987130.v9
Acutis, M., Tadiello, T., Perego, A., Di Guardo, A., Schillaci, C., &
Valkama, E. (2022). EX-TRACT: An excel tool for the estima-
tion of standard deviations from published articles. Environ-
mental Modelling and Software, 147, 105236. https://doi.org/10.
1016/j.envsoft.2021.105236
Aguilera, E., Lassaletta, L., Gattinger, A., & Gimeno, B. S. (2013).
Managing soil carbon for climate change mitigation and adap-
tation in Mediterranean cropping systems: A meta-analysis.
Agriculture, Ecosystems & Environment, 168, 25–36. https://doi.
org/10.1016/j.agee.2013.02.003
A´lvaro-Fuentes, J., & Paustian, K. (2011). Potential soil carbon
sequestration in a semiarid Mediterranean agroecosystem
under climate change: Quantifying management and climate
effects. Plant Soil, 338, 261–272. https://doi.org/10.1007/s11104-
010-0304-7
Angers, D. A., & Eriksen-Hamel, N. S. (2008). Full-inversion tillage
and organic carbon distribution in soil profiles: A meta-analy-
sis. Soil Science Society of America Journal, 72, 1370–1374.
https://doi.org/10.2136/sssaj2007.0342
Aria, M., & Cuccurullo, C. (2017). Bibliometrix: An R-tool for com-
prehensive science mapping analysis. Journal of informetrics,
11(4), 959–975, Elsevier. https://doi.org/10.1016/j.joi.2017.
08.007
Borenstein, M., Hedges, L., Higgins, J., & Rothstein, H. (2009).
Introduction to meta-analysis. John Wiley & Sons Ltd.
Bot, A., & Benites, J. (2005). The importance of soil organic matter:
key to drought-resistant soil and sustained food production.
FAO SOILS BULLETIN 80, Food and agriculture organization
of the united nations, Rome.
Bouma, E. (2005). Development of comparable agro-climatic zones
for the international exchange of data on the efficacy and crop
safety of plant protection products. EPPO Bulletin, 35, 233–238.
https://doi.org/10.1111/j.1365-2338.2005.00830.x
Chen, D., & Chen, H. (2013). Using the Köppen classification to
quantify climate variation and change: An example for 1901–
2010. Environmental Development, 6, 69–79. https://doi.org/10.
1016/j.envdev.2013.03.007
Churchman, G. J., Singh, M., Schapel, A., Sarkar, B., & Bolan, N.
(2020). Clay minerals as the key to the sequestration of carbon
in soils. Clays and Clay Minerals, 68, 135–143. https://doi.org/
10.1007/s42860-020-00071-z
Climate Change Knowledge Portal, for Development Practitioners
and Policy makers. (2021). https://climateknowledgeportal.
worldbank.org/download-data
Copeland, P. J., & Crookston, R. K. (1992). Crop sequence affects
nutrient composition of corn and soybean grow under high fer-
tility. Agronomy Journal, 84, 503–509.
Corsi, S., Friedrich, T., Kassam, A., Pisante, M., & Moraes, S. J.
(2012). Soil organic carbon accumulation and greenhouse
gas emisson reductions from conservation agriculture: A lit-
erature review, integrated crop management. FAO (Eds.),
food and agriculture Organization of the United Nations,
Rome.
Dexter, A. R., Richard, G., Arrouays, D., Czy, E. A., Jolivet, C., &
Duval, O. (2008). Complexed organic matter controls soil physi-
cal properties. Geoderma, 144, 620–627. https://doi.org/10.
1016/j.geoderma.2008.01.022
Du, Z., Angers, D. A., Ren, T., Zhang, Q., & Li, G. (2017). The effect of
no-till on organic C storage in Chinese soils should not be over-
emphasized: A meta-analysis. Agriculture, Ecosystems & Environ-
ment, 236, 1–11. https://doi.org/10.1016/j.agee.2016.11.007
Duiker, S. W., & Lal, R. (2000). Carbon budget study using CO2 flux
measurements from a no till system in Central Ohio. Soil and
Tillage Research, 54, 21–30. https://doi.org/10.1016/S0167-1987
(99)00101-4
European Commission. (2021). The new common agricultural
policy: 2023–2027. https://agriculture.ec.europa.eu/common-
agricultural-policy/cap-overview/new-cap-2023-27_en#docu
ments
FAO. (2016). Save and Grow in practice: Maize, rice and wheat.
