Journal of Arid Environments 198 (2022) 104696

Contents lists available at ScienceDirect

Journal of Arid Environments

journal homepage: www.elsevier.com/locate/jaridenv

Root litter decomposition in a sub-Sahelian agroforestry parkland

dominated by Faidherbia albida
Lorène Siegwart a, Isabelle Bertrand a, *, Olivier Roupsard a, b, c, Maxime Duthoit a, b,
Christophe Jourdan a, b
a
UMR Eco&Sols, Univ Montpellier, CIRAD, INRAE, IRD, Montpellier SupAgro, Montpellier, France
b
CIRAD, UMR Eco&Sols, F-34398, Montpellier, France
c
LMI IESOL, 18524, Dakar, Senegal

A R T I C L E I N F O A B S T R A C T

Keywords: In agroforestry systems, fine roots grow at several depths due to the mixture of trees and annual crops. The
Soil organic carbon decomposition of fine roots contributes to soil organic carbon stocks and may impact soil fertility, particularly in
Soil nutrients poor soils, such as those encountered in sub-Sahelian regions. The aim of our study was to measure the
Root litter quality
decomposition rate of root litter from annual and perennial species according to soil depth and location under
Soil depth
and far from trees in a sub-Sahelian agroforestry parkland.
Vigna unguiculata
Pennisetum glaucum Soil characteristics under and far from the trees were analysed from topsoil to 200 cm depth. Faidherbia tree,
pearl millet and cowpea root litter samples were buried in litterbags for 15 months at 20, 40, 90 and 180 cm
depths.
Root litter decomposition was mainly impacted by soil moisture and soil depth. Faidherbia decomposed more
slowly (36 ± 12% remaining mass after 15 months) than cowpea and pearl millet roots (23 ± 7% and 29 ± 11%
respectively). Pearl millet aboveground biomass, at harvesting time, was twice as high under (992 g m− 2) than
far (433 g m− 2) from the tree, and belowground biomass (0–200 cm of depth) was 30.9 g m− 2 and 19.6 g m− 2
under and far from the tree, respectively. Faidherbia fine roots contributed slightly (p-value < 0.1) to higher
stocks of C under the tree (7761 ± 346 g m− 2) than far from it (5425 ± 558 g m− 2) and from 0 cm down to 200
cm depth.

1. Introduction challenge in highly weathered tropical soils with low C contents, and
agroforestry practices may contribute to overcoming this challenge.
In the current context of global warming, soil carbon (C) sequestra­ In agroforestry systems, the diversity of the plant species and new
tion can contribute to mitigating the greenhouse effect (Nair et al. ecological niches for biodiversity (Leaky 1996) lead to an enrichment of
2009a, 2009b; Chenu et al., 2019). In the tropics, C sequestration can aerial, root and microbial biomasses (Nair et al., 2009b; Lagerlöf et al.,
more specifically contribute to the improvement of food security and to 2014) with a trade-off between soil fertility improvement and compe­
climate change adaptation (Paustian et al., 2016). Tropical soils are tition for growth resources (Rao et al., 1997). C inputs in agroforestry
characterized by lower nutrient contents (Feller and Beare 1997) and systems are mostly related to the decomposition of aboveground
more rapid C turnover than those in temperate systems (Six et al., 2002). biomass (tree litterfall and crop residues) and belowground biomass
A recent synthesis based on 48 studies performed on tropical soils from originating from tree and crop root turnover and/or mortality and rhi­
13 countries demonstrated that the main determinants of soil organic zodeposition (Kuzyakov and Domanski 2000). Cardinael (2015) esti­
carbon (SOC) accumulation were C inputs, duration of the experiments mated that tree and crop fine roots each contribute 30% to organic
and management practices (Fujisaki et al., 2018). However, this syn­ matter input in agroforestry systems. Fine roots are generally more
thesis did not consider agroforestry practices due to the lack of refer­ recalcitrant than aerial aboveground biomass to soil microbial decom­
ences, although agroforestry is assumed to enhance C storage in soils position (Rasse et al., 2005; Bertrand et al., 2006; Freschet et al., 2013),
(Smith et al., 2014). Increasing soil C sequestration is a current and they have the potential to increase soil C stocks. While several

