ArticlePDF Available
HAL Id: hal-01495108
https://hal-agroparistech.archives-ouvertes.fr/hal-01495108
Submitted on 16 Jan 2020
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Increased soil organic carbon stocks under agroforestry:
A survey of six dierent sites in France
Rémi Cardinael, Tiphaine Chevallier, Aurelie Cambou, Camille Beral, Bernard
G. Barthès, Christian Dupraz, Celine Durand, Ernest Kouakoua, Claire Chenu
To cite this version:
Rémi Cardinael, Tiphaine Chevallier, Aurelie Cambou, Camille Beral, Bernard G. Barthès, et
al.. Increased soil organic carbon stocks under agroforestry: A survey of six dierent sites
in France. Agriculture, Ecosystems and Environment, Elsevier Masson, 2017, 236, pp.243–255.
�10.1016/j.agee.2016.12.011�. �hal-01495108�
1
Increased soil organic carbon stocks under agroforestry: a survey of six different sites in
1
France
2
Rémi Cardinaela,b,c, Tiphaine Chevalliera*, Aurélie Camboua,d, Camille Bérale, Bernard G.
3
Barthèsa, Christian Duprazf, Céline Duranda, Ernest Kouakouaa, Claire Chenub
4
a IRD, UMR Eco&Sols, Montpellier SupAgro, 2 place Viala, 34060 Montpellier, France
5
b AgroParisTech, UMR Ecosys, Avenue Lucien Brétignières, 78850 Thiverval-Grignon, France
6
c CIRAD, UPR AIDA, Avenue d’Agropolis, 34398 Montpellier, France (present address)
7
d AgroCampus Ouest centre d'Angers, UPSP EPHor, 2 Rue André le Nôtre, 49045 Angers,
8
France (present address)
9
e Agroof, 9 plan de Brie, 30140 Anduze, France
10
f INRA, UMR System, Montpellier SupAgro, 2 place Viala, 34060 Montpellier, France
11
* Corresponding author. Tel.: +33 04.67.61.53.08. E-mail address: remi.cardinael@cirad.fr
12
13
ABSTRACT
14
Agroforestry systems are land use management systems in which trees are grown in
15
combination with crops or pasture in the same field. In silvoarable systems, trees are
16
intercropped with arable crops, and in silvopastoral systems trees are combined with pasture
17
for livestock. These systems may produce forage and timber as well as providing ecosystem
18
services such as climate change mitigation. Carbon (C) is stored in the aboveground and
19
belowground biomass of the trees, and the transfer of organic matter from the trees to the soil
20
can increase soil organic carbon (SOC) stocks. Few studies have assessed the impact of
21
agroforestry systems on carbon storage in soils in temperate climates, as most have been
22
2
undertaken in tropical regions. This study assessed five silvoarable systems and one
23
silvopastoral system in France. All sites had an agroforestry system with an adjacent, purely
24
agricultural control plot. The land use management in the inter-rows in the agroforestry systems
25
and in the control plots were identical. The age of the study sites ranged from 6 to 41 years after
26
tree planting. Depending on the type of soil, the sampling depth ranged from 20 to 100 cm and
27
SOC stocks were assessed using equivalent soil masses. The aboveground biomass of the trees
28
was also measured at all sites. In the silvoarable systems, the mean organic carbon stock
29
accumulation rate in the soil was 0.24 (0.09-0.46) Mg C ha-1 yr-1 at a depth of 30 cm and 0.65
30
(0.004-1.85) Mg C ha-1 yr-1 in the tree biomass. Increased SOC stocks were also found in deeper
31
soil layers at two silvoarable sites. Young plantations stored additional SOC but mainly in the
32
soil under the rows of trees, possibly as a result of the herbaceous vegetation growing in the
33
rows. At the silvopastoral site, the SOC stock was significantly greater at a depth of 30 to 50
34
cm than in the control. Overall, this study showed the potential of agroforestry systems to store
35
C in both soil and biomass in temperate regions.
36
37
Keywords: Alley cropping, Soil organic carbon storage, Equivalent soil mass, Aboveground
38
biomass, Belowground biomass
39
40
1. Introduction
41
Soils play an essential role in the global carbon budget (Houghton, 2007). Currently, the land
42
sink (including soil and vegetation) absorbs about 30% of the carbon (C) emitted to the
43
atmosphere through the burning of fossil fuel and cement production (Le Quéré et al., 2014).
44
Since 1850, the depletion of soil organic carbon (SOC) in cultivated lands has transferred about
45
3
70 Gt C to the atmosphere (Amundson, 2001; Lal, 2004a). The potential of these SOC depleted
46
soils as future C sinks through SOC sequestration has now been recognized (Paustian et al.,
47
1997; Freibauer et al., 2004; Smith, 2004). In France, SOC stocks have been estimated at 3.1-
48
3.3 Gt C in the top 30 cm of soils (Arrouays et al., 2001; Martin et al., 2011). Based on the SOC
49
saturation capacity (Hassink, 1997), assuming that the quantity of stable SOC is limited by the
50
amount of fine particles, Angers et al. (2011) found that the median saturation deficit of French
51
arable topsoils was 8.1 g C kg-1 soil. About 70% of French agricultural topsoils are, therefore,
52
unsaturated in SOC and have the potential for additional SOC storage. Increasing SOC stocks
53
is often seen as a win-win strategy (Lal, 2004a; Janzen, 2006) as it allows the transfer of CO2
54
from the atmosphere to the soil while improving soil quality and fertility (Lal, 2004b).
55
Several agricultural practices have been developed to increase SOC stocks. For instance, the
56
introduction of cover crops (Constantin et al., 2010; Poeplau and Don, 2015) or grasslands
57
(Conant et al., 2001; Soussana et al., 2004) in the cropping sequence has proven effective. The
58
effect of no-till farming on SOC stocks is disputed and highly variable (Luo et al., 2010; Virto
59
et al., 2012; Dimassi et al., 2013) and seems to depend on the amount of C transferred from the
60
crops to the soil (Virto et al., 2012). Agroforestry is a general term for agroecosystems in which
61
trees are intercropped with crops or pasture (Nair, 1993). Silvoarable systems intercrop trees
62
and arable crops and silvopastoral systems combine trees, pasture and livestock. These are
63
recognized as possible land use management systems that can maintain or increase SOC stocks,
64
both in tropical (Albrecht and Kandji, 2003) and temperate regions (Peichl et al., 2006;
65
Bambrick et al., 2010; Wotherspoon et al., 2014). However, most studies only consider the
66
surface soil layers (to a depth of < 20 or 30 cm) whereas trees grown in agroforestry can be
67
very deep rooted (Mulia and Dupraz, 2006; Cardinael et al., 2015a) and affect deep SOC stocks.
68
A recent study in the Mediterranean region of France showed that an 18-year-old silvoarable
69
system with hybrid walnuts intercropped with durum wheat increased SOC stocks by 0.25 ±
70
4
0.03 Mg C ha-1 yr-1 in the 0-30 cm layer and by 0.35 ± 0.04 Mg C ha-1 yr-1 from 0 to 100 cm
71
compared to an adjacent agricultural plot (Cardinael et al., 2015b). Furthermore, although trees
72
affect the spatial distribution of organic matter inputs to the soil (Rhoades, 1997), sampling
73
protocols have not always taken account of the potential impact on the spatial distribution of
74
SOC stocks. Some authors showed that SOC stocks were greater in the tree rows than in the
75
inter-rows, and found no gradients within the inter-rows (Peichl et al., 2006; Upson and
76
Burgess, 2013). Bambrick et al., (2010) found that the spatial distribution of SOC stocks varied
77
with the time after tree planting. Few studies have estimated SOC storage in agroforestry
78
systems in temperate conditions (Howlett et al., 2011; Mosquera Losada et al., 2011; Upson
79
and Burgess, 2013) and these studies sometimes do not have control plots without trees for
80
comparison, making it difficult to evaluate the precise effect of agroforestry on SOC stocks
81
(Pellerin et al., 2013).
82
This study set out i) to quantify organic carbon stocks in soils and in the tree biomass in six
83
agroforestry systems with adjacent agricultural control plots under different soil and climate
84
conditions in France, ii) to study the spatial distribution of SOC stocks as a function of the
85
distance from individual trees and the tree rows and iii) to estimate the SOC stock accumulation
86
rates for these agroforestry systems.
87
88
2. Materials and methods
89
2.1 The six agroforestry sites
90
Each study site had an agroforestry system and an adjacent agricultural control plot. Before tree
91
planting, the agroforestry plot was part of the agricultural plot, with the same soil use and
92
management (crop rotation, fertilization, soil tillage). After tree planting, the soil management
93
of the agroforestry inter-rows and of the agricultural plot remained identical. Rows of trees
94
5
were planted in the agroforestry fields, with natural or sown grasses between the trees. Five
95
sites, Restinclières (RE), Châteaudun (CH), Melle (ME), Saint-Jean d’Angely (SJ), and
96
Vézénobres (VE), were silvoarable systems with no grazing. Only one site, Theix (TH), was a
97
silvopastoral system with regular grazing. Four sites were owned and managed by farmers and
98
Restinclières (RE) and Theix (TH) were experimental research sites.
99
100
101
102
Figure 1. Location and description of the six study cases under agroforestry systems sampled
103
in France.
104
105
106
107
6
Table 1 Site characteristics.
108
109
CH: Châteaudun, ME: Melle, SJ: Saint-Jean-d’Angély, VE: Vézénobres, RE: Restinclières, TH: Theix.
110
111
112
Mean annual temperature
(°C)
Mean annual rainfall
(mm)
Soil type
(FAO)
Soil depth
(cm)
Soil texture
clay/silt/sand (g kg-1)
Soil pH
in water
Agroforestry
Control
11.1
595
Luvisol
0-30
200/700/100
190/710/100
7.0
11.7
810
Luvisol
0-30
240/660/100
260/630/110
5.8
12.9
850
Luvisol
0-20
560/370/70
500/410/90
7.7
14.5
1037
Fluvisol
0-30
110/410/480
90/370/540
8.3
30-60
100/440/460
80/370/550
8.3
15.4
873
Fluvisol
0-30
173/406/421
176/413/411
8.0
30-50
178/416/406
177/421/402
8.1
50-70
250/501/249
243/507/250
8.2
70-100
309/582/109
307/586/107
8.3
7.7
800
Andosol
0-20
340/300/360
380/360/260
6.5
20-50
320/280/400
360/380/260
6.5
7
Table 2 Description of the agroforestry plots.
113
CH: Châteaudun, ME: Melle, SJ: Saint-Jean-d’Angély, VE: Vézénobres, RE: Restinclières, TH: Theix.
114
Site
Tree species
Age
(yrs)
Density
(trees ha-1)
Distance between
trees in tree rows
(m)
Width of
inter-rows
(m)
Width of
tree rows
(m)
Area occupied by tree
rows in the AF plot
(%)
Crops
CH
Hybrid walnut
6
34
10
24
2
8
wheat, rapeseed
ME
Hybrid walnut
6
35
8
27
2
7
wheat, rapeseed,
sunflower
SJ
Black walnut
41
102
7
12
2
14
sunflower, wheat,
barley
VE
Hybrid walnut
18
100
10
9
2
18
rapeseed, wheat,
potato, garlic
RE
Hybrid walnut
18
110
4-12
11
2
16
durum wheat,
rapeseed,
chickpea
TH
Wild cherry
26
200
7
No row
No row
No row
ryegrass, fescue
8
The CH silvoarable site was located in Châteaudun (Fig. 1), in the department of Eure-et-Loir
115
(longitude 1°17’58’’ E, latitude 48°06’08’’ N, elevation 147 m a.s.l.). The mean temperature
116
was 11.1°C and the mean annual rainfall 595 mm (years 2001-2013, INRA CLIMATIK,
117
https://intranet.inra.fr/climatik). The soil was a silty loam Luvisol (IUSS Working Group WRB,
118
2007) (Table 1). Hybrid walnut trees (Juglans regia × nigra cv. NG23) were planted in
119
February 2008 at a density of 34 trees ha-1. The trees were planted 10 m apart within the rows,
120
with 26 m between rows. A mix of ryegrass (Lolium perenne L.) and tall fescue (Festuca
121
arundinacea Schreb.) was sown in August 2007 in two meter wide strips along the tree rows
122
before the trees were planted. After tree planting, wheat (Triticum aestivum L. subsp. aestivum)
123
and rapeseed (Brassica napus L.) were grown in rotation in the control plot and in the inter-
124
rows (Table 2). The mean fresh grain yield was 7.5-8 t ha-1 for wheat, and 3.8 t ha-1 for rapeseed.
125
All crop residues were left in the field after harvest. The agroforestry inter-rows and the control
126
plot were ploughed every three years to a depth of 22 cm and harrowed to 8 cm the other years.
127
The ME silvoarable site was located in Melle (Fig. 1), in the department of Deux-Sèvres
128
(longitude 0°10’37’’ W, latitude 46°11’54’’ N, elevation 107 m a.s.l.). The mean temperature
129
was 11.7°C and the mean annual rainfall 810 mm (years 1990-2013, INRA CLIMATIK,
130
https://intranet.inra.fr/climatik). The soil was a silty loam Luvisol (IUSS Working Group WRB,
131
2007) (Table 1). Hybrid walnut trees (Juglans regia × nigra cv. NG23) were planted in 2008 at
132
a density of 35 trees ha-1. The trees were planted 8 m apart within the rows, with 29 m between
133
rows. Sheep fescue (Festuca ovina L.) was sown in 2008 in two meter wide strips along the
134
tree rows before the trees were planted. After tree planting, wheat (Triticum aestivum L. subsp.
135
aestivum), rapeseed (Brassica napus L.) and sunflower (Helianthus annuus L.) were grown in
136
rotation in the control plot and in the inter-rows (Table 2). The mean fresh grain yield was 8-
137
8.5 t ha-1 for wheat, 3.3 t ha-1 for rapeseed and 2.5 t ha-1 for sunflower. Crop residues were
138
usually exported, but this was counterbalanced by the application of manure in both the
139
9
agroforestry inter-rows and the control plot (the farmer was unable to specify the application
140
rates, but they were similar for both plots). Before the spring crop (sunflower), a winter cover
141
crop was sown to prevent soil erosion and nitrate leaching. This cover crop was a mix of radish
142
(Raphanus sativus L.), phacelia (Phacelia tanacetifolia Benth.) and mustard (Sinapis alba L.).