Rome.
FAO. (2017a). Global Soil Organic Carbon Map—Leaflet. Global Soil
Partnership, Rome, Italy. 5 pp.
FAO. (2017b). Conservation agriculture, Rome, Italy, 2pp.
FAO. (2020). Global Soil Organic Carbon Map (GSOCmap) Version
1.5. https://doi.org/10.4060/ca7597en
Gonzalez-Sanchez, E. J., Ordoñez-Fernandez, R., Carbonell-
Bojollo, R., Veroz-Gonzalez, O., & Gil-Ribes, J. A. (2012). Meta-
analysis on atmospheric carbon capture in Spain through the
use of conservation agriculture. Soil and Tillage Research, 122,
52–60. https://doi.org/10.1016/j.still.2012.03.001
TADIELLO ET AL. 19 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Gurevitch, J., & Hedges, L. V. (1999). Statistical issues in ecological
meta-analyses. Ecology, 80, 8.
Gurevitch, J., Koricheva, J., Nakagawa, S., & Gavin, S. (2018).
Meta-analysis and the science of research synthesis. Nature,
555, 175–182. https://doi.org/10.1038/nature25753
Haddaway, N. R., Hedlund, K., Jackson, L. E., Kätterer, T.,
Lugato, E., Thomsen, I. K., Jørgensen, H. B., & Isberg, P. E.
(2017). How does tillage intensity affect soil organic carbon? A
systematic review. Environmental Evidence, 6, 30. https://doi.
org/10.1186/s13750-017-0108-9
Hedges, L., & Olkin, I. (1985). Statistical methods for meta-analysis.
Academic Press.
Hedges, L. V., Gurevitch, J., & Curtis, P. S. (1999). The meta-
analysis of response ratios in experimental ecology. Ecology,
80(4), 1150–1156. https://doi.org/10.2307/177062
Hernandez-Ochoa, I., & Asseng, S. (2018). Cropping systems and
climate change in humid subtropical environments. Agronomy,
8, 19. https://doi.org/10.3390/agronomy8020019
Hungate, B. A., van Groenigen, K. J., Six, J., Jastrow, J. D.,
Luo, Y., de Graaff, M. A., van Kessel, C., & Osenberg, C. W.
(2009). Assessing the effect of elevated carbon dioxide on soil
carbon: A comparison of four meta-analyses. Global Change
Biology, 15, 2020–2034. https://doi.org/10.1111/j.1365-2486.
2009.01866.x
Jones, R. J. A., Hiederer, R., Rusco, E., & Montanarella, L. (2005).
Estimating organic carbon in the soils of Europe for policy sup-
port. The European Journal of Soil Science, 56, 655–671. https://
doi.org/10.1111/j.1365-2389.2005.00728.x
Kämpf, I., Hölzel, N., Störrle, M., Broll, G., & Kiehl, K. (2016).
Potential of temperate agricultural soils for carbon sequestra-
tion: A meta-analysis of land-use effects. Science of the Total
Environment, 566–567, 428–435. https://doi.org/10.1016/j.
scitotenv.2016.05.067
Kim, N., Zabaloy, M. C., Guan, K., & Villamil, M. B. (2020). Do
cover crops benefit soil microbiome? A meta-analysis of current
research. Soil Biology and Biochemistry, 142, 107701. https://
doi.org/10.1016/j.soilbio.2019.107701
Kopittke, P. M., Dalal, R. C., Finn, D., & Menzies, N. W. (2017).
Global changes in soil stocks of carbon, nitrogen, phosphorus,
and Sulphur as influenced by long-term agricultural produc-
tion. Global Change Biology, 23, 2509–2519. https://doi.org/10.