* Corresponding author.
E-mail address: [email protected] (I. Bertrand).

https://doi.org/10.1016/j.jaridenv.2021.104696
Received 28 June 2021; Received in revised form 7 October 2021; Accepted 3 December 2021
Available online 12 December 2021
0140-1963/© 2021 Elsevier Ltd. All rights reserved.
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

studies have demonstrated the chemical characteristics responsible for moisture sensors buried at 15, 30, 50, 75, 100, 125, 150, 175 and 200 cm
slow root decomposition rates (Machinet et al., 2009; Cotrufo et al., in an open area of the plot (far from the trees and close to the weather
2013; Prieto et al., 2016), the impact of soil depth has been less studied, station). The rainfall was measured on site with an automatic tipping
although roots occur at different depths in the soil profile. This is bucket (Texas Electronics, model TE525 mm). Data were recorded every
particularly true for agroforestry systems where tree, herbaceous and 30 min over the entire study period. The average daily temperature and
crop roots colonize different soil layers, especially at depth (Cardinael soil moisture and sum of rainfall were calculated (Fig. 1). The soil is a
et al., 2015; Germon et al., 2016; Battie-Laclau et al., 2020). sandy tropical ferralitic soil (Maignien 1965); it is classified as an Are­
Root litter decomposition depends not only on litter quality but also nosol (IUSS Working group WRB, 2014). The water table is located at
on pedoclimatic conditions (Makkonen et al., 2012) and soil microbial approximately 5–6 m depending on the season.
communities and activity (Herman et al., 2012). These biotic and abiotic The studied agroforestry system was composed mainly of Faidherbia
soil characteristics are strongly impacted by the introduction of trees in albida trees (85% occurrence), with a density of 6.8 trees ha-1, which
arable lands. The introduction of trees causes spatial heterogeneity in represents an average canopy cover of 5.14% measured over an area of
soil temperature and humidity (Monteith et al., 1991; Rao et al., 1997; 15 ha (Roupsard et al., 2020; Rahimi et al., 2021). Faidherbia trees were
Lin 2007) as well as soil microbial biomass abundance and composition associated with groundnut (Arachis hypogaea) or pearl millet (Pennisetum
(Chander et al., 1998; Guillot et al., 2021; Liu et al., 2019) and soil C glaucum) according to annual rotations. In June 2018, pearl millet was
stocks (Cardinael 2015). In a recent study performed in the same area as manually sown in the studied plot at a distance of 80 cm between each
our study site, Roupsard et al. (2020) demonstrated that the whole pearl sowing pocket. Cowpea (Vigna unguiculata) was sown at the same time in
millet plant dry mass was 2.2 times higher under the Faidherbia tree a neighbouring plot with the same soil type and climatic conditions.
crown than far from the tree. As a consequence, biomass inputs may be There was no amendment applied to these plots, and harvesting was
more important near trees, which could induce a modification of the soil conducted in October 2018.
chemical and physical properties. While the impact of trees on crop
yield, climatic conditions and soil C stocks at a local scale was previously 2.2. Above- and belowground biomass sampling
investigated in shallow soil horizons (Oelbermann et al., 2004; Oelber­
mann and Voroney 2007; Lin 2007; Roupsard et al., 2020), to our The biomass sampling campaign was conducted in October 2018,
knowledge, no studies have investigated the impact of trees on deep soil immediately before the pearl millet crop harvest. According to the large
characteristics or on tree and crop root decomposition. volume of soil to excavate and sieve (each pit was 8 m3), only 2 pits
Soil properties may vary with soil depth given that the total organic C could be prepared, one under and one far from a tree. We sampled 3
content and microbial biomass decrease with depth (Hicks Pries et al., walls in each pit, thereby assuming independence of the results. The pit
2018). Soil temperature and moisture tend to be less subject to varia­ under the tree was chosen under a Faidherbia individual representative
tions in deeper soil layers than in topsoil. Hicks Pries et al. (2018) of the tree population (Diatta 2021), with a height of 13.5 m and a