143
The soil was ploughed every year to a depth of 20 cm in both the agroforestry inter-rows and
144
the control plot. The agroforestry system was established on a moderate slope, while the control
145
plot was flat.
146
The SJ silvoarable site was located in Saint-Jean-d’Angély (Fig. 1), in the department of
147
Charente-Maritime (longitude 0°13’57’’ W, latitude 46°00’39’’ N, elevation 152 m a.s.l.). The
148
mean temperature was 12.9°C and the mean annual rainfall 850 mm (years 1990-2013, INRA
149
CLIMATIK, https://intranet.inra.fr/climatik). The soil was a carbonated silty clay Luvisol
150
(IUSS Working Group WRB, 2007) (Table 1). Black walnut trees (Juglans nigra L.) were
151
planted in 1973 at a density of 102 trees ha-1. The trees were planted 7 m apart within the tree
152
rows, with 14 m between rows. The rows of trees were two meters wide, and covered by
153
spontaneous herbaceous vegetation. After tree planting, sunflower (Helianthus annuus L.),
154
wheat (Triticum aestivum L. subsp. aestivum) and barley (Hordeum vulgare L.) were grown in
155
rotation in the control plot and in the inter-rows (Table 2). Crop residues were left in the field
156
after harvest. The soil was ploughed every three years to a depth of 10-20 cm in both the
157
agroforestry inter-rows and the control plot.
158
The VE silvoarable site was located in Vézénobres (Fig. 1), in the department of Gard
159
(longitude 4°06’37’’ E, latitude 46°00’39’’ N, elevation 102 m a.s.l.). The climate was sub-
160
humid Mediterranean with a mean temperature of 14.5°C and a mean annual rainfall of 1037
161
mm (mean 1995-2007, experimental site weather station). The soil was a deep sandy loam
162
alluvial Fluvisol (IUSS Working Group WRB, 2007) (Table 1) originating from deposits from
163
the granitic Cevennes mountain range and was, therefore, not calcareous. Hybrid walnut trees
164
10
(Juglans regia × nigra cv. NG23) were planted in 1995 at a density of 100 trees ha-1. The trees
165
were planted 10 m apart with the rows, with 10 m between rows. The tree rows were two meters
166
wide and were covered by spontaneous herbaceous vegetation. In the inter-rows, rapeseed
167
(Brassica napus L.) and wheat (Triticum aestivum L. subsp. aestivum) were grown in rotation
168
until 2010 (Table 2). In 2011, the farm changed over to organic farming and potatoes were
169
planted (Solanum tuberosum L.). In 2012 garlic (Allium sativum L.) was grown in the inter-
170
rows. In 2013 the inter-rows were left fallow and in 2014 sunflower (Helianthus annuus L.)
171
was sown. The same crops were grown in the control plot, except in 2011 when wheat (Triticum
172
aestivum L. subsp. aestivum) was sown and in 2012 when the control was left fallow. The soil
173
was occasionally ploughed to a depth of 20 cm in both the agroforestry inter-rows and the
174
control plot.
175
The RE site was located in Prades-le-Lez, at the Restinclières experimental site (Fig. 1), in the
176
department of Hérault (longitude 04°01’ E, latitude 43°43’ N, elevation 54 m a.s.l.). A full
177
description of this site is given in the study by Cardinael et al. (2015b). The climate was sub-
178
humid Mediterranean with a mean temperature of 15.4°C and a mean annual rainfall of 873
179
mm (years 19952013, experimental site weather station). The soil was a deep carbonated
180
sandy loam Fluvisol (IUSS Working Group WRB, 2007) (Table 1). Hybrid walnut trees
181
(Juglans regia × nigra cv. NG23) were planted in 1995 and the density was 110 trees ha-1 at the
182
time of the study (Table 2). The trees were planted 4 to 8 m apart along the rows with 13 m
183
between rows. The two meter wide tree rows were covered by spontaneous herbaceous
184
vegetation. They were mainly intercropped with durum wheat (Triticum turgidum L. subsp.
185
durum) but also with rapeseed (Brassica napus L.) and chickpea (Cicer arietinum L.). The soil
186
was regularly ploughed to a depth of 20 cm in both the agroforestry inter-rows and the control
187
plot.
188
11
The TH silvopastoral site was located at the Theix experimental site (Fig. 1), in the department
189
of Puy-de-Dôme (longitude 3°01’39’’ E, latitude 45°42’58’’ N, elevation 829 m a.s.l.). The
190
mean temperature was 7.7°C and the mean annual rainfall 800 mm (years 1990-2013, INRA
191
CLIMATIK, https://intranet.inra.fr/climatik). The soil was a clay loam Andosol (IUSS
192
Working Group WRB, 2007) (Table 1). Wild cherry trees (Prunus avium L.) were planted in
193
1988 at a density of 200 trees ha-1 on a natural permanent pasture. The trees were planted 7 m
194
apart and the soil was uniformly covered by a permanent pasture, mainly ryegrass (Lolium
195
perenne L.) and fescue (Festuca sp.), in both the control and agroforestry plots (Table 2). There
196
was no distinction between tree rows and inter-rows in terms of soil cover and management.
197
The pasture was regularly grazed by sheep in both the control and agroforestry plots.
198
199
2.2 Soil sampling protocol
200
The sampling protocol was defined to allow for the spatial distribution of SOC stocks owing to
201
the presence of trees and rows of trees, with sampling points at varying distances from the trees.
202
The agroforestry designs varied between sites with different distances between the trees within
203
the rows and between the rows. The sampling protocol was flexible to take account of these
204
differences but consistent enough to allow comparisons between sites. A sampling pattern was
205
defined with sampling points in transects around one tree. This was a rectangle with dimensions
206
, where L is the distance between tree rows and d is the distance between trees in the rows
207
(Fig. 2). This pattern is a quarter of the Voronoi polygon which is the elementary space defined
208
by the half distances between the sampled tree and its neighbors, as commonly used to estimate
209
root biomass (Levillain et al., 2011; Picard et al., 2012). At all sites, nine soil samples per
210
pattern were taken at fixed positions around the trees, at 1, 2 and 3 m in the tree row, in the
211
inter-row in front of the tree, and in the inter-row between two trees. If L ≥ 8 m, soil samples
212
12
were additionally taken at mid-distance
, and, if L ≥ 16 m, soil samples were also taken at
. If
213
d ≥ 8 m, soil samples were also taken at
. This sampling pattern was applied three times in the
214
agroforestry plots at all sites. Two sampling patterns were oriented north of the tree rows (if the
215
rows were oriented east-west) or west of the rows (if the rows were oriented north-south) and
216
one sampling pattern was oriented south or east, respectively. Thirty-six sampling points were,
217
therefore, defined for the agroforestry plot at the CH site, twenty-four at the SJ site, thirty at the
218
VE site, and twenty-seven at the TH site (Table 3). In the control plots, a simpler sampling
219
pattern was applied in triplicate. This pattern was a rectangle with dimensions
, with soil
220
samples taken at each corner (12 sampling points).
221
At the ME site, the agricultural control plot was flat, whereas the agroforestry plot was on a
222
moderate slope. The SOC-rich topsoil in the agroforestry plot might, therefore, have been
223
eroded before the start of the experiment. To take account of this topography difference, six
224
soil samples from the middle of the inter-row (two sampling positions for each of the three
225
sampling patterns) were used as an alternate arable control. Because the inter-rows were 27 m
226
wide and the 6-year-old trees were only 3 m high, the soil in the middle of the inter-rows had
227
probably not yet been affected by the presence of trees. 30 sampling points were defined in the
228
agroforestry plot and 6 in the control plot (Table 3).
229
The RE site had been the subject of a previous study (Cardinael et al., 2015b) to map SOC
230
stocks at plot scale. The sampling protocol at this site was, therefore, very dense: 100 soil
231
samples were taken from the agroforestry plot and 93 from the control plot (Table 3). Sampling
232
points were located every 5 m along a regular grid (25 × 25 m), and at 1, 2 and 3 m around nine
233
walnut trees, in the inter-rows and in the tree rows.
234
The sampling depths were 30 cm at the CH and ME sites, 20 cm at the SJ site, 60 at the VE
235
site, 100 at the RE site and 50 cm at the TH site. At the SJ site, the sampling depth corresponded
236
13
to the maximum soil depth. Soil samples were taken every 10 cm depth from the surface, except
237
at the RE site (at 10 cm and every 20 cm from 10 cm).
238
239
Figure 2. Sampling pattern for the agroforestry sites (except for the RE site). L is the distance
240
between tree rows, d is the distance between trees on the rows.
241
242
14
2.3 Bulk density measurement
243
The soil samples were collected in April 2014 at all sites, except at the RE site which was
244
sampled in May 2013. Soil samples were taken every 10 cm from the surface using a 500-cm3
245
cylinder, except at the RE site where soil samples were taken every 20 cm depth after the top
246
10 cm. After air-drying in the lab, the soil samples were oven-dried at 105°C for 48 hours,
247
sieved to 2 mm and weighed without coarse particles > 2 mm. The bulk density (g cm-3) was
248
calculated as the ratio of the dry mass of fine soil (< 2 mm) to the cylinder volume.
249
250
2.4 Organic carbon analysis
251
The soil samples were dried at 40°C and ball milled until they passed through a 200 µm mesh
252
sieve. The presence of inorganic carbon was tested with HCl. If the soil contained inorganic
253
carbon, carbonates were removed by acid fumigation, as described in Harris et al. (2001). This
254
was the case for samples from the SJ and RE sites. 30 mg of soil were placed in open Ag-foil
255
capsules. The capsules were then placed in the wells of a microtiter plate and 50 µL of
256
demineralized water was added to each capsule. The microtiter plate was placed in a vacuum
257
desiccator with a beaker filled with 100 mL of concentrated HCl. The samples were exposed to
258
HCl vapor for 8 h and then dried at 40°C for 48 h. The capsules were then enclosed in a bigger
259
tin capsule. All samples were analyzed for organic carbon concentration using a CHN elemental
260
analyzer (Carlo Erba NA 2000, Milan, Italy).
261
262
2.5 SOC stock calculation
263
The SOC stock at soil sample level (mg C cm-3) is defined as the product of the SOC
264
concentration (mg C g-1) and the bulk density (g cm-3) and is then calculated for each soil profile
265
15
(kg C m-2) by summing the SOC stocks in the samples through the profile. For each site, the
266
SOC stocks were calculated on an equivalent soil mass (ESM) basis (Ellert and Bettany, 1995)
267
to enable comparison between all locations (control, tree rows, inter-rows) even where the soil
268
bulk density varied within the same site. SOC stocks in the agroforestry plot (Mg C ha-1) were
269
calculated by adding the tree row and inter-row SOC stocks, weighted by their respective
270
relative surface areas:
271
         
 
272
273
where p is the percentage of tree row surface area in the agroforestry plot (Table 2).
274
275
The delta SOC stock (Mg C ha-1) at a given depth was expressed as the difference in the SOC
276
stock between the agroforestry and the control plot:
277
     
278
279
The SOC stock accumulation rates under an agroforestry system at a given depth was calculated
280
by dividing the delta SOC stock by the number of years since tree planting.
281
282
283
2.6 Tree aboveground and belowground biomass
284
At each site, 10 to 20 trees were measured to estimate the aboveground biomass. As the trees
285
in the farmers’ fields could not be felled, the aboveground biomass was estimated by
286
multiplying the volume of the trunk and branches by the wood density, using the global wood
287
density database (Chave et al., 2009). The trunk volume was estimated as the sum of the volume
288
of three truncated cones, from the soil surface up to 1.30 m, from 1.30 m to the first branch and
289
16
from the first branch to the top of the tree. The trunk diameter was measured 5 cm above the
290
soil surface, at 1.30 m (Diameter at Breast Height, DBH) and below the first branch. The total
291
height (Htot) and merchantable height (H) of the trees were also measured. The volume of the
292
first order branches (branches arising directly off the trunk) was also estimated by measuring
293
the diameter of the branches at the trunk and the length of the branches and branch volumes
294
were calculated as cone volumes. For the RE site, three trees were felled to measure the trunk
295
and branch biomass directly. The carbon concentrations of the trunk and branches of the
296
Juglans regia × nigra cv. NG23 were measured. As it was not possible to sample wood from
297
the tree trunks at the other sites, the C concentrations were considered to be the same for Prunus
298
avium and Juglans nigra. This simplification was possible because these trees are slow growing
299
species and there is usually little variation in their wood C concentration (462.7 to 499.7 mg C
300
g-1 DM) (Lamlom and Savidge, 2003). It was also assumed that young and old trees had the
301
same wood density and C concentration.
302
So far as we are aware, there is no allometric equation for estimating the belowground biomass
303
of temperate agroforestry trees and so the equation proposed by Cairns et al. (1997) for
304
temperate forests was used:
305
    
306
where RB is the total root biomass (Mg C ha-1), AB is the aboveground biomass (Mg C ha-1)
307
and Age is the age of the plantation (yr).
308
309
2.7 Statistical analyses
310
The influence of the sampling location in the inter-rows (in front of a tree or between two trees)
311
on the SOC concentration, bulk density and SOC stock was determined using mixed effects
312
17
models. This analysis was done at each site using the nlme package (Pinheiro et al., 2013). An
313
ANOVA was performed on these models. Mixed effects models were then fitted for each site
314
using the whole soil data set. The SOC concentration, bulk density and SOC stock were
315
compared as a function of depth, location (control, tree row, inter-row) and distance from the
316
closest tree. An ANOVA was performed on these models. The SOC stock were compared
317
between tree rows and inter-rows, between inter-rows and the control plot and between the
318
agroforestry plot and the control plot. The statistical analyses were performed using R version
319
3.1.1 (R Development Core Team, 2013), at a significance level of < 0.05.
320
321
3. Results
322
3.1 Soil bulk density
323
At all sites, the soil bulk density increased significantly with increasing soil depth (Table 3, S1).