1111/gcb.13513
Koricheva, J., Gurevitch, J., & Mengersen, K. (Eds.). (2013). Hand-
book of meta-analysis in ecology and evolution. Princeton Uni-
versity Press.
Kottek, M., Grieser, J., Beck, C., Rudolf, B., & Rubel, F. (2006).
World map of the Köppen-Geiger climate classification
updated. Metz, 15, 259–263. https://doi.org/10.1127/0941-2948/
2006/0130
Lal, R. (2004). Soil carbon sequestration impacts on global climate
change and food security. Science, 304, 1623–1627. https://doi.
org/10.1126/science.1097396
Lal, R. (2020). Soil organic matter and water retention. Agriculture
Journal, 112(5), 3265–3277. https://doi.org/10.1002/agj2.20282
Li, Y., Li, Z., Chang, S. X., Cui, S., Jagadamma, S., Zhang, Q., &
Cai, Y. (2020). Residue retention promotes soil carbon accumu-
lation in minimum tillage systems: Implications for conserva-
tion agriculture. Science of the Total Environment, 740, 140147.
https://doi.org/10.1016/j.scitotenv.2020.140147
Lorenz, K., & Lal, R. (2018). Carbon sequestration in agricultural
ecosystems. Springer Berlin Heidelberg.
Lugato, E., Panagos, P., Bampa, F., Jones, A., & Montanarella, L.
(2014). A new baseline of organic carbon stock in European
agricultural soils using a modelling approach. Global Change
Biology, 20, 313–326. https://doi.org/10.1111/gcb.12292
Luo, Z., Wang, E., & Sun, O. J. (2010). Can no-tillage stimulate car-
bon sequestration in agricultural soils? A meta-analysis of
paired experiments. Agriculture, Ecosystems & Environment,
139, 224–231. https://doi.org/10.1016/j.agee.2010.08.006
Mavridis, D., Moustaki, I., Wall, M., & Salanti, G. (2017). Detecting
outlying studies in meta-regression models using a forward
search algorithm: Forward search in meta-analysis. Research
Synthesis Methods, 8, 199–211. https://doi.org/10.1002/jrsm.1197
Mazzoncini, M., Antichi, D., Bene, C. D., Risaliti, R., Petri, M., &
Bonari, E. (2016). Soil carbon and nitrogen changes after
28 years of no-tillage management under Mediterranean condi-
tions 10. The European Journal of Agronomy, 77, 156–165
10.1016/j.eja.2016.02.011.
Meurer, K. H. E., Haddaway, N. R., Bolinder, M. A., & Kätterer, T.
(2018). Tillage intensity affects total SOC stocks in boreo-
temperate regions only in the topsoil—A systematic review
using an ESM approach. Earth-Science Reviews, 177, 613–622.
https://doi.org/10.1016/j.earscirev.2017.12.015
Mihailescu, E., & Bruno Soares, M. (2020). The influence of climate
on agricultural decisions for three European crops: A system-
atic review. Frontiers in Sustainable Food Systems, 4, 64.
https://doi.org/10.3389/fsufs.2020.00064
Mondal, S., Chakraborty, D., Bandyopadhyay, K., Aggarwal, P., &
Rana, D. S. (2020). A global analysis of the impact of zero-
tillage on soil physical condition, organic carbon content, and
plant root response. Land Degradation and Development, 31,
557–567. https://doi.org/10.1002/ldr.3470
Nakagawa, S., Lagisz, M., Jennions, M. D., Koricheva, J.,
Noble, D. W., Parker, T. H., & O'Dea, R. E. (2021). Methods for
testing publication bias in ecological and evolutionary meta-
analyses. https://doi.org/10.32942/osf.io/k7pmz
Nakagawa, S., Noble, D. W. A., Senior, A. M., & Lagisz, M. (2017).