showed that slower root decomposition could be responsible for higher circumference at breast height of 2.84 m.
stable C stocks at soil depth. A recent meta-analysis by Balesdent et al. The aboveground parts of pearl millet were collected in two subplots,
(2018) performed on 112 grassland, forest and cropland sites demon­ each measuring 2 m × 2 m. One subplot was located under the selected
strated that the subsoil (30–100 cm depth) stored 47% of the C in the Faidherbia tree crown (1.5 m from the trunk, crown radius of 5 m),
first metre of the soil profile. This deep C storage despite a lower litter whereas the second subplot was located far from any tree and at a
input is related to root mortality and rhizodeposition and to the reduced minimum distance of 30 m from the first subplot. The subplots used for
decomposition rates at depth (Guenet et al., 2013). Data are lacking on biomass quantification were at the same location as the pits. Each sub­
root litter dynamics and C stocks in deep soil horizons, especially for plot included four pearl millet pockets. The vegetative biomass was split
tropical areas. into ears, stems, leaves and stumps. All samples were oven-dried for 48 h
The aim of our study was thus to measure soil characteristics, at 65 ◦ C before weighing.
including soil C stocks and the root decomposition rate, according to soil After the aboveground biomass was sampled, two pits of 2 m × 2 m
depth in a sub-Sahelian agroforestry park dominated by Faidherbia × 2 m were dug at the same locations. For each pit, roots were sorted by
albida trees and to account for the tree effect. manually sieving the soil at 2 mm from a total soil volume of 8 m3 and
We hypothesized that (i) soil fertility, indicated by the C and nutrient split according to the plant species (pearl millet and Faidherbia tree) and
contents, would be higher in the topsoil than in deeper soil layers and the corresponding soil layer (0–40 cm, 40–100 cm, 100–150 cm and
under trees than far from the trees due to the presence of leguminous 150–200 cm). Given the small quantity of pearl millet roots found at
tree species, (ii) the root litter decomposition rate would mainly depend great depth, the root biomasses of pearl millet in soil layers 100–150 and
on the plant species (i.e., root tissue quality), roots would decompose 150–200 cm were summed for each profile. Then, Faidherbia roots were
faster under trees and (iii) root litter would decompose more slowly in sorted manually, and their diameter (D) was measured with a digital
deep soil layers than in topsoil because of the stable pedoclimatic calliper to separate fine roots (D < 2 mm) from medium roots (10 mm >
conditions. D ≥ 2 mm). All samples were washed on a 0.5 mm sieve and oven-dried
for 48 h at 65 ◦ C before weighing. The belowground biomass was
2. Materials and methods assessed for 2 subplots × 2 plant species × 4 soil layers (× 2 root
diameter categories for Faidherbia) after correction for the ash content.
2.1. Study site To this end, a subsample of 1 g of the washed root sample was burned at
500 ◦ C for 4 h to remove organic matter, and the remaining mineral ash
The “Faidherbia-Flux” collaborative observatory for greenhouse gas was weighed and deducted from the dry root mass.
balance and ecosystem services (https://lped.info/wikiObsSN/?Faidhe Supplementary roots were collected between 0 and 40 cm depth in
rbia-Flux) is located in the natural agro-silvo-pastoral parkland of Sob the neighbouring plot planted with cowpea and prepared as described
(14◦ 29′ 45N, 16◦ 27′ 13W), 135 km East of Dakar, on the Bambey-Fatick above for millet and Faidherbia roots.
transect, West Senegal, (Roupsard et al., 2020). The climate is
sub-Sahelian, with an average annual rainfall of 500 mm (Lalou et al., 2.3. Litterbag experiment
2019) and an average temperature of 29.6 ◦ C (Ndiaye et al., 2001). The
soil temperature (◦ C) was measured with thermocouples buried at 0, 2, We performed a 464-day root litter decomposition experiment with
5, 15, 30, 60, 100, 150 and 200 cm. The soil volumetric water content fine roots of Faidherbia tree, cowpea and pearl millet. A subsample of
(m3H2O m-3soil) was measured with TDR (time domain reflectometry) 1.5 g root litter was inserted in 10 cm × 20 cm nylon mesh screens of 1