324
In the top 30 cm, the bulk density ranged from 0.7 to 1.6 g cm-3 depending on the site. There
325
was no significant difference in bulk density between the tree row and the inter-row except in
326
the top 10 cm at the ME, SJ and RE sites, where it was lower in the tree row than in the inter-
327
row and in the control (Table 3, S1). There was no significant difference between the control
328
and the inter-row at any depth, except at the RE site where the bulk density was higher in the
329
top 10 cm in the control plot (Table 3, S1).. The distance from the closest tree had no significant
330
effect on the bulk density except at the SJ site (Table S1). There was no significant difference
331
in the inter-row between samples collected in front of a tree or between two trees at any of the
332
sites or at any depth (p-value 0.18), except at the ME site (p-value = 0.03) (Table S1).
333
334
335
18
Table 3 Mean soil bulk density (g cm-3) and mean soil organic carbon (SOC) concentrations (mg C g-1) with associated standard errors.
336
Number of soil samples
Bulk density (g cm-3)
SOC concentration (mg C g-1)
Site
Soil depth (cm)
Tree row
Inter-row
Control
Tree row
Inter-row
Control
Tree row
Inter-row
Control
CH
0-10
12
24
12
1.09 ± 0.03
1.10 ± 0.02
1.18 ± 0.02
19.44 ± 1.00
16.44 ± 0.26
14.88 ± 0.38
10-20
12
24
12
1.12 ± 0.02
1.13 ± 0.02
1.16 ± 0.03
13.58 ± 0.31
14.39 ± 0.34
14.56 ± 0.48
20-30
12
24
12
1.15 ± 0.02
1.20 ± 0.01
1.25 ± 0.02
11.76 ± 0.65
12.07 ± 0.48
11.78 ± 0.35
ME
0-10
12
18
6
1.04 ± 0.03
1.27 ± 0.02
1.31 ± 0.01
21.30 ± 0.63
13.01 ± 0.19
12.80 ± 0.43
10-20
12
18
6
1.28 ± 0.02
1.29 ± 0.02
1.37 ± 0.03
13.14 ± 0.26
12.03 ± 0.50
12.02 ± 0.40
20-30
12
18
6
1.21 ± 0.01
1.34 ± 0.01
1.35 ± 0.02
10.35 ± 0.21
8.38 ± 0.44
8.68 ± 0.93
SJ
0-10
8
16
12
0.67 ± 0.03
0.76 ± 0.02
0.78 ± 0.01
58.60 ± 1.88
49.49 ± 1.28
32.89 ± 0.33
10-20
8
16
12
0.84 ± 0.03
0.78 ± 0.03
0.88 ± 0.04
35.60 ± 0.82
32.01 ± 0.67
24.86 ± 1.12
VE
0-10
12
18
10
1.06 ± 0.04
0.98 ± 0.03
0.91 ± 0.02
17.25 ± 0.49
15.95 ± 0.37
15.00 ± 1.11
10-20
12
18
10
1.12 ± 0.02
1.18 ± 0.02
1.24 ± 0.03
13.72 ± 0.40
13.50 ± 0.49
13.19 ± 0.70
20-30
12
18
10
1.16 ± 0.03
1.25 ± 0.01
1.31 ± 0.02
11.38 ± 0.30
10.83 ± 0.25
10.89 ± 0.68
30-40
12
18
10
1.29 ± 0.04
1.39 ± 0.02
1.47 ± 0.04
10.82 ± 0.27
10.31 ± 0.29
8.55 ± 0.78
40-50
12
18
10
1.30 ± 0.05
1.37 ± 0.03
1.34 ± 0.03
10.52 ± 0.33
8.25 ± 0.35
5.79 ± 0.69
50-60
12
18
10
1.36 ± 0.04
1.39 ± 0.04
1.37 ± 0.06
9.74 ± 0.35
7.16 ± 0.62
5.28 ± 0.86
RE
0-10
40
60
93
1.10 ± 0.02
1.23 ± 0.03
1.41 ± 0.01
21.59 ± 0.76
9.78 ± 0.13
9.33 ± 0.06
10-30
40
60
93
1.49 ± 0.01
1.60 ± 0.02
1.61 ± 0.00
10.16 ± 0.16
9.57 ± 0.12
8.94 ± 0.05
30-50
40
60
93
1.71 ± 0.01
1.67 ± 0.02
1.73 ± 0.00
7.29 ± 0.15
6.95 ± 0.11
6.82 ± 0.10
50-70
40
60
93
1.73 ± 0.01
1.77 ± 0.01
1.80 ± 0.00
6.07 ± 0.11
5.89 ± 0.07
5.77 ± 0.06
70-100
40
60
93
1.68 ± 0.00
1.71 ± 0.00
1.74 ± 0.00
6.49 ± 0.16
6.29 ± 0.06
6.09 ± 0.06
0-10
27
10
0.75 ± 0.02
0.69 ± 0.02
64.00 ± 2.40
67.83 ± 2.45
10-20
27
10
0.79 ± 0.01
0.75 ± 0.01
46.97 ± 1.15
49.31 ± 0.89
TH
20-30
27
10
0.80 ± 0.02
0.73 ± 0.02
38.82 ± 0.88
40.56 ± 0.86
30-40
27
10
0.82 ± 0.01
0.78 ± 0.02
32.90 ± 0.70
29.92 ± 0.75
40-50
19
10
0.80 ± 0.01
0.79 ± 0.03
28.65 ± 0.76
22.69 ± 1.25
337
19
At the TH silvopastoral site, no distinction was made between tree rows and inter-rows (uniform cover), values are for the whole agroforestry plot.
338
CH: Châteaudun, ME: Melle, SJ: Saint-Jean-d’Angély, VE: Vézénobres, RE: Restinclières, TH: Theix.
339
340
20
341
Figure 3. Soil organic carbon concentration (mg C g-1) at the different sites. Transparent rectangles represent standard errors. At the TH
342
silvopastoral site, no distinction was made between tree rows and inter-rows (uniform cover), values are for the whole agroforestry plot.
343
21
3.2 Soil organic carbon concentration
344
The SOC concentration decreased significantly with increasing soil depth, except in the
345
ploughed layer, where it was uniform (Table 3, S1). At all sites, the SOC concentration in the
346
top 10 cm was significantly higher in the tree row than in the inter-row (Fig. 3). However, there
347
was no significant difference in the inter-row between samples collected in front of a tree and
348
between two trees (p-value 0.32) at any site and at any depth. The SOC concentration
349
depended significantly on the distance from the trees only at the oldest site (SJ, p-value < 0.001)
350
(Table S1). At sites CH, SJ and RE, the SOC concentration in the top 10 cm was significantly
351
higher in the inter-rows than in the control plot (Fig. 3, Table 3). At the VE and RE silvoarable
352
sites, the SOC concentration was significantly higher in the inter-row than in the control below
353
30 cm (Fig. 3, Table 3). At the TH silvopastoral site, the SOC concentration below 30 cm was
354
also significantly higher in the silvopasture than in the tree-less pasture (Fig. 3, Table 3).
355
356
3.3 Soil organic carbon stock
357
The SOC stock was mainly influenced by depth and location (Table S1). In the inter-row, there
358
was no significant difference between samples collected in front of a tree and between two trees
359
(p-value 0.30). The distance from the closest tree had no significant effect on the SOC stock
360
(p-value 0.5) except at the SJ site (p-value = 0.005) (Table S1). In the silvoarable systems,
361
the SOC stock was significantly higher in the tree rows than in the inter-rows in the top 10 cm,
362
even in young plantations (CH and ME sites) (Fig. 4). The SOC stock was also significantly
363
higher in the inter-rows than in the control at depths of 10 cm at the CH site, 20 cm at the SJ
364
site and 30 cm at the RE site, as happened for SOC concentration (Fig. 4). Unlike, at the VE
365
site, the SOC stock was higher in the inter-rows than in the control below 30 cm (Fig. 4). At
366
22
the TH silvopastoral site, the SOC stock below 30 cm was higher in the agroforestry plot than
367
in the control.
368
In the top 30 cm, the delta SOC stock between silvoarable systems and control plots was
369
significantly positive except at the ME and VE sites (Table 4). For the silvoarable sites, the
370
delta SOC stock ranged from 0.5 to 4.5 Mg C ha-1 in the top 30 cm (Table 4), and was about 19
371
Mg C ha-1 in the top 20 cm for the oldest silvoarable system (SJ). At the RE and VE silvoarable
372
sites, the delta SOC stock was significantly positive below 30 cm depth. At the TH silvopastoral
373
site, the delta SOC stock was not significantly different in the top 30 cm (-0.16 ± 0.25 Mg C
374
ha-1) but was significantly positive for the whole soil profile (0.49 ± 0.27 Mg C ha-1) down to
375
60 cm (Table 4).
376
377
3.4 Carbon stock in the tree biomass
378
The wood density of Juglans regia × nigra cv. NG23 was 0.62 g cm-3, that of Juglans nigra
379
was 0.59 g cm-3 and that of Prunus avium was 0.54 g cm-3. The C concentrations of the trunk
380
and branches of 18-year-old Juglans regia × nigra cv. NG23 were 445.71 ± 1.04 and 428.64 ±
381
1.70 mg C g-1 DM, respectively. At the silvoarable sites, the organic carbon stocks in the
382
aboveground biomass of the trees ranged from 0.02 to 26.64 Mg C ha-1 depending on the tree
383
density and age (Table 5). The aboveground tree C stock was the highest at the silvopastoral
384
site, reaching about 37 Mg C ha-1. The estimated C stocks in the tree belowground biomass
385
ranged from 0.01 to 6.61 Mg C ha-1 at the silvoarable sites and was more than 9 Mg C ha-1 at
386
the TH silvopastoral site (Table 5).
387
388
389
23
Table 4 Soil organic carbon stock (Mg C ha-1) and SOC stock accumulation rate (Mg C ha-1 yr-1).
390
Site
Cumulativ
e ESM
(Mg ha-1)
Approximate
soil depth
(cm)
Cumulative SOC stock (Mg C ha-1)

(Mg C ha-1)
SOC stock accumulation rate (Mg C ha-1 yr-1)
Tree row
Inter row
AF
Control
AF Control
AF/Control
Tree row/Control
Inter-row/Control
CH
1000
0-10
19.4 ± 1.0
16.4 ± 0.3
16.7 ± 0.3
14.9 ± 0.4
1.8 ± 0.5*
0.30 ± 0.08*
0.76 ± 0.18*
0.26 ± 0.08*
2100
0-20
34.8 ± 1.2
32.5 ± 0.5
32.7 ± 0.5
31.0 ± 0.9
1.7 ± 1.0*
0.28 ± 0.17*
0.63 ± 0.25*
0.25 ± 0.17*
3250
0-30
48.4 ± 1.7
46.6 ± 1.0
46.7 ± 1.0
45.0 ± 1.1
1.7 ± 1.4*
0.29 ± 0.24*
0.57 ± 0.33*
0.27 ± 0.25*
1000
0-10
21.2 ± 0.6
13.0 ± 0.2
13.6 ± 0.2
12.2 ± 0.3
1.4 ± 0.4*
0.24 ± 0.07*
1.50 ± 0.11*
0.14 ± 0.07*
ME
2200
0-20
37.2 ± 0.6
27.7 ± 0.5
28.4 ± 0.5
26.4 ± 0.9
2.0 ± 1.1*
0.33 ± 0.18*
1.79 ± 0.19*
0.22 ± 0.18*
3500
0-30
51.1 ± 0.8
39.9 ± 0.9
40.7 ± 0.9
40.1 ± 1.7
0.5 ± 2.0
0.09 ± 0.33
1.83 ± 0.32*
-0.04 ± 0.33
SJ
700
0-10
40.6 ± 1.1
34.6 ± 0.9
35.5 ± 0.8
23.0 ± 0.2
12.4 ± 0.8*
0.30 ± 0.02*
0.43 ± 0.03*
0.28 ± 0.02*
1450
0-20
67.7 ± 1.1
59.8 ± 1.0
60.9 ± 0.9
42.1 ± 0.8
18.8 ± 1.2*
0.46 ± 0.03*
0.62 ± 0.03*
0.43 ± 0.03*
VE
900
0-10
15.5 ± 0.5
14.6 ± 0.4
14.8 ± 0.3
13.5 ± 1.0
1.3 ± 1.0*
0.07 ± 0.06*
0.11 ± 0.06*
0.06 ± 0.06*
2000
0-20
31.2 ± 0.8
29.5 ± 0.8
29.8 ± 0.6
27.9 ± 1.5
1.9 ± 1.6*
0.11 ± 0.09*
0.18 ± 0.09*
0.09 ± 0.09*
3150
0-30
44.7 ± 1.0
42.4 ± 0.9
42.8 ± 0.8
40.8 ± 2.0
2.0 ± 2.2
0.11 ± 0.12
0.21 ± 0.12*
0.09 ± 0.12
4400
0-40
58.1 ± 1.2
55.1 ± 1.2
55.7 ± 1.0
51.8 ± 2.5
3.9 ± 2.7*
0.22 ± 0.15*
0.35 ± 0.16*
0.19 ± 0.16*
5700
0-50
72.0 ± 1.5
66.8 ± 1.3
67.7 ± 1.1
61.2 ± 3.2
6.5 ± 3.4*
0.36 ± 0.19*
0.60 ± 0.20*
0.31 ± 0.19*
7050
0-60
85.3 ± 1.9
77.1 ± 1.6
78.6 ± 1.4
68.6 ± 4.1
10.0 ± 4.3*
0.56 ± 0.24*
0.93 ± 0.25*
0.48 ± 0.25*
RE
1000
0-10
21.6 ± 1.0
9.8 ± 0.4
11.7 ± 0.3
9.3 ± 0.1
2.3 ± 0.4*
0.13 ± 0.02*
0.68 ± 0.05*
0.02 ± 0.02*
4000
0-30
52.8 ± 1.4
37.9 ± 0.6
40.3 ± 0.5
35.8 ± 0.2
4.5 ± 0.6*
0.25 ± 0.03*
0.95 ± 0.08*
0.12 ± 0.03*
7300
0-50
77.1 ± 1.5
62.0 ± 0.7
64.4 ± 0.6
59.4 ± 0.2
5.0 ± 0.6*
0.28 ± 0.04*
0.98 ± 0.08*
0.14 ± 0.04*
10700
0-70
98.1 ± 1.5
82.4 ± 0.7
84.9 ± 0.6
79.7 ± 0.3
5.1 ± 0.7*
0.29 ± 0.04*
1.02 ± 0.08*
0.15 ± 0.04*
15700
0-100
130.4 ± 1.5
113.7 ± 0.7
116.4 ± 0.7
110.1 ± 0.3
6.3 ± 0.7*
0.35 ± 0.04*
1.13 ± 0.09*
0.20 ± 0.05*
TH
700
0-10
-
-
44.2 ± 3.4
47.1 ± 1.6
-2.9 ± 3.8
-0.11 ± 0.14
-
-
1450
0-20
-
-
80.4 ± 5.0
84.1 ± 1.9
-3.7 ± 5.3
-0.14 ± 0.20
-
-
2200
0-30
-
-
110.2 ± 6.1
114.3 ± 2.3
-4.1 ± 6.5
-0.16 ± 0.25
-
-
3000
0-40
-
-
137.6 ± 6.5
138.2 ± 2.3
-0.5 ± 6.9
-0.02 ± 0.26
-
-
3800
0-50
-
-
169.3 ± 6.5
156.5 ± 2.7
12.8 ± 7.0*
0.49 ± 0.27*
-
-
24
Associated errors are standard errors. Approximate depths are presented here to give a better understanding of the ESM for a given site but do not
391
correspond to the precise mass of the profile, which may vary between tree rows, inter-rows and the control (Ellert and Bettany, 1995). At the TH
392
silvopastoral site, no distinction was made between tree rows and inter-rows (uniform cover). Significantly different (p-value < 0.05) delta SOC
393
stock ( and additional SOC storage rate are followed by *. ESM: Equivalent Soil Mass, AF: Agroforestry. CH: Châteaudun, ME: Melle,
394
SJ: Saint-Jean-d’Angély, VE: Vézénobres, RE: Restinclières, TH: Theix.