Meta-evaluation of meta-analysis: Ten appraisal questions for
biologists. BMC Biology, 15, 18. https://doi.org/10.1186/s12915-
017-0357-7
Nasta, P., Palladino, M., Sica, B., Pizzolante, A., Trifuoggi, M.,
Toscanesi, M., Giarra, A., D'Auria, J., Nicodemo, F.,
Mazzitelli, C., Lazzaro, U., Di Fiore, P., & Romano, N. (2020).
Evaluating pedotransfer functions for predicting soil bulk den-
sity using hierarchical mapping information in Campania,
Italy. Geoderma Regional, 21, e00267. https://doi.org/10.1016/j.
geodrs.2020.e00267
Ogle, S. M., Alsaker, C., Baldock, J., Bernoux, M., Breidt, F. J.,
McConkey, B., & Vazquez-Amabile, G. G. (2019). Climate and
soil characteristics determine where no-till management can
store carbon in soils and mitigate greenhouse gas emissions.
Scientific Reports, 9, 1–8. https://doi.org/10.1038/s41598-019-
47861-7
Ogle, S. M., Breidt, F. J., & Paustian, K. (2005). Agricultural man-
agement impacts on soil organic carbon storage under moist
and dry climatic conditions of temperate and tropical regions.
Biogeochemistry, 72, 87–121. https://doi.org/10.1007/s10533-
004-0360-2
20 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Ogle, S. M., Swan, A., & Paustian, K. (2012). No-till management
impacts on crop pro- ductivity, carbon input and soil carbon
sequestration. Agriculture, Ecosystems & Environment, 149,
37–49. https://doi.org/10.1016/j.agee.2011.12.010
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I.,
Hoffmann, T. C., Mulrow, C. D., et al. (2021). The PRISMA
2020 statement: An updated guideline for reporting systematic
reviews. BMJ, 372, n71. https://doi.org/10.1136/bmj.n71
Paul, E. A., Paustian, K., Elliott, E. T., & Cole, C. V. (1997). Soil
organic matter in temperate agroecosystems: Long-term experi-
ments in North America (pp. 15–42). CRC Press. https://doi.org/
10.1201/9780367811693
Peel, M. C., Finlayson, B. L., & McMahon, T. A. (2007). Updated
world map of the Köppen-Geiger climate classification. Hydrol-
ogy and Earth System Sciences, 11, 1633–1644. https://doi.org/
10.5194/hess-11-1633-2007
Piccoli, I., Chiarini, F., Carletti, P., Furlan, L., Lazzaro, B., Nardi, S.,
Berti, A., Sartori, L., Dalconi, M. C., & Morari, F. (2016). Disentan-
gling the effects of conservation agriculture practices on the vertical
distribution of soil organic carbon. Evidence of poor carbon seques-
tration in North-Eastern Italy. Agriculture, Ecosystems & Environ-
ment, 230, 68–78. https://doi.org/10.1016/j.agee.2016.05.035
Pisante, M., Stagnari, F., Acutis, M., Bindi, M., Brilli, L., Di
Stefano, V., & Carozzi, M. (2015). Conservation agriculture and
climate change. In M. Farooq & K. H. M. Siddique (Eds.), Con-
servation agriculture (pp. 579–620). Springer International Pub-
lishing. https://doi.org/10.1007/978-3-319-11620-4_22
Poeplau, C., & Don, A. (2015). Carbon sequestration in agricultural
soils via cultivation of cover crops—A meta-analysis. Agricul-
ture, Ecosystems & Environment, 200, 33–41. https://doi.org/10.
1016/j.agee.2014.10.024
Post, W. M., Emanuel, W. R., & Zinke, P. J. G. A.. (1982). Soil car-
bon pools and world life zones. Nature, 298, 156–159. https://
doi.org/10.1038/298156a0.Stangenberger
Powlson, D. S., Stirling, C. M., Thierfelder, C., White, R. P., &
Jat, M. L. (2016). Does conservation agriculture deliver climate
change mitigation through soil carbon sequestration in tropical
agro-ecosystems? Agriculture, Ecosystems & Environment, 220,
164–174. https://doi.org/10.1016/j.agee.2016.01.005
Pravalie, R., Patriche, C., Borrelli, P., Panagos, P., Rosca, B.,
Dumitrascu, M., Nita, I. A., Savulescu, I., Birsan, M. V., &
Bandoc, G. (2021). Arable lands under the pressure of multiple
land degradation processes. A global perspective. Environmental
Research, 194, 110697. https://doi.org/10.1016/j.envres.2020.110697
Prout, J. M., Shepherd, K. D., McGrath, S. P., Kirk, G. J. D., &
Haefele, S. M. (2021). What is a good level of soil organic mat-
ter? An index based on organic carbon to clay ratio. The
European Journal of Soil Science, 72, 2493–2503. https://doi.