2
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Fig. 1. Dynamics of rainfall (a), volumetric

soil water content (b) and soil temperature
(c) over time according to soil depth (from
the topsoil to 200 cm deep) from the
beginning (d0: 10/15/18) to the end of the
experiment (d5: 01/22/20). On the x-axis,
d0 to d5 correspond to the sampling dates
after 1.5 (d1), 3 (d2), 6 (d3), 9 (d4) and 15
(d5) months of decomposition. Data are
presented as daily averages. The wet season
is represented in blue, and the dry season is
represented in yellow. (For interpretation of
the references to color in this figure legend,
the reader is referred to the Web version of
this article.)

mm (Diatex), hereafter referred to as root litterbags. The mesh size of 1 In total, 3 plant species × 2 locations (subplots) × 4 soil depths × 3
mm allowed all decomposer communities, including small in­ pit walls × 5 sampling dates = 360 litterbags were buried. However, due
vertebrates, to penetrate the nylon mesh and establish themselves on the to a shortage in the initial sampling of root biomass encountered in the
decomposing roots (Handa et al., 2014). On October 15th, 2018, cor­ pits, cowpea and Faidherbia litter samples were not buried at 40 and 90
responding to the harvest period, three litterbag replicates per plant cm and thus were not collected on all dates (see Fig. 2).
species (Faidherbia tree, pearl millet or cowpea) were buried at four soil After litterbag collection, the remaining root litter was carefully
depths (20, 40, 90 and 180 cm) on 3 different walls (east, north and west retrieved, and the soil adhering to the decomposed roots was carefully
walls) in each subplot (located under and far from the tree). The litter­ removed by shaking by hand before being oven-dried for 48 h at 65 ◦ C.
bags were buried at approximately 50 cm perpendicular to the pit walls Ash corrections were made on a subsample to remove soil particle
to prevent desiccation or temperature fluctuations as much as possible. contamination as previously described. The relative humidity of the soil
Each hole made to insert the litterbags perpendicular to the pit wall was around the litterbags was measured from the oven-dried soil samples.
filled with soil from the same hole. Each litterbag was replicated five The remaining dry mass in each litterbag was calculated as
times to allow five sampling campaigns (d1 to d5), which were sched­
Mf
uled after 1.5, 3, 6, 9 and 15 months of root decomposition. The first RDM = × 100
Mi
months of decomposition corresponded to the dry season (d1 to d3),
while the wet season started immediately before the fourth litterbag where RDM is the remaining dry mass (%), Mi is the initial litter dry mass
sampling (see Fig. 1). The last sampling occurred during the next dry (g) and Mf is the final litter dry mass (g).
season (d5). The remaining root dry mass according to a time axis for each species

3
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Fig. 2. Sampling strategy in the two pits (far from and under the tree), for each root litter type (pearl millet, cowpea and Faidherbia), at four depths (20, 40, 90 and
180 cm) and for five sampling dates (after 1.5 (d1), 3 (d2), 6 (d3), 9 (d4) and 15 (d5) months). Each litterbag was replicated on three pit walls (northern (N), eastern
(E) and western (W) soil profiles). Missing treatments are due to root sample shortages.