395
396
397
398
25
399
Figure 4. Soil organic carbon stock (kg C m-3) at the different sites. Bars represent standard errors. Approximate depths are presented but refer to
400
equivalent soil mass. At the TH silvopastoral site, no distinction was made between tree rows and inter-rows (uniform cover), values are
401
for the whole agroforestry plot.
402
26
403
Figure 5. Total organic carbon stock (Mg C ha-1) of the different sites. AF: agroforestry, C: agricultural control. SOC: Soil organic carbon, ABG:
404
Aboveground, BLG: Belowground. CH: Châteaudun, ME: Melle, SJ: Saint-Jean-d’Angély, VE: Vézénobres, RE: Restinclières, TH:
405
Theix. Studied depths vary between sites: 30 cm for CH, 30 cm for ME, 20 cm for SJ, 60 cm for VE, 100 cm for RE and 50 cm for TH.
406
Different lowercase letters indicate a significant (p-value < 0.05) difference of SOC stock between AF and C plots per site, and different
407
uppercase letters indicate a significant difference (p-value < 0.05) in the total organic carbon stock between AF and C plots per site.
408
27
3.5 Total carbon stock of the different systems
409
At the silvoarable sites, the total C stock (SOC + biomass) ranged from about 50 Mg C ha-1 to
410
125 Mg C ha-1, and reached 220 Mg C ha-1 at the TH silvopastoral site (Fig. 5). The total C
411
stock was always higher in the agroforestry systems than in the control plots. In the young
412
plantations (CH and ME), the total C stock was mainly SOC, with tree C stock accounting for
413
less than 0.01% of the total C stock. At oldest sites, up to 75% of the difference between total
414
C stock in the agroforestry systems and control plots was explained by the tree biomass (Fig.
415
5).
416
417
3.6 Organic carbon accumulation rate in soil and tree biomass
418
The mean SOC stock accumulation rate in the top 30 cm in the silvoarable systems was 0.18
419
Mg C ha-1 yr-1 (0.09 to 0.29 Mg C ha-1 yr-1). This rate reached 0.24 Mg C ha-1 yr-1 when the SJ
420
silvoarable site and its shallow soil (20 cm) was taken into account. At the RE site, the SOC
421
stock accumulation rate was 0.25 Mg C ha-1 yr-1 in the top 30 cm, and 0.35 Mg C ha-1 yr-1 in the
422
top 100 cm, with a SOC stock accumulation rate of about 0.1 Mg C ha-1 yr-1 in the 30-100 cm
423
layer (Table 4). Tree rows contributed about 20% to 50% to the SOC stock accumulation rate
424
although they covered only 7% to 18% of the agroforestry surface area.
425
The C accumulation rate in the tree biomass in CH and ME young plantations was negligible
426
(0.004 and 0.02 Mg C ha-1 yr-1, respectively) (Table 5). In the older and denser silvoarable
427
sites, this rate ranged from 0.62 to 1.85 Mg C ha-1 yr-1, and was 1.76 Mg C ha-1 yr-1 at the TH
428
silvopastoral site (Table 5).
429
430
431
28
Table 5 Tree characteristics, aboveground and belowground carbon stocks at the various sites.
432
Site
Age
(yr)
DBH (cm)
Height of
merchantable
timber (m)
Total height
(m)
C stock of
merchantable
timber (kg C tree-1)
ABG tree C
stock
(kg C tree-1)
ABG tree C
stock
(Mg C ha-1)
Estimated BEG
tree C stock
(Mg C ha-1)
Estimated total tree C stock
accumulation rate
(Mg C ha-1 yr-1)
CH
6
2.6 ± 0.2
1.45 ± 0.04
2.12 ± 0.11
0.44 ± 0.06
0.49 ± 0.07
0.017 ± 0.002
0.01 (0.01-0.01)
0.004 ± 0.0004
ME
6
5.5 ± 0.3
1.13 ± 0.03
3.18 ± 0.13
1.18 ± 0.12
2.07 ± 0.19
0.073 ± 0.007
0.03 (0.03-0.04)
0.02 ± 0.001
SJ
41
29.9 ± 1.3
3.11 ± 0.23
13.18 ± 0.10
41.44 ± 2.36
194.56 ± 14.94
19.85 ± 1.52
5.55 (3.28-9.38)
0.62 ± 0.10
VE
18
31.7 ± 1.5
4.17 ± 0.18
15.52 ± 0.36
56.85 ± 3.77
266.44 ± 19.90
26.64 ± 1.99
6.61 (4.00-10.95)
1.85 ± 0.27
RE
18
25.5 ± 1.4
4.49 ± 0.39
11.21 ± 0.65
46.23 ± 2.47
98.93 ± 7.80
10.88 ± 0.86
2.99 (1.89-4.72)
0.77 ± 0.11
TH
26
30.7 ± 1.4
4.10 ± 0.23
14.70 ± 0.32
53.80 ± 1.76
183.46 ± 2.66
36.69 ± 0.53
9.13 (5.34-15.63)
1.76 ± 0.25
Errors represent standard errors. Number of measured trees: CH=24, ME=20, SJ=10, VE=10, RE=9 except for biomass measurements where n=3,
433
and TH=10. Values in brackets represent the 95% prediction interval for estimating the belowground biomass (Cairns et al., 1997). ABG:
434
Aboveground, BEG: Belowground. CH: Châteaudun, ME: Melle, SJ: Saint-Jean-d’Angély, VE: Vézénobres, RE: Restinclières, TH: Theix.
435
436
29
437
Figure 6. SOC stock accumulation rates as a function of plantation age. Values are for the
438
approximate top 30 cm, except for the SJ site (approximate top 20 cm, maximum soil
439
depth).
440
30
4. Discussion
441
4.1 Spatial variation of SOC stock in silvoarable systems
442
The sampling protocol was designed to take account of the spatial distribution of SOC stocks
443
as a function of distance from the trees. Sampling in the inter-rows in front of a tree or between
444
two trees did not affect the estimation of SOC stocks. The protocol could, therefore, be
445
simplified for instance by sampling only in front of a tree or by sampling along the diagonal of
446
the sampling pattern, which was equivalent to a quarter of the Voronoi polygon (Levillain et
447
al., 2011). Field sampling would then be less costly and less time-consuming.
448
The distance from the trees had no effect on SOC stocks in the inter-rows, except at the oldest
449
SJ site. At this 41-year-old site, the width of the cropped alley had been reduced over the past
450
10 years owing to light competition, which might explain the gradient of SOC stocks observed.
451
At the RE site, Cardinael et al., (2015b) suggested that close to the trees, organic C input coming
452
from tree fine root senescence (Cardinael et al., 2015a; Germon et al., 2016), exudates and
453
leaves might be compensated by a decrease in organic C input from crop residues owing to
454
lower yields (Dufour et al., 2013). The same hypothesis might apply at the VE site, where no
455
SOC stock gradient was found in the inter-rows (same tree density, same tree species and tree
456
age as the RE site). Consequently, fewer soil samples could be taken to estimate SOC stock in
457
the inter-rows. However, these two 18-year-old sites had a high tree density, the distance
458
between two tree rows (11 m and 13 m) being almost the same as the mean tree height (15 m
459
and 11 m). It is possible that a SOC stock gradient may appear with time in the inter-rows in
460
low-density plantations with a large distance between two tree rows (> 30 m). This gradient
461
effect could also depend on the tree species. This hypothesis could be tested in the future at the
462
CH and ME sites.
463
31
At all the silvoarable sites, the SOC stock was higher in the tree rows than in the inter-rows and
464
in the control plot, especially in the topsoil layer (0-10 cm). Tree rows therefore had a
465
considerable effect on SOC storage, contributing up to 50% of the additional SOC storage at
466
silvoarable plot scale for only a small surface area. There were two main sources of organic
467
matter returned to the soil in the tree rows: carbon from the trees (litter, fine roots and exudates)
468
and carbon from the herbaceous vegetation. At the RE site, the aboveground and belowground
469
biomass of the herbaceous vegetation in the tree rows was 2.13 Mg C ha-1 and 0.74 Mg C ha-1,
470
respectively (unpublished data). The C input to the soil from this vegetation in the tree rows
471
could, therefore, be up to 2.9 Mg C ha-1 yr-1. The spaces between the trees along the tree rows
472
could be considered comparable to grass strips or natural grassland because of the herbaceous
473
cover and the lack of soil tillage. Converting annual crop cultivation to grassland was shown to
474
be very efficient in terms of SOC storage by Conant et al., (2001), Arrouays et al., (2002), and
475
Soussana et al., (2004) with SOC stock accumulation rates ranging from 0.49 Mg C ha-1 yr-1 to
476
1.01 Mg C ha-1 yr-1 in the top 30 cm. Based on their results and on the high SOC stocks also
477
measured in the topsoil in tree rows of young plantations with small tree biomass, we suggest
478
that a major part of the SOC storage in the tree rows is due to the herbaceous vegetation. There
479
was no clear difference between sown and natural herbaceous vegetation in the tree rows,
480
although the highest SOC stock accumulation rate was obtained for sown grass (ME site, 1.3
481
Mg C ha-1 yr-1). However, the management of these tree rows seems to be a key factor for
482
increasing the SOC storage capacity of silvoarable systems. Several studies showed that
483
including legumes in the composition of grasslands increased herbage productivity (Tilman et
484
al., 2001; Marquard et al., 2009; Prieto et al., 2015) and SOC storage (Steinbeiss et al., 2008;
485
Lange et al., 2015).
486
487
488
32
4.2 SOC stock accumulation rates in silvoarable systems
489
In the five silvoarable systems studied, the mean SOC stock accumulation rate in the top 30 cm
490
was 0.24 (0.09-0.46) Mg C ha-1 yr-1. This estimate for silvoarable plots with an average age of
491
17.8 -yr, is slightly lower than previously suggested for 20-yr-old agroforestry systems in
492
France (0.30 (0.03-0.41) Mg C ha-1 yr-1) by Pellerin et al. (2013) based on a literature review
493
but it is of the same order of magnitude. The SOC stock accumulation rate was also slightly
494
lower than those reported by Oelbermann et al. (2006) for a 13-yr-old Canadian alley cropping
495
system combining hybrid poplars and wheat, soybean and maize grown in rotation (0.30 Mg C
496
ha-1 yr-1 in the top 20 cm and 0.39 Mg C ha-1 yr-1 in the top 40 cm). As well as, Peichl et al.
497
(2006) reported a SOC stock accumulation rate of 1.04 Mg C ha-1 yr-1 in the top 20 cm for a
498
13-yr-old hybrid poplar and Norway spruce-barley agroforestry system. Overall, our estimated
499
SOC stock accumulation rate is slightly lower than most published results (Lorenz and Lal,
500
2014; Kim et al., 2016). However, as reported by Cardinael et al. (2015b), our study estimated
501
SOC storage in silvoarable systems using the equivalent soil mass, which gives more accurate
502
results when soil bulk density is modified by changes in land use (Ellert and Bettany, 1995;
503
Ellert et al., 2002), as was the case in these systems, especially in the tree rows. Furthermore,
504
most fields in our study were owned and managed by farmers. Although this fact may generate
505
some uncertainties, it has the advantage of taking account of a broad variety of practices that
506
are commonly used by farmers.
507
At the two 18-year-old silvoarable sites (RE and VE) there was a significant increase in deep
508
SOS stocks (below 30 cm). At the VE site this might be partially due to a slightly higher sand
509
content in the control plot than in the agroforestry plot below 30 cm. At the RE site, this increase
510
might result from a high density of deep tree fine roots (Mulia and Dupraz, 2006; Cardinael et
511
al., 2015a). Although the SOC stock accumulation rate was lower than in topsoil layers, deep
512
soil layers might then be able to store a large amount of SOC over a longer period owing to
513
33
better SOC stabilization conditions (Rasse et al., 2005). However, little is known about the
514
effect of fresh organic matter input on deep soil layers and some authors found that this might
515
stimulate the mineralization of old organic matter (Fontaine et al., 2004, 2007).