org/10.1111/ejss.13012
Prout, J. M., Shepherd, K. D., McGrath, S. P., Kirk, G. J. D.,
Hassall, K. L., & Haefele, S. M. (2022). Changes in organic car-
bon to clay ratios in different soils and land uses in England
and Wales over time. Scientific Reports, 12, 5162. https://doi.
org/10.1038/s41598-022-09101-3
R Core Team. (2018). R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna,
Austria. https://www.R-project.org/
Raudenbush, S. W. (2009). Analyzing effect sizes: Random effects
models. In H. Cooper, L. V. Hedges, & J. C. Valentine (Eds.),
The handbook of research synthesis and meta-analysis (2nd ed.,
pp. 295–315). Russell Sage Foundation.
Rohatgi, A. (2020). WebPlotDigitizer. website: https://automeris.io/
WebPlotDigitizer, version: 4.3, location: Pacifica, California, USA
Rosenthal, R. (1979). The “file drawer problem” and tolerance for
null results. Psychol Bull, 86, 638–641.
Ruiz, I., Almagro, M., García de Jalon, S., Solà, M. D. M., &
Sanz, M. J. (2020). Assessment of sustainable land management
practices in Mediterranean rural regions. The Journal of Envi-
ronmental Management, 276, 111293. https://doi.org/10.1016/j.
jenvman.2020.111293
Schillaci, C., Perego, A., Valkama, E., Märker, M., Saia, S., Veronesi, F.,
Lipani, A., Lombardo, L., Tadiello, T., Gamper, H. A., Tedone, L.,
Moss, C., Pareja-Serrano, E., Amato, G., Kühl, K., Damatirca, C.,
Cogato, A., Mzid, N., Eeswaran, R.,…Acutis, M. (2021). New pedo-
transfer approaches to predict soil bulk density using WoSIS soil
data and environmental covariates in Mediterranean agro-ecosys-
tems. Science of the Total Environment, 146, 609. https://doi.org/10.
1016/j.scitotenv.2021.146609
Senior, A. M., Grueber, C. E., Kamiya, T., Lagisz, M., O'Dwyer, K.,
Santos, E. S. A., & Nakagawa, S. (2016). Heterogeneity in eco-
logical and evolutionary meta-analyses: Its magnitude and
implications. Ecology, 97, 3293–3299. https://doi.org/10.1002/
ecy.1591
Spinoni, J., Vogt, J., Naumann, G., Carrao, H., & Barbosa, P. (2015).
Towards identifying areas at climatological risk of desertifica-
tion using the Köppen-Geiger classification and FAO aridity
index: Towards identifying areas at climatological risk of
desertification. International Journal of Climatology, 35, 2210–
2222. https://doi.org/10.1002/joc.4124
Sterne, J. A. C., & Egger, M. (2001). Funnel plots for detecting bias
in meta-analysis: Guidelines on choice of axis. Journal of Clini-
cal Epidemiology, 54(10), 1046–1055. https://doi.org/10.1016/
s0895-4356(01)00377-8
Sun, W., Canadell, J. G., Yu, L., Yu, L., Zhang, W., Smith, P.,
Fischer, T., & Huang, Y. (2020). Climate drives global soil car-
bon sequestration and crop yield changes under conservation
agriculture. Global Change Biology, 26, 3325–3335. https://doi.
org/10.1111/gcb.15001
Tadiello, T., Perego, A., Valkama, E., Schillaci, C., & Acutis, M.