at each location and each soil depth gave the decomposition kinetics, 2.4. Initial litter quality
where the Y intercept was named d0. The time axis was expressed on
standardized days depending on the soil temperature at each soil depth. Initial root chemical qualities were determined for the three plant
The time was normalized by temperature using the method published by species (pearl millet, cowpea and Faidherbia tree). C fractions (soluble
Mary et al. (1999) at a reference temperature of 25 ◦ C, arbitrarily chosen compounds, cellulose, hemicellulose and lignin) were assessed with a
as commonly used by Balesdent and Recous (1997): fibre analyser (Fibretherm®, Gerhardt) on a 500 mg root litter sub­
sample following the Van Soest protocol (Goering and Van Soest 1970).
Dmeas
Dcorr25 = K×(T− Tref ) Root C and N elemental composition was determined with an automatic
e−
elemental analyser (Flash, 2000, ThermoFisher Scientific) on 3 mg
where Dcorr25 (days) is the time normalized at Tref , Tref (◦ C) is the subsamples of root litter. For the total root P content, 50 mg of litter
reference temperature (25 ◦ C), Dmeas is the measured time (days), T (◦ C) powder was mixed with 65% HNO3 and mineralized for 15 min at 200 ◦ C
is the average soil temperature of each day, and K is the thermal coef­ in a Milestone Ethos Easy microwave under standard and blank condi­
ficient (K = 0.115 for kinetics of SOC decomposition at 25 ◦ C (Balesdent tions. The total P content was quantified colorimetrically with the yel­
and Recous 1997)). Then, for each root species at each location and each low vanadomolybdate assay (Koenig and Johnson 1942).
soil depth, the decomposition kinetics were determined by a regression The proportion of C originating from roots and remaining in the soil
between the remaining dry mass and standardized time. To better fit our after 15 months of decomposition was calculated for 2 species (pearl
data, two linear regressions were applied for individual decomposition millet and Faidherbia fine roots) at both locations (under and far from
kinetics: the first regression with a k1 coefficient was based on the first the tree) and at 3 depths (0–40, 40–100 and 100–200 cm, which
sampling date (from d0 to d1), and the second regression had a k2 co­ matched the root biomass sampling and litterbag experimental setup) by
efficient (from d1 to d5). Two linear regressions × 3 root species × 2 multiplying the root carbon content (%) and the remaining mass at d5
locations × 4 soil depths = 48 coefficients were obtained. Regressions (%). Decomposition data were missing at 90 cm depth for Faidherbia,
with an R-squared value lower than 50% were removed from the and an average proportion of C at 20 and 180 cm was thus used. Then,
dataset. this calculated proportion was multiplied by the living root carbon
stocks (gC m− 3) to give the amount of remaining C originating from
roots and remaining in the soil after 15 months of decomposition for