516
There was no change in the SOC stock accumulation rates with time in the silvoarable systems
517
(Fig. 6) but very old sites (> 40 year old) were under-represented in this study. It is therefore
518
difficult to assess the possible effect of tree age on the SOC accumulation rate. Tree growth
519
increases organic litter production with time but competition with the intercrop also increases,
520
potentially causing a decrease in crop yields such as cereals (Dufour et al., 2013). In a recent
521
meta-analysis, Kim et al., (2016) found a slight decrease in the SOC stock accumulation rates
522
in very old agroforestry systems, which was attributed to the soil reaching a new SOC stock
523
equilibrium. Based on technical limits (soil depth, water holding capacity, field size), Pellerin
524
et al., (2013) and Chenu et al., (2014) estimated that about 4 M ha of arable land could be
525
converted to silvoarable systems in France. Given the estimated SOC stock accumulation rate
526
in this study, this would mean that 3.6 105 to 1.84 106 Mg C could be stored annually in the soil.
527
528
4.3 Carbon storage in silvopastoral systems
529
The silvopastoral system set up on an andosol on permanent grassland (Tables 3 and 4) had no
530
more additional SOC in the top 30 cm than grassland without trees. This site had been under
531
pasture for decades before tree planting. It had a high SOC concentration (about 65 mg C g-1 at
532
0-10 cm) and the soil was possibly at a steady state so that it could not store additional SOC, at
533
least in fine soil fractions (Hassink, 1997). On a Patagonian andosol, Dube et al., (2012) also
534
found that there was no significant difference in the SOC stocks in the top 40 cm of a
535
silvopastoral system compared to a natural pasture. At our site, there was a significant effect of
536
the silvopastoral system on SOC concentration and stock in the 30-50 cm layer: the SOC
537
34
concentration in the silvopastoral system was about 29 mg C g-1 while in the grassland control
538
it was only about 23 mg C g-1. It is possible that these deep soil layers in grasslands might be
539
less SOC-saturated than topsoil layers and that roots from agroforestry trees could, therefore,
540
contribute to additional SOC storage at depth. Haile et al. (2010) also found that trees affected
541
deep SOC storage in silvopastoral systems. The biomass production of pastures in silvopastoral
542
systems is usually less sensitive to shade than that of annual crops such as cereals grown in
543
silvoarable systems (Moreno et al., 2007a, b; Moreno, 2008), except for N2 fixing species
544
(Carranca et al., 2015). Furthermore, grass under the tree cover can have a longer growing
545
season (Puerto et al., 1990) and forage quality can be improved under tree canopies (Cubera et
546
al., 2009). Therefore, silvopastoral systems might support a higher tree density than silvoarable
547
systems (Benavides et al., 2009; Devkota et al., 2009), resulting in higher C stocks in the tree
548
biomass (> 35 Mg C ha-1 in this case).
549
550
4.4 Carbon storage in the tree biomass
551
The C stock in the tree biomass in the young plantations was negligible but, in the old
552
plantations, C storage was greater in the tree biomass than in the soil (Fig. 5). The C
553
accumulation rate in the tree biomass was higher in the old plantations than in young
554
plantations. This is explained by the much higher total leaf area of old trees compared to very
555
young trees and, therefore, by a higher photosynthesis capacity (Stephenson et al., 2014).
556
However, estimates of the tree root biomass may be underestimated by the forest allometrics
557
used. The architecture of agroforestry trees is different from forest trees owing to a lower
558
intraspecific competition and to pruning. Moreover, agroforestry trees have been shown to be
559
very deep rooted owing to soil tillage and to competition with intercrops (Mulia and Dupraz,
560
2006; Cardinael et al., 2015a).
561
35
Carbon stock in the tree biomass is not usually considered as a long-term C sink in the same
562
way as the SOC stock but the residence time of C in the harvested biomass depends on the fate
563
of wood products and can be as long as many decades for timber wood (Profft et al., 2009;
564
Bauhus et al., 2010), which was the case for the trees grown at the sites studied. Branches could
565
be used as a substitute for fossil fuel to produce energy (Kürsten, 2000; Cardinael et al., 2012)
566
or be returned to the soil as ramial chipped wood amendments (Barthès et al., 2010).
567
568
5. Conclusion
569
This study showed the potential of agroforestry systems to increase carbon stock in both the
570
soil and tree biomass under different pedo-climatic conditions in France. The sampling protocol
571
evaluated the spatial distribution of SOC stock and the results showed that it could be simplified
572
for future studies. SOC stocks accumulated mainly in the tree rows and mainly in the top 30 cm
573
of soil, but at deeper soil layers in two silvoarable sites, as well. Further studies are required to
574
gain a better assessment of the effect of agroforestry on deep SOC stock. Allometric equations
575
should be developed for trees grown in temperate agroforestry systems to reduce the uncertainty
576
of tree root biomass estimates. Very old sites (> 40 years old) were under-represented in our
577
dataset and long-term experimental agroforestry sites are required to assess the effect of trees
578
on soil carbon over long periods.
579
580
Acknowledgments
581
This study was financed by the French Environment and Energy Management Agency
582
(ADEME), following a call for proposals as part of the REACCTIF program (Research on
583
Climate Change Mitigation in Agriculture and Forestry). This study was part of the funded
584
36
project AGRIPSOL (Agroforestry for Soil Protection), coordinated by Agroof. Rémi Cardinael
585
was also funded by La Fondation de France. Two anonymous reviewers provided many
586
excellent comments that improved the quality of this manuscript. We are very grateful to the
587
farmers who allowed us to take samples in their fields and to Eric Villeneuve (INRA) for his
588
help at the Theix site. We should also like to thank Daniel Billou (UPMC), Manon Villeneuve
589
(IRD), Patricia Mahafaka and Clément Renoir for their help in the field and in the laboratory.
590
591
592
References
593
Albrecht, A., Kandji, S.T., 2003. Carbon sequestration in tropical agroforestry systems. Agric.
594
Ecosyst. Environ. 99, 1527.
595
Amundson, R., 2001. The carbon budget in soils. Annu. Rev. Earth Planet. Sci. 29, 535562.
596
Angers, D.A., Arrouays, D., Saby, N.P.A., Walter, C., 2011. Estimating and mapping the
597
carbon saturation deficit of French agricultural topsoils. Soil Use Manag. 27, 448452.
598
Arrouays, D., Balesdent, J., Jayet, P.A., Soussana, J.-F., Stengel, P., 2002. Contribution à la
599
lutte contre l’effet de serre. Stocker du carbone dans les sols agricoles de France? Expertise
600
Scientifique Collective, Synthèse du rapport. INRA, Paris.
601
Arrouays, D., Deslais, W., Badeau, V., 2001. The carbon content of topsoil and its geographical
602
distribution in France. Soil Use Manag. 17, 711.
603
Bambrick, A.D., Whalen, J.K., Bradley, R.L., Cogliastro, A., Gordon, A.M., Olivier, A.,
604
Thevathasan, N. V, 2010. Spatial heterogeneity of soil organic carbon in tree-based
605
intercropping systems in Quebec and Ontario, Canada. Agrofor. Syst. 79, 343353.
606
Barthès, B.G., Manlay, R.J., Porte, O., 2010. Effets de l’apport de bois raméal sur la plante et
607
le sol : une revue des résultats expérimentaux. Cah. Agric. 19, 280287.
608
37
Bauhus, J., van der Meer, P., Kanninen, M., 2010. Ecosystem Goods and Services from
609
Plantation Forests. Earthscan, London, UK.
610
Benavides, R., Douglas, G.B., Osoro, K., 2009. Silvopastoralism in New Zealand: Review of
611
effects of evergreen and deciduous trees on pasture dynamics. Agrofor. Syst. 76, 327350.
612
Cairns, M.A., Brown, S., Helmer, E.H., Baumgardner, G.A., 1997. Root biomass allocation in
613
the World’s upland forests. Oecologia 111, 1–11.
614
Cardinael, R., Mao, Z., Prieto, I., Stokes, A., Dupraz, C., Kim, J.H., Jourdan, C., 2015a.
615
Competition with winter crops induces deeper rooting of walnut trees in a Mediterranean
616
alley cropping agroforestry system. Plant Soil 391, 219235.
617
Cardinael, R., Chevallier, T., Barthès, B.G., Saby, N.P.., Parent, T., Dupraz, C., Bernoux, M.,
618
Chenu, C., 2015b. Impact of alley cropping agroforestry on stocks, forms and spatial
619
distribution of soil organic carbon - A case study in a Mediterranean context. Geoderma
620
259260, 288299.
621
Cardinael, R., Thevathasan, N., Gordon, A., Clinch, R., Mohammed, I., Sidders, D., 2012.
622
Growing woody biomass for bioenergy in a tree-based intercropping system in southern
623
Ontario, Canada. Agrofor. Syst. 86, 279286.
624
Carranca, C., Castro, I. V., Figueiredo, N., Redondo, R., Rodrigues, A.R.F., Saraiva, I.,
625
Maricato, R., Madeira, M.A. V, 2015. Influence of tree canopy on N2 fixation by pasture
626
legumes and soil rhizobial abundance in Mediterranean oak woodlands. Sci. Total
627
Environ. 506507, 8694.
628
Chave, J., Coomes, D., Jansen, S., Lewis, S.L., Swenson, N.G., Zanne, A.E., 2009. Towards a
629
worldwide wood economics spectrum. Ecol. Lett. 12, 35166.
630
Chenu, C., Klumpp, K., Bispo, A., Angers, D., Colnenne, C., Metay, A., 2014. Stocker du
631
carbone dans les sols agricoles: évaluation de leviers d’action pour la France. Innov.
632
38
Agron. 37, 2337.
633
Conant, R.T., Paustian, K., Elliott, E.T., 2001. Grassland management and conversion into
634
grassland: effects on soil carbon. Ecol. Appl. 11, 343355.
635
Constantin, J., Mary, B., Laurent, F., Aubrion, G., Fontaine, A., Kerveillant, P., Beaudoin, N.,
636
2010. Effects of catch crops, no till and reduced nitrogen fertilization on nitrogen leaching
637
and balance in three long-term experiments. Agric. Ecosyst. Environ. 135, 268278.
638
Cubera, E., Nunes, J.M., Madeira, M., Gazarini, L., 2009. Influence of Quercus ilex trees on
639
herbaceous production and nutrient concentrations in southern Portugal. J. Plant Nutr. Soil
640
Sci. 172, 565571.
641
Devkota, N.R., Kemp, P.D., Hodgson, J., Valentine, I., Jaya, I.K.D., 2009. Relationship
642
between tree canopy height and the production of pasture species in a silvopastoral system
643
based on alder trees. Agrofor. Syst. 76, 363374.
644
Dimassi, B., Cohan, J.-P., Labreuche, J., Mary, B., 2013. Changes in soil carbon and nitrogen
645
following tillage conversion in a long-term experiment in Northern France. Agric. Ecosyst.
646
Environ. 169, 1220.
647
Dube, F., Espinosa, M., Stolpe, N.B., Zagal, E., Thevathasan, N. V., Gordon, A.M., 2012.
648
Productivity and carbon storage in silvopastoral systems with Pinus ponderosa and
649
Trifolium spp., plantations and pasture on an Andisol in Patagonia, Chile. Agrofor. Syst.
650
86, 113128.
651
Dufour, L., Metay, A., Talbot, G., Dupraz, C., 2013. Assessing Light Competition for Cereal
652
Production in Temperate Agroforestry Systems using Experimentation and Crop
653
Modelling. J. Agron. Crop Sci.
654
Ellert, B.H., Bettany, J.R., 1995. Calculation of organic matter and nutrients stored in soils
655
under contrasting management regimes. Can. J. Soil Sci. 75, 529538.
656
39
Ellert, B.H., Janzen, H.H., Entz, T., 2002. Assessment of a method to measure temporal change
657
in soil carbon storage. Soil Sci. Soc. Am. J. 66, 16871695.
658
Fontaine, S., Bardoux, G., Abbadie, L., Mariotti, A., 2004. Carbon input to soil may decrease
659
soil carbon content. Ecol. Lett. 7, 314320.
660
Fontaine, S., Barot, S., Barré, P., Bdioui, N., Mary, B., Rumpel, C., 2007. Stability of organic
661
carbon in deep soil layers controlled by fresh carbon supply. Nature 450, 277281.
662
Freibauer, A., Rounsevell, M.D.A., Smith, P., Verhagen, J., 2004. Carbon sequestration in the
663
agricultural soils of Europe. Geoderma 122, 123.
664
Germon, A., Cardinael, R., Prieto, I., Mao, Z., Kim, J.H., Stokes, A., Dupraz, C., Laclau, J.-P.,
665
Jourdan, C., 2016. Unexpected phenology and lifespan of shallow and deep fine roots of
666
walnut trees grown in a silvoarable Mediterranean agroforestry system. Plant Soil 401,
667
409426.
668
Haile, S.G., Nair, V.D., Nair, P.K.R., 2010. Contribution of trees to carbon storage in soils of
669
silvopastoral systems in Florida, USA. Glob. Chang. Biol. 16, 427438.
670
Harris, D., Horwath, W.R., Van Kessel, C., 2001. Acid fumigation of soils to remove carbonates
671
prior to total organic carbon or carbon-13 isotopic analysis. Soil Sci. Soc. Am. J. 65, 1853
672
1856.
673
Hassink, J., 1997. The capacity of soils to preserve organic C and N by their association with
674
clay and silt particles. Plant Soil 191, 7787.
675
Houghton, R.A., 2007. Balancing the global carbon budget. Annu. Rev. Earth Planet. Sci. 35,
676
313347.
677
Howlett, D.S., Moreno, G., Mosquera Losada, M.R., Nair, P.K.R., Nair, V.D., 2011. Soil carbon
678
storage as influenced by tree cover in the Dehesa cork oak silvopasture of central-western
679
Spain. J. Environ. Monit. 13, 1897904.
680
40
IUSS Working Group WRB, 2007. World Reference Base for Soil Resources 2006, first update
681
2007. World Soil Resources Reports No. 103. FAO, Rome.
682
Janzen, H.H., 2006. The soil carbon dilemma: Shall we hoard it or use it? Soil Biol. Biochem.
683
38, 419424.
684
Kim, D.-G., Kirschbaum, M.U.F., Beedy, T.L., 2016. Carbon sequestration and net emissions
685
of CH4 and N2O under agroforestry: Synthesizing available data and suggestions for
686
future studies. Agric. Ecosyst. Environ. 226, 6578.
687
Kürsten, E., 2000. Fuelwood production in agroforestry systems for sustainable land use and
688
CO2-mitigation. Ecol. Eng. 16, S69S72.