(2022). Computation of total soil organic carbon stock and its
standard deviation from layered soils. In Computation of total
soil organic carbon stock and its standard deviation from layered
soils (p. 101662). MethodsX. https://doi.org/10.1016/j.mex.2022.
101662
Thornthwaite, C. W. (1948). An approach toward a rational classifi-
cation of climate. Geographical Review, 38, 55–94. https://doi.
org/10.2307/210739
Tiefenbacher, A., Sandén, T., Haslmayr, H.-P., Miloczki, J.,
Wenzel, W., & Spiegel, H. (2021). Optimizing carbon sequestra-
tion in croplands: A synthesis. Agronomy, 11, 882. https://doi.
org/10.3390/agronomy11050882
Toledo, D. M., Galantini, J., Dalurzo, H., Vazquez, S., & Bollero, G.
(2013). Methods for assessing the effects of land use changes on
carbon stocks of subtropical Oxisols. Soil Science Society
of America Journal, 77, 1542–1552. https://doi.org/10.2136/
sssaj2013.03.0087
Tremblay, N., Bouroubi, Y. M., Bélec, C., William, R. M.,
Kitchen, N. R., Thomason, W. E., Ebelhar, S., Mengel, D. B.,
TADIELLO ET AL. 21 of 22
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Raun, W. R., Francis, D. D., Vories, E. D., & Monasterio, I. O.
(2012). Corn response to nitrogen is influenced by soil texture
and weather. Agr. J., 104, 1658–1671. https://doi.org/10.2134/
agronj2012.0184
Underwood, E. C., Klausmeyer, K. R., Cox, R. L., Busby, S. M.,
Morrison, S. A., & Shaw, M. R. (2009). Expanding the global
network of protected areas to save the imperiled Mediterranean
biome. Conservation Biology, 23, 43–52. https://doi.org/10.1111/
j.1523-1739.2008.01072.x
Viechtbauer, W. (2010). Conducting meta-analyses in R with the
metafor package. Journal of Statistical Software, 36(3), 1–48.
https://doi.org/10.18637/jss.v036.i03
Virto, I., Barré, P., Burlot, A., & Chenu, C. (2012). Carbon input dif-
ferences as main factor explaining the variability in soil organic
C storage in no-tilled compared to inversion tilled agrosystems.
Biogeochemistry, 108, 17–26. https://doi.org/10.1007/s10533-
011-9600-4
West, T. O., & Post, W. M. (2002). Soil organic carbon sequestration
rates by tillage and crop rotation: A global data analysis. Soil
Science Society of America Journal, 66, 1930–1946. https://doi.
org/10.2136/sssaj2002.1930
Xu, X., Shi, Z., Li, D., Rey, A., Ruan, H., Craine, J. M., Liang, J.,
Zhou, J., & Luo, Y. (2016). Soil properties control
decomposition of soil organic carbon: Results from data assimi-
lation analysis. Geoderma, 262, 235–242. https://doi.org/10.
1016/j.geoderma.2015.08.038
Zhao, J., Yang, Y., Zhang, K., Jeong, J., Zeng, Z., & Zang, H. (2020).
Does crop rotation yield more in China? A meta-analysis. Field
Crops Research, 245, 107659. https://doi.org/10.1016/j.fcr.2019
SUPPORTING INFORMATION
Additional supporting information can be found online
in the Supporting Information section at the end of this
article.
How to cite this article: Tadiello, T., Acutis, M.,
Perego, A., Schillaci, C., & Valkama, E. (2023). Soil
organic carbon under conservation agriculture in
Mediterranean and humid subtropical climates:
Global meta-analysis. European Journal of Soil
Science, 74(1), e13338. https://doi.org/10.1111/ejss.
13338
22 of 22 TADIELLO ET AL.
 13652389, 2023, 1, D
ow
nloaded from
 https://bsssjournals.onlinelibrary.wiley.com/doi/10.1111/ejss.13338 by Luonnonvarakeskus, W
iley Online Library on [03/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License