4
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

each species and at each depth. For pearl millet as an annual crop, the 3 individuals each.
totality of the root carbon entered the soil at each harvest and thus at The initial difference in the quality of the root litter from the three
each year. For Faidherbia as a perennial tree, we considered that 0.56% plant species was analysed with one-way analysis of variance for each
of the root carbon was entering the soil each year according to the acacia variable (soluble fraction, hemicellulose, cellulose, lignin, total C, total
root turnover (Jha and Prasad Mohapatra 2010). This calculation was N, and total P contents and C:N). To analyse the variations in the hu­
not performed for cowpea because the root biomass in the soil profile midity of the soil in contact with the litterbags, a linear mixed model was
was not assessed for pearl millet or for Faidherbia. applied to the relative humidity, with location, soil depth, plant root
species and sampling date as fixed factors and the 3 replicated profiles as
2.5. Soil sampling and analyses random factors. To analyse the effect of location, soil depth and plant
species on root litter decomposition, linear mixed models were applied
The soil sampling campaign was conducted in late October 2018 to the remaining litter dry mass on each sampling date and to the k1 and
immediately after the pits were dug. In each pit (under and far from the k2 decomposition rates, with soil depth, location and plant species as
tree), in three out of four faces (taken as replicates), soil samples were fixed factors and the 3 replicated profiles as random factors.
collected at different depths (0–10 cm, 10–20 cm, 20–40 cm, 40–70 cm, To analyse the carbon inputs from the roots (Fig. 10), linear mixed
70–100 cm, 100–130 cm, 130–160 cm, 160–200 cm). The soil sampling models were applied to the soil C stocks, to the tree living fine root C
was more detailed than the experimental design of the litterbags to stocks, to the pearl millet living root C stocks and to the remaining C in
obtain a precise characterization of the soil profile. Soil was sampled the soil after 15 months of decomposition (all data in gC m− 3), with
where the litterbags were buried. Each sample was analysed by the location, soil depth and plant root species as fixed factors and the 3
LAMA laboratory (IRD-US Imago, Dakar, Senegal) for total soil C and N replicated profiles as random factors. For all the linear mixed models
contents by dry combustion (Matejovic 1997). The mineral C content and analyses of variance, lme4 and car packages were used. The
was assumed to be insignificant, and the measured total soil C was thus normality of the residues was always verified with a Shapiro-Wilk test,
associated with soil organic C. Soil pH was measured in a 1:2.5 and the homogeneity of the variances was verified with a Bartlett test.
soil-water suspension. Available phosphorus was determined according When necessary (p-values < 5%), Box-Cox (boxcox) or Yeo Johnson
to the Olsen method and was measured by the malachite green method (jtrans) transformations were applied.
(Ohno and Zibilske 1991). Soil mineral N was extracted with a 1:4 soil-1 A simple ordination of all the variables was conducted for a principal
M KCl solution, NO3- and NH4+ were determined by continuous flow component analysis with the “vegan” and “factoextra” packages. Among
colorimetry (SKALARSA 3000 flow analyser), and the sum of NO3- and the soil depths that were analysed for the initial soil characterization,
NH4+ represented the mineral soil N content. Soil texture was deter­ only 4 depths were selected for this analysis (10–20, 20–40, 70–100 and
mined based on five fractions (clay, silt (fine + coarse), sand (fine + 100–130 cm) to match the experimental design of the litterbags. Wilk’s
coarse)). tests allowed the identification of qualitative variables (location, depth
The soil bulk density was assessed according to the cylinder method and plant species) that significantly separated the individuals with the
(Blake and Hartge 1986) in each pit (under and far from the tree) on two “FactoInvestigate” package.
out of four faces (as replicates) at ten soil depths (10, 20, 40, 60, 80, 100,
120, 140, 160 and 180 cm). 3. Results
SOC stocks were calculated at each location (under and far from the
tree) and each soil depth following the ‘M1’ method described by Poe­ 3.1. Effects of depth and location on soil characteristics
plau et al. (2017) as follows:
The soil texture was globally very sandy, with more than 70% sand in
Cstock i,j = BDmean × Ctot × w every sample (Fig. 3), but the texture was significantly impacted by soil
depth (Supplementary Table 1), with soils richer in clay and lower in
where Cstock i,j is the soil C stock at location j in soil layer i (g m-2), w is coarse sand in the deeper layers (Fig. 3a). Location impacted only fine
the width of soil layer i (m), BDmean is the mean bulk density of soil layer i sand, with a higher content far from the tree (p-value = 1.94 × 10− 2),
(g m− 3) and Ctot is the amount of total soil C measured in soil layer i at while an interaction between soil depth and location was observed for
location j (g g− 1soil). To compare the surface soil layers with the deep the silt content (p-value = 7.63 × 10− 5 combined with soil depth).
layers while the compaction was different due to ploughing of the The total C and N contents were not significantly impacted by
topsoil layers, we also calculated the C stock at an equivalent soil mass location (Supplementary Table 1). However, the soil total C and total N
following the method presented by Ellert and Bettany (1995). contents tended to be higher under the tree than far from the tree
The total SOC stock in the whole soil profile was calculated for each (Fig. 4a and b). In this poor Arenosol, the total soil C did not exceed
location as the sum of the SOC stock in each layer. 0.45% in the surface layer. At both locations, soil depth strongly affected
the total C (F = 30.17, p-value = 3.0 × 10− 11) and N contents (F = 11.30,
2.6. Statistical analyses p-value = 1.2 × 10− 6) with a strong decrease from a depth of 30 cm. The
C:N ratio, soil pH, and soil available phosphorus