689
Lal, R., 2004a. Soil carbon sequestration to mitigate climate change. Geoderma 123, 122.
690
Lal, R., 2004b. Soil carbon sequestration impacts on global climate change and food security.
691
Science 304, 16231627.
692
Lamlom, S.H., Savidge, R.H., 2003. A reassessment of carbon content in wood: variation within
693
and between 41 North American species. Biomass and Bioenergy 25, 381388.
694
Lange, M., Eisenhauer, N., Sierra, C.A., Bessler, H., Engels, C., Griffiths, R.I., Mellado-
695
Vázquez, P.G., Malik, A.A., Roy, J., Scheu, S., Steinbeiss, S., Thomson, B.C., Trumbore,
696
S.E., Gleixner, G., 2015. Plant diversity increases soil microbial activity and soil carbon
697
storage. Nat. Commun. 6, 6707.
698
Le Quéré, C., Peters, G.P., Andres, R.J., Andrew, R.M., Boden, T., Ciais, P., Friedlingstein, P.,
699
Houghton, R.A., Marland, G., Moriarty, R., Sitch, S., Tans, P., Arneth, A., Arvanitis, A.,
700
Bakker, D.C.E., Bopp, L., Canadell, J.G., Chini, L.P., Doney, S.C., Harper, A., Harris, I.,
701
House, J.I., Jain, A.K., Jones, S.D., Kato, E., Keeling, R.F., Klein Goldewijk, K.,
702
Körtzinger, A., Koven, C., Lefèvre, N., Omar, A., Ono, T., Park, G.-H., Pfeil, B., Poulter,
703
B., Raupach, M.R., Regnier, P., Rödenbeck, C., Saito, S., Schwinger, J., Segschneider, J.,
704
41
Stocker, B.D., Tilbrook, B., van Heuven, S., Viovy, N., Wanninkhof, R., Wiltshire, A.,
705
Zaehle, S., Yue, C., 2014. Global carbon budget 2014. Earth Syst. Sci. Data Discuss. 7,
706
521610.
707
Levillain, J., Thongo M’Bou, A., Deleporte, P., Saint-André, L., Jourdan, C., 2011. Is the simple
708
auger coring method reliable for below-ground standing biomass estimation in Eucalyptus
709
forest plantations? Ann. Bot. 108, 221230.
710
Lorenz, K., Lal, R., 2014. Soil organic carbon sequestration in agroforestry systems. A review.
711
Agron. Sustain. Dev. 34, 443454.
712
Luo, Z., Wang, E., Sun, O.J., 2010. Can no-tillage stimulate carbon sequestration in agricultural
713
soils? A meta-analysis of paired experiments. Agric. Ecosyst. Environ. 139, 224231.
714
Marquard, E., Weigelt, A., Temperton, V.M., Roscher, C., Schumacher, J., Buchmann, N.,
715
Fischer, M., Weisser, W.W., Schmid, B., 2009. Plant species richness and functional
716
composition drive overyielding in a six-year grassland experiment. Ecology 90, 3290
717
3302.
718
Martin, M.P., Wattenbach, M., Smith, P., Meersmans, J., Jolivet, C., Boulonne, L., Arrouays,
719
D., 2011. Spatial distribution of soil organic carbon stocks in France. Biogeosciences 8,
720
10531065.
721
Moreno, G., 2008. Response of understorey forage to multiple tree effects in Iberian dehesas.
722
Agric. Ecosyst. Environ. 123, 239244.
723
Moreno, G., Obrador, J.J., García, A., 2007a. Impact of evergreen oaks on soil fertility and crop
724
production in intercropped dehesas. Agric. Ecosyst. Environ. 119, 270280.
725
Moreno, G., Obrador, J.J., García, E., Cubera, E., Montero, M.J., Pulido, F., Dupraz, C., 2007b.
726
Driving competitive and facilitative interactions in oak dehesas through management
727
practices. Agrofor. Syst. 70, 2540.
728
42
Mosquera Losada, M.R., Freese, D., Rigueiro-Rodríguez, A., 2011. Carbon Sequestration in
729
European Agroforestry Systems, in: Kumar, B.M., Nair, P.K.R. (Eds.), Carbon
730
Sequestration Potential of Agroforestry Systems. Springer Netherlands, pp. 4359.
731
Mulia, R., Dupraz, C., 2006. Unusual fine root distributions of two deciduous tree species in
732
southern France: What consequences for modelling of tree root dynamics? Plant Soil 281,
733
7185.
734
Nair, P.K., 1993. An introduction to agroforestry. Kluwer, Dordrecht, The Netherlands.
735
Oelbermann, M., Voroney, R.P., Thevathasan, N. V., Gordon, A.M., Kass, D.C.L.,
736
Schlönvoigt, A.M., 2006. Soil carbon dynamics and residue stabilization in a Costa Rican
737
and southern Canadian alley cropping system. Agrofor. Syst. 68, 2736.
738
Paustian, K., Andrén, O., Janzen, H., Lal, R., Smith, P., Tian, G., Tiessen, H., Van Noordwijk,
739
M., Woomer, P., 1997. Agricultural soils as a sink to mitigate CO2 emissions. Soil Use
740
Manag. 13, 230244.
741
Peichl, M., Thevathasan, N. V, Gordon, A.M., Huss, J., Abohassan, R.A., 2006. Carbon
742
sequestration potentials in temperate tree-based intercropping systems, southern Ontario,
743
Canada. Agrofor. Syst. 66, 243257.
744
Pellerin, S., Bamière, L., Angers, D., Béline, F., Benoît, M., Butault, J.P., Chenu, C., Colnenne-
745
David, C., De Cara, S., Delame, N., Doreau, M., Dupraz, P., Faverdin, P., Garcia-Launay,
746
F., Hassouna, M., Hénault, C., Jeuffroy, M.., Klumpp, K., Metay, A., Moran, D., Recous,
747
S., Samson, E., Savini, I., Pardon, L., 2013. How can French agriculture contribute to
748
reducing greenhouse gas emissions? Abatement potential and cost of ten technical
749
measures. Synopsis of the study report, INRA, Paris (France).
750
Picard, N., Saint-André, L., Henry, M., 2012. Manual for building tree volume and biomass
751
allometric equations. From field measurement to prediction, CIRAD. ed. FAO.
752
43
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., R Development Core Team, 2013. nlme: Linear
753
and Nonlinear Mixed Effects Models. R package version 3.1-111.
754
Poeplau, C., Don, A., 2015. Carbon sequestration in agricultural soils via cultivation of cover
755
crops A meta-analysis. Agric. Ecosyst. Environ. 200, 3341.
756
Prieto, I., Violle, C., Barre, P., Durand, J., Ghesquiere, M., Litrico, I., 2015. Complementary
757
effects of species and genetic diversity on productivity and stability of sown grasslands.
758
Nat. Plants 15.
759
Profft, I., Mund, M., Weber, G.-E., Weller, E., Schulze, E.-D., 2009. Forest management and
760
carbon sequestration in wood products. Eur. J. For. Res. 128, 399413.
761
Puerto, A., Rico, M., Matias, M., Garcia, J., 1990. Variation in structure and diversity in
762
Mediterranean grasslands related to trophic status and grazing intensity. J. Veg. Sci. 1,
763
445452.
764
R Development Core Team, 2013. R: A language and environment for statistical computing.
765
Rasse, D.P., Rumpel, C., Dignac, M.F., 2005. Is soil carbon mostly root carbon? Mechanisms
766
for a specific stabilisation. Plant Soil 269, 341356.
767
Rhoades, C.C., 1997. Single-tree influences on soil properties in agroforestry: lessons from
768
natural forest and savanna ecosystems. Agrofor. Syst. 35, 7194.
769
Smith, P., 2004. Carbon sequestration in croplands: the potential in Europe and the global
770
context. Eur. J. Agron. 20, 229236.
771
Soussana, J.-F., Loiseau, P., Vuichard, N., Ceschia, E., Balesdent, J., Chevallier, T., Arrouays,
772
D., 2004. Carbon cycling and sequestration opportunities in temperate grasslands. Soil Use
773
Manag. 20, 219230.
774
Steinbeiss, S., Beßler, H., Engels, C., Temperton, V.M., Buchmann, N., Roscher, C.,
775
Kreutziger, Y., Baade, J., Habekost, M., Gleixner, G., 2008. Plant diversity positively
776
44
affects short-term soil carbon storage in experimental grasslands. Glob. Chang. Biol. 14,
777
29372949.
778
Stephenson, N.L., Das, A.J., Condit, R., Russo, S.E., Baker, P.J., Beckman, N.G., Coomes,
779
D.A., Lines, E.R., Morris, W.K., Rüger, N., Álvarez, E., Blundo, C., Bunyavejchewin, S.,
780
Chuyong, G., Davies, S.J., Duque, Á., Ewango, C.N., Flores, O., Franklin, J.F., Grau, H.R.,
781
Hao, Z., Harmon, M.E., Hubbell, S.P., Kenfack, D., Lin, Y., Makana, J.-R., Malizia, A.,
782
Malizia, L.R., Pabst, R.J., Pongpattananurak, N., Su, S.-H., Sun, I.-F., Tan, S., Thomas,
783
D., van Mantgem, P.J., Wang, X., Wiser, S.K., Zavala, M.A., 2014. Rate of tree carbon
784
accumulation increases continuously with tree size. Nature 507, 9093.
785
Tilman, D., Reich, P.B., Knops, J., Wedin, D., Mielke, T., Lehman, C., 2001. Diversity and
786
productivity in a long-term grassland experiment. Science 294, 843845.
787
Upson, M.A., Burgess, P.J., 2013. Soil organic carbon and root distribution in a temperate
788
arable agroforestry system. Plant Soil 373, 4358.
789
Virto, I., Barré, P., Burlot, A., Chenu, C., 2012. Carbon input differences as the main factor
790
explaining the variability in soil organic C storage in no-tilled compared to inversion tilled
791
agrosystems. Biogeochemistry 108, 1726.
792
Wotherspoon, A., Thevathasan, N. V., Gordon, A.M., Voroney, R.P., 2014. Carbon
793
sequestration potential of five tree species in a 25-year-old temperate tree-based
794
intercropping system in southern Ontario, Canada. Agrofor. Syst. 88, 631643.
795
Zanne, A.E., Lopez-Gonzalez, G., Coomes, D.A., Llic, J., Jansen, J., Lewis, S.L., Miller, R.B.,
796
Swenson, N.G., Wiemann, M.C., Chave, J., 2009. Data from: Towards a worldwide wood
797
economics spectrum. Dryad Digital Repository. http://dx.doi.org/10.5061/dryad.234.
798
799
800
45
Table S1 ANOVA on the linear mixed-effects (LME) model for SOC content, bulk density and SOC stock in the agroforestry plots as a function
801
of depth, location (inter-row or tree row), distance to the closest tree, and interactions between these.
802
Soil organic carbon content
Bulk density
Soil organic carbon stock
Site
F-value
Pr(>F)
F-value
Pr(>F)
F-value
Pr(>F)
CH
Depth
64.982
<0.0001
10.956
0.0001
22.341
<0.0001
Location
2.246
0.137
3.153
0.079
1.890
0.173
Distance
0.394
0.532
0.266
0.607
0.379
0.540
Depth×Location
8.078
0.0006
0.672
0.513
6.908
0.002
Depth×Distance
0.576
0.564
0.296
0.744
0.570
0.568
Location×Distance
0.227
0.635
0.226
0.636
0.472
0.494
ME
Depth
140.956
<0.0001
20.473
<0.0001
24.004
<0.0001
Location
130.363
<0.0001
78.246
<0.0001
116.989
<0.0001
Distance
0.012
0.911
7.257
0.008
0.016
0.900
Depth×Location
51.699
<0.0001
15.888
<0.0001
45.731
<0.0001
Depth×Distance
1.627
0.202
1.910
0.154
2.895
0.063
Location×Distance
0.004
0.949
0.162
0.688
0.144
0.705
SJ
Depth
370.623
<0.0001
7.285
0.0104
284.905
<0.0001
Location
35.543
<0.0001
0.356
0.554
33.719
<0.0001
Distance
15.183
0.0004
0.691
0.411
8.827
0.005
Depth×Location
6.719
0.014
6.305
0.017
9.250
0.004
Depth×Distance
4.101
0.0501
7.985
0.008
10.264
0.002
Location×Distance
0.987
0.327
1.534
0.223
0.728
0.399
VE
Depth
110.547
<0.0001
39.920
<0.0001
19.071
<0.0001
Location
24.017
<0.0001
5.956
0.016
23.272
<0.0001
Distance
0.001
0.980
0.674
0.413
0.083
0.773
Depth×Location
2.801
0.019
1.998
0.082
2.243
0.053
Depth×Distance
0.086
0.994
0.917
0.472
0.151
0.980
Location×Distance
0.278
0.599
0.095
0.758
0.075
0.785
RE
Depth
703.719
<0.0001
391.32
<0.0001
723.666
<0.0001
Location
223.367
<0.0001
23.90
<0.0001
66.935
<0.0001
Distance
2.229
0.1387
2.12
0.1491
2.353
0.1283
Depth×Location
272.736
<0.0001
10.04
<0.0001
68.377
<0.0001
Depth×Distance
2.338
0.0173
0.68
0.7137
1.775
0.0784
Location×Distance
4.425
0.0380
1.25
0.2666
3.285
0.0731
46
TH
Depth
89.206
<0.0001
2.739
0.033
59.624
<0.0001
Location
0.040
0.842
0.577
0.449
0.032
0.859
Distance
1.511
0.222
6.966
0.010
0.446
0.506
Depth×Location
0.673
0.612
0.817
0.517
0.622
0.648
Depth×Distance
0.225
0.924
0.750
0.560
0.341
0.850
Location×Distance
0.235
0.629
1.663
0.200
0.001
0.975
CH: Châteaudun, ME: Melle, SJ: Saint-Jean-d’Angély, VE: Vézénobres, RE: Restinclières, TH: Theix.
803
804
47
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
... In der Erforschung von Agroforstsystemen kommt dem Zustand der Böden und den damit zusammenhängenden Stoffhaushalten sowie der Boden-Biodiversität eine besondere Bedeutung zu. Insbesondere die Gehölzstreifen können mit ihrer mehrjährigen Vegetation und ausbleibenden Bodenbearbeitung den Boden vor Erosion und Degradation schützen und damit die Agrarlandschaft nachhaltig beeinflussen (Torralba et al. 2016;Basche und DeLonge 2019;Beule et al. 2022;Jacobs et al. 2022 1 Einleitung auf die gesamt Fläche (Böhm et al. 2014) und dem lokalen Düngungseffekt durch nährstoffreiche Laubstreu (Wachendorf et al. 2020), werden die stofflichen Auswirkungen von Gehölzen vor allem in Bezug auf den Kohlenstoffgehalt der Böden betont (Nair 2011;Zomer et al. 2016;Cardinael et al. 2017;Hübner et al. 2021;Mayer et al. 2022 ...
... Die räumliche Heterogenität der Gehölzstreifen wurde hingegen nicht ausreichend berücksichtigt und in den meisten Fällen der Gehölzeinfluss für das Gesamtsystem sogar unterrepräsentiert (Seitz et al. 2017;Beuschel et al. 2019;de Abreu et al. 2020). Cardinael et al. (2017) haben in ihrem Probennahmedesign die Heterogenität des Gehölzstreifens erstmals sehr detailliert erfasst und aus den Ergebnissen geschlussfolgert, dass das Probennahmedesign für weitere Studien vereinfacht werden kann. Ein entsprechender Ansatz wird in diesem Leitfaden mit dem Mischprobendesign vorgestellt, welches die Heterogenität des Gehölzstreifens und des angrenzenden Acker-bzw. ...
... Mit dem Fokus auf einem Langzeitmonitoring der Landnutzungsänderung zeigen wissenschaftliche Untersuchungen in diesem Bereich unterschied-bung der oberen 30 cm empfohlen (Wiesmeier et al. 2020). Die Unterteilung in 0-10 und 10-30 cm ist dabei zwingend erforderlich, da in den oberen Zentimetern, insbesondere in den ersten Jahren, die stärksten Veränderungen und Unterschiede zu erwarten sind (Cardinael et al. 2017 Wenn möglich, sollte außerdem die Untersuchung einer benachbarten Nicht-Agroforstfläche (Referenzfläche) mit gleichem Acker bzw. Grünlandmanagement zum gleichen Zeitpunkt erfolgen. ...
Method
Full-text available
Dieser Leitfaden richtet sich an alle Wissenschaftler*innen sowie Citizen-Science-Gruppen, die bodenkundliche Untersuchungen in Agroforstsystemen mit streifenförmiger Anordnung der Gehölzkulturen vornehmen möchten. Ein besonderer Fokus liegt auf dem Probennahmedesign als Fundament und Grundvoraussetzung für die Vergleichbarkeit von wissenschaftlichen Studien, sowie einer Auswahl an zu untersuchenden bodenkundlichen Parametern und geeigneten Methoden. Das erläuterte Forschungsdesign ist als Mindestanforderung für die Untersuchung eines Agroforstsystems als Gesamtsystem zu verstehen. Außerdem kann es als Basis für weiterführende Untersuchungen und spezifische Fragestellungen im Agroforstkontext dienen. Das übergeordnete Ziel dieses Leitfadens ist es, langfristig eine hohe Vergleichbarkeit von Daten aus verschiedenen Agroforstsystemen zu gewährleisten. Die aus der Zusammenführung von Daten unterschiedlicher Fallstudien gewonnenen Informationen sollen in der Zukunft Wissenschaftler*innen, Praktiker*innen und Entscheidungsträger*innen dabei unterstützen, objektive Aussagen sowie fundierte Entscheidungen im Management und der Förderung von Agroforstsystemen und -projekten zu treffen.
... Modern alley cropping systems, whereby annual crops are grown between tree lines or hedges, have shown to achieve higher system yields than sole cropping systems and can provide diverse ecosystem services, such as water regulation, soil protection and biodiversity (Don et al., 2018;Palma et al., 2007;Smith et al., 2012). In addition, trees and hedges contribute to climate change mitigation through enhanced carbon sequestration in aboveground and belowground biomass, as well as soil organic carbon (Cardinael et al., 2017;Golicz et al., 2021). ...
... Tree rows in short rotation coppice alley cropping systems ( Fig. 1d) are, for example, generally characterised by higher tree density than alley cropping systems for fruit and timber trees (Fig. 1e). Furthermore, regular biomass harvesting in these systems causes dynamic changes in aboveground biomass throughout the harvesting cycle, whereas fruit and timber tree-based systems develop more gradually over time (Cardinael et al., 2017;Huber et al., 2018;Lamerre et al., 2015). These differences in design and management influence how the microclimate is altered within an agroforestry system, and which components of the water balance, such as evapotranspiration (e.g. ...
Article
Climate change scenarios predict an increased occurrence of droughts and heatwaves, as well as extreme rainfall events in Central Europe. Alley cropping, which is the inclusion of rows of trees and shrubs in agricultural land, could enhance the resilience of cropping systems, as these systems are expected to positively modify the microclimate and water balance of croplands. This review analyses the effect of alley cropping on the micro-climate and water balance, based on the available evidence from temperate alley cropping systems. Within alley cropping systems, the tree rows generate gradients in microclimatic variables, whereby strongest effects are observed in or close to the tree rows. Field-scale studies on light intensity (n=20), wind speed (n=4) and surface runoff (n=3) all reported a reduction compared to sole cropping systems. Effects on air temperature (n=10), relative humidity (n=5) and evapotranspiration (n=6) varied among studies, with the majority reporting a decrease in daytime temperatures (50% of studies), variable effects on relative humidity (60%) and an increase in evapotranspiration (50%) due to higher evapotranspiration by trees. Highest variation among studies was found for soil moisture, with 41% of studies reporting temporal and spatial differences within the system. This variation among studies likely depends on the purpose of the trees (short rotation coppice vs. fruit and/or timber trees) and design of the system. Also site context, such as topography, landscape diversity and climate, could play a role, but these factors are rarely taken into account. Only few studies investigated landscape-scale effects (n=3), such as groundwater recharge and moisture recycling. Future research should investigate the role of site context in the functioning of alley cropping systems and quantify landscape-scale effects. The process understanding gained from those studies will contribute to designing alley cropping systems that enhance the climate change resilience of current central European cropping systems.
... Several definitions describe the term, but scientists from the early 1980s have broadly employed the International Council for Research in Agroforestry (ICRAF) definition. A.F.S. is defined as the land use management system in which trees and crops or pasture are grown in the same field [17]. The arrangement can be either in spatial arrangement or in a time sequence". ...
Article
Full-text available
Aims: Most of the land area of Saudi Arabia is either arid or hyper-arid. In the past decades, many efforts have been exerted to increase the green cover in Saudi Arabia, the most recent of which is the Saudi Green Initiative (S.G.I.), launched in 2021. S.G.I.’s main objectives are to increase the green land surface area and decrease carbon emissions. In this paper, the role of dryland agroforestry in mitigating the effects of climate change was reviewed, and its contribution to fulfilling S.G.I. was discussed. Methodology: Previously published literature, scholarly research articles, and conference proceeding papers, on agroforestry systems (A.F.S), carbon sequestration and nutrient dynamics under A.F.S over the past 34 years were critically reviewed, examined, and analysed to find various applications of AFS for climate change mitigation and carbon sinks with focus on arid land. Results: Forests are a vital source for climate change mitigation and adaptation and play a vital role as carbon sinks. A.F.S, eco-friendly and environmentally viable land use and management, provide immense potential to sequester carbon (C). A.F.S. is a reliable tool for increasing C sequestration. As a result of the worth granted to non-timber products, the application of A.F.S. could likewise reduce C emissions to the air by reducing the odds of concrete cutting of trees. Moreover, tree components are a source of C for the soil by means of root and leave decomposition. Conclusion: In the perspective of the high threat facing humanity paused the climate variability and climate change, many nations and countries have taken various measures to tackle it which included protecting natural forests, afforestation, managed the natural regeneration of green cover. A.F.S leads to better land-use efficiency, increases the green cover, and thus helps in mitigating climate change.
... Thus, in the short term, we can infer that there is a perception of qualitative improvement from the adoption of agroforestry, at least in superficial soil layers; however, this benefit has not been proven in quantitative terms. Nevertheless, Cardinael et al. (2015), Cardinael et al. (2017), Cardinael et al. (2018) showed that there was an increase in soil organic matter in an agroforestry system from the larger amount of biomass added to the system. In agreement with the results found in this study, the increase in fertility perceived by farmers and mentioned in the studies analyzed may be related to an increase in soil organic material in agroforestry systems. ...
Article
Full-text available
Soils have many ecological functions and provide various ecosystem services including support for global food and fuel production. However, FAO reports indicate that approximately one-third of the planet’s arable lands show levels of degradation from processes including soil erosion, low levels of nutrients, acidification, salinization, compaction, sealing, and contamination. These conditions are also found in Brazil where soil degradation is largely caused by inadequate land management. Worldwide, strategic policies have been presented to mitigate this problem, with emphasis on sustainable agriculture. Among them, agroforestry has been identified as a viable system for mitigating and recovering degraded areas. Agroforestry techniques have been developed and tested but are still not understood by farmers, due to their complexity. This study aimed to analyze experiences and studies with agroforestry reported from Australia, some countries in Africa, and Brazil to search for similarities in these complex systems and identify possible correlations to support the hypothesis that land recovery can be enhanced through soil management using agroforestry. A Sankey diagram was developed to illustrate relationships among problems, the adoption of agroforestry and improvements, and the most important contributions. Data analysis shows that the main problems related to soil degradation are soil erosion and decreased soil fertility, while the adoption of agroforestry systems proved to improve different aspects of soil quality and to be a safe path to sustainable agricultural production. To obtain more information on the adoption of these systems in different locations, soils, and climates, it is important to implement policies for reducing land degradation. Furthermore, the assessment of the economic, environmental and social benefits of improving soil fertility and decreasing erosion in agroforestry systems is necessary to validate the use of agroforestry as a sustainable agricultural practice.
... Then, using Equation (1) of Janzen et al. (2022) (despite all its assumptions and limitations), we can calculate an annual C gain (Cg) of the world's soil (Pg C yr − 1 ) due to the current NPP input: Similarly, comparing C gain (Cg) 0.234 to initial C stock (H) of 58.5 Mg C ha − 1 , we have 0.004 yr − 1 (i.e., 4 per 1000). Interestingly the rate ~ 0.2 Mg C ha − 1 yr − 1 is typical of SOC sequestration rates reported for cover cropping, agroforestry or other practices (Poeplau and Don, 2015;Cardinael et al., 2017;Minasny et al., 2017). ...
Article
Full-text available
Soil organic carbon (SOC) sequestration is the transfer of CO2 from the atmosphere into soil organic matter. It, therefore, relies on photo- synthesis and plant-derived carbon (C) input, which usually occurs through biomass production. Janzen et al. (2022) reminded us that when calculating SOC sequestration potential, we should recognise the source of C input to the soil as estimated by Net Primary Production (NPP). Indeed, increasing plant biomass production via NPP has been discussed as the most important driver of many SOC sequestration strategies (Soussana et al., 2019). Janzen et al. described a simple back-of-the-envelope calculation to demonstrate the limits of SOC stock increase as defined by the current NPP. While such a straightforward approach is reasonable to get a rough guestimate, it is important to recognise that there are limits to such a simplified modelling approach which carries significant uncertainties. In this comment, we discuss the limitations of such an approach and the way forward. Moreover, we show that Janzen et al.’s calculation con- tains inaccurate assumptions. When
... Ex-mining reclamation land currently has the opportunity to contribute to the development of cajuput agroforestry cultivation, especially in non-forest areas or Area Penggunaan Lain (APL) (Kodir et al. 2016). Agroforestry is a land-use system that combines agricultural and forestry crops has diversified production, ecological and social protection (Cardinael et al. 2017;Kaur et al. 2017;Suryani and Dariah 2012;Tarigan et al. 2019). Cajuput is highly adaptive developed with agroforestry systems (Priswantoro et al. 2021) to support food security programs . ...
Article
Full-text available
Cajuput (Melaleuca cajuputi) has the potential to be developed with monoculture and agroforestry patterns in ex-coal mining land. This study aimed to analyze the oil yield and cineole content of cajuput planted with monoculture and agroforestry and patterns in ex-coal mining land. The research design used a split-split plot design with three factors: planting pattern, fertilizer dosage, and lemongrass (Cymbopogon nardus) plant spacing. The variables measured included the oil yield and cineole content of cajuput. This study was conducted in cajuput monoculture and agroforestry patterns with lemongrass G2 variety and Sitrona 2 Agribun variety, with various spacing. The study was conducted for six months, with one harvest of cajuput leaves in the sixth month. The results showed that lemongrass’s planting pattern and plant spacing significantly affected cajuput oil yield. However, treatment of the dosage of fertilizer had no significant effect on cajuput oil yield. The treatment of lemongrass plant spacing only significantly affected cineole content, but planting patterns and fertilizer dosages had no significant effect. The cajuput agroforestry with the Sitrona 2 Agribun variety produced the highest oil yield (2.84%) and cineole content (50.70%), compared to the monoculture pattern and cajuput agroforestry with G2 variety. The lemongrass plant spacing of 0.5 m x 0.5 m produced the highest oil yield (2.73%) and cineole content (52.98%) compared to 1 m x 1 m plant spacing Cymbopogon nardus. Keywords: agroforestry, cineole, essential oil, ex-coal mining, Melaleuca cajuputi
Thesis
Full-text available
Agriculture is under increasing pressure to produce more food with less environmental impacts and in the face of a changing climate. Management practices capable of sequestering soil carbon (C) and improving overall soil health hold promise for sustainable intensification, as well as climate change mitigation and adaptation. As market and policy-based incentives develop to support these practices, however, it is critical that adequate sampling protocols, minimum viable data sets, and thresholds of management responses to soil health indicators are identified across the diversity of cropping systems and edaphoclimatic conditions. Much of the research into the impacts of agricultural management on soil C and soil health have been conducted in the Midwest, over the short-term, and to a shallow depth. Soil C dynamics and other soil health indicators are strongly influenced by climate and mineralogy, necessitating more research across a range of edaphoclimatic conditions. Further, detectable changes in soil C take decades to accrue, requiring long-term research. Proper accounting of changes in C stocks on a given acreage for climate mitigation strategies and economic incentive programs also necessitates sampling to a sufficient depth (minimum 1 meter or a root-limiting layer). Using long-term, on-farm interventions, controlling for cropping system, climate and soil type, this work investigates the impact of soil health practices on soil C in surface and subsurface soils, as well as on a suite of physical, chemical, and biological soil properties commonly used to assess soil health. Deep soil cores at a long-term, industrial scale, agricultural research station in a Mediterranean-type climate indicated that 19 years of cover cropping with annual composted poultry manure applications (4t ha-1) increased soil C to a depth of 200 cm by +21.8 Mg ha-1 relative to a -4.8 Mg ha-1 loss under conventional management (Chapter 1). Trends also indicated potential losses of -13.4 Mg ha-1 under conventional management with cover cropping, despite increases of +1.4 Mg ha-1 in the surface 0-30 cm, stressing the importance of deep soil sampling for greenhouse gas accounting purposes. Continuing the theme of deep soil C, a nearby regional survey of 10+ yr old hedgerows and adjacent cultivated fields across four soil types showed a strong impact of hedgerows on soil C to a depth of 100 cm, with an average difference of 3.85 kg C m-2 (0-100 cm) and few differences across the four soil types (Chapter 2). Most differences occurred in the surface 0-10 cm and the subsoil at 50-100 cm, indicating a dual role of surface management (litter accumulation, reduced disturbance) and deep, woody perennial roots. Soil type differences were only apparent in one of the four soil types, which differed substantially in parent material, mineralogy, and degree of weathering. Soil type did not influence the management effect and may indicate broad potential for hedgerows as a climate mitigation strategy. The magnitude of this strategy is limited, however, by the extent of hedgerows on a given farm/ranch. Revegetation of field margins with hedgerows also had a positive impact on a broad suite of physical, chemical, and biological parameters (0-20 cm) commonly associated with soil health (Chapter 3). Hedgerow values were greater than cultivated fields for nearly every indicator in the surface 0-10 cm, commonly 2-3 times greater. Fewer, smaller differences were observed at 10-20 cm. Total soil C and N, available C, microbial biomass C, aggregate stability, and surface hardness were some of the most sensitive and least variable indicators of management type. Texture, pH, and bulk density were more indicative of soil type. A composite of variables was necessary to explain most of the variation in the data, indicating the complexity of soil health.
Article
The carbon reduction potential and accounting principles in agroforestry outlined in this article cover four sectors – comparable to other land use-based strategies. 1) above-ground biomass, 2) below-ground biomass, 3) the soil, and 4) the up- and downstream sector. The recommendations cover ten frequently discussed topics and concerns, namely 1) additionally, 2) quantifiability, 3) displacement effects, 4) contribution to food security, 5) additional emissions, 6) longevity and durability, 7) traceability, 8) transaction and opportunity costs, 9) synergies and compromises with other goals, and 10) security, trust and transparency. If the recommendations developed are taken into account, the authors conclude that the climate protecting and mitigation services of agroforestry in the form of carbon reduction potential can and should be rewarded by climate or carbon certificates. On the one hand, this could be seen as an innovative and promising way of financing future agroforestry systems; on the other hand, it must be ensured that the measures meet minimum scientific and social requirements. If planned scientifically sound, reliable, transparent and ethically just, there is a good chance to create with climate certificates for agroforestry some contributing solutions for EUs ambitious climate target plan, a 55 percent greenhouse gas reduction by 2030 compared to 1990.
Article
Agricultural soils contribute significantly to global CO2 emissions. Soils can be both a CO2 sink and source depending on vegetation cover and the physical‐chemical characteristics of the soil environment. This study uses a new 14C dating approach to elucidate soil physical and geochemical properties that drive the production and transport of CO2 in cultivated and noncultivated drainage ditch soil environments. As expected, CO2 production and fluxes decrease with soil depth. Relating depth‐specific subsurface radiocarbon of CO2 (14CO2) to the atmospheric scenario generated from the testing of thermonuclear bombs in the 1950s and 1960s provides a means to date the labile organic C pools responsible for soil respiration that are otherwise generalized as having a modern age (≤1950 AD). Depth‐specific 14CO2 measurements showed that CO2 in shallow soils (≤22.5‐cm depth) have a younger age than CO2 in deeper soils (>22.5‐cm depth). The 14CO2 of surface emissions were comparable to those of shallow soils (≤22.5‐cm depth), showing that respiration in the deeper soils (>22.5‐cm depth) do not contribute significantly to surface emissions, and the higher CO2 concentrations at depth in soil result from processes limiting CO2 transport to surface. Coupling the 14C data with CO2 flux profiles confirms that CO2 effluxes are derived from modern soil organic C (<10 yr since sequestered) from the shallowest soil depths in all soil environments studied, but changes in physical soil processes governing gas transport at depth could release these deeper CO2 stores. The radiocarbon thermonuclear bomb pulse is a tool to identify CO2 transport in agriculture soils. Subsurface CO2 fluxes decrease with soil depth. Higher CO2 concentrations at depth in soil result from processes limiting CO2 transport to surface.
Article
Full-text available
Soil organic carbon plays a major role in the global carbon budget, and can act as a source or a sink of atmospheric carbon, thereby possibly influencing the course of climate change. Changes in soil organic carbon (SOC) stocks are now taken into account in international negotiations regarding climate change. Consequently, developing sampling schemes and models for estimating the spatial distribution of SOC stocks is a priority. The French soil monitoring network has been established on a 16 km × 16 km grid and the first sampling campaign has recently been completed, providing around 2200 measurements of stocks of soil organic carbon, obtained through an in situ composite sampling, uniformly distributed over the French territory. We calibrated a boosted regression tree model on the observed stocks, modelling SOC stocks as a function of other variables such as climatic parameters, vegetation net primary productivity, soil properties and land use. The calibrated model was evaluated through cross-validation and eventually used for estimating SOC stocks for mainland France. Two other models were calibrated on forest and agricultural soils separately, in order to assess more precisely the influence of pedo-climatic variables on SOC for such soils. The boosted regression tree model showed good predictive ability, and enabled quantification of relationships between SOC stocks and pedo-climatic variables (plus their interactions) over the French territory. These relationships strongly depended on the land use, and more specifically, differed between forest soils and cultivated soil. The total estimate of SOC stocks in France was 3.260 ± 0.872 PgC for the first 30 cm. It was compared to another estimate, based on the previously published European soil organic carbon and bulk density maps, of 5.303 PgC. We demonstrate that the present estimate might better represent the actual SOC stock distributions of France, and consequently that the previously published approach at the European level greatly overestimates SOC stocks.
Article
Full-text available
Increasing attention is being paid by farmers and extension services to soil amendment with small branches, especially as chipped ramial wood (CRW), but the scientific validation of this practice is incomplete. The present work summarizes statistically significant results regarding the effects of such branch amendments, both buried and mulched, on crop yield and soil properties in temperate and tropical regions. Broadly speaking, soil amendment with CRW has a positive effect on crop yield, except for the crop that immediately follows the first burying of CRW in sandy soils (which has mainly been tested in temperate regions); however this negative effect can be limited when nitrogen is simultaneously applied. Moreover, CRW application increases soil organic matter content, stimulates soil biological activities, improves medium-term nutrient availability and - especially as mulch - soil hydro-physical properties (moisture, porosity, structure, etc.). The effects of CRW application can be affected by several factors such as tree species and amendment characteristics (amount, periodicity, chip size, etc.), however the results available do not allow precise recommendations. Moreover, the benefit of CRW as compared with non-woody amendments is poorly documented.
Article
Full-text available
Background and Aims Fine roots play a major role in the global carbon cycle through respiration, exudation and decomposition processes, but their dynamics are poorly understood. Current estimates of root dynamics have principally been observed in shallow soil horizons (<1 m), and mainly in forest systems. We studied walnut (Juglans regia × nigra L.) fine root dynamics in an agroforestry system in a Mediterranean climate, with a focus on deep soils (down to 5 m), and root dynamics throughout the year. Methods Sixteen minirhizotron tubes were installed in a soil pit, at depths of 0.0–0.7, 1.0–1.7, 2.5–3.2 and 4.0–4.7 m and at two distances from the nearest trees (2 and 5 m). Fine root (diameter ≤ 2 mm) dynamics were recorded across three diameter classes every 3 weeks for 1 year to determine their phenology and turnover in relation to soil depth, root diameter and distance from the tree row. Results Deep (>2.5 m) root growth occurred at two distinct periods, at bud break in spring and throughout the winter i.e., after leaf fall. In contrast, shallow roots grew mainly during the spring-summer period. Maximum root elongation rates ranged from 1 to 2 cm day⁻¹ depending on soil depth. Most root mortality occurred in upper soil layers whereas only 10 % of fine roots below 4 m died over the study period. Fine root lifespan was longer in thicker and in deeper roots with the lifespan of the thinnest roots (0.0–0.5 mm) increasing from 129 days in the topsoil to 190 at depths > 2.5 m. Conclusions The unexpected growth of very deep fine roots during the winter months, which is unusual for a deciduous tree species, suggests that deep and shallow roots share different physiological strategies and that current estimates based on the shortest root growth periods (i.e., during spring and summer) may be underestimating root production. Although high fine root turnover rates might partially result from the minirhizotron approach used, our results help gain insight into some of the factors driving soil organic carbon content.
Article
Full-text available
Ce numéro d'Innovations Agronomiques comprend les articles correspondant aux présentations du colloque « Atténuation des Gaz à Effet de Serre par l’agriculture » qui s'est tenu à Versailles le 4 juin 2014.
Article
Full-text available
L_amendement du sol avec des branches, notamment avec des bois raméaux fragmentés (BRF), suscite un intérêt croissant chez les agriculteurs et les services de vulgarisation, mais la validation scientifique de cette pratique est incomplète. Cet article synthétise les résultats statistiquement significatifs concernant l_effet d_apports enfouis ou paillés (mulch) de BRF sur les cultures et le sol, en milieu tempéré et tropical. L_apport de BRF a généralement un effet positif sur le rendement agricole, sauf pour la culture qui suit immédiatement un premier enfouissement en sol sableux (lequel a surtout été testé en milieu tempéré ; le résultat reste peu documenté en milieu tropical) ; cet effet négatif peut cependant être limité par un apport d_engrais azoté. Par ailleurs, l_apport de BRF, surtout en mulch, améliore les propriéteés physico-hydriques du sol (humidité, porosité, structure), enrichit le sol en matière organique, stimule l_activité biologique, et accroît la disponibilité des nutriments à moyen terme. Les effets des BRF sont modulés par plusieurs facteurs, comme l_essence forestière utilisée et les modalités d_apport (dose, périodicité, dimension des fragments, etc.) ; mais les résultats répertoriés ne permettent guère de formuler des recommandations précises. Par ailleurs, l'intérêt des BRF par rapport aux amendements organiques non ligneux est mal documenté
Article
Full-text available
Accurate assessment of anthropogenic carbon dioxide (CO2) emissions and their redistribution among the atmosphere, ocean, and terrestrial biosphere is important to better understand the global carbon cycle, support the development of climate policies, and project future climate change. Here we describe data sets and a methodology to quantify all major components of the global carbon budget, including their uncertainties, based on the combination of a range of data, algorithms, statistics, and model estimates and their interpretation by a broad scientific community. We discuss changes compared to previous estimates, consistency within and among components, alongside methodology and data limitations. CO2 emissions from fossil fuel combustion and cement production (EFF) are based on energy statistics and cement production data, respectively, while emissions from land-use change (ELUC), mainly deforestation, are based on combined evidence from land-cover-change data, fire activity associated with deforestation, and models. The global atmospheric CO2 concentration is measured directly and its rate of growth (GATM) is computed from the annual changes in concentration. The mean ocean CO2 sink (SOCEAN) is based on observations from the 1990s, while the annual anomalies and trends are estimated with ocean models. The variability in SOCEAN is evaluated with data products based on surveys of ocean CO2 measurements. The global residual terrestrial CO2 sink (SLAND) is estimated by the difference of the other terms of the global carbon budget and compared to results of independent dynamic global vegetation models forced by observed climate, CO2, and land-cover-change (some including nitrogen–carbon interactions). We compare the mean land and ocean fluxes and their variability to estimates from three atmospheric inverse methods for three broad latitude bands. All uncertainties are reported as ±1σ, reflecting the current capacity to characterise the annual estimates of each component of the global carbon budget. For the last decade available (2004–2013), EFF was 8.9 ± 0.4 GtC yr−1, ELUC 0.9 ± 0.5 GtC yr−1, GATM 4.3 ± 0.1 GtC yr−1, SOCEAN 2.6 ± 0.5 GtC yr−1, and SLAND 2.9 ± 0.8 GtC yr−1. For year 2013 alone, EFF grew to 9.9 ± 0.5 GtC yr−1, 2.3% above 2012, continuing the growth trend in these emissions, ELUC was 0.9 ± 0.5 GtC yr−1, GATM was 5.4 ± 0.2 GtC yr−1, SOCEAN was 2.9 ± 0.5 GtC yr−1, and SLAND was 2.5 ± 0.9 GtC yr−1. GATM was high in 2013, reflecting a steady increase in EFF and smaller and opposite changes between SOCEAN and SLAND compared to the past decade (2004–2013). The global atmospheric CO2 concentration reached 395.31 ± 0.10 ppm averaged over 2013. We estimate that EFF will increase by 2.5% (1.3–3.5%) to 10.1 ± 0.6 GtC in 2014 (37.0 ± 2.2 GtCO2 yr−1), 65% above emissions in 1990, based on projections of world gross domestic product and recent changes in the carbon intensity of the global economy. From this projection of EFF and assumed constant ELUC for 2014, cumulative emissions of CO2 will reach about 545 ± 55 GtC (2000 ± 200 GtCO2) for 1870–2014, about 75% from EFF and 25% from ELUC. This paper documents changes in the methods and data sets used in this new carbon budget compared with previous publications of this living data set (Le Quéré et al., 2013, 2014). All observations presented here can be downloaded from the Carbon Dioxide Information Analysis Center (doi:10.3334/CDIAC/GCP_2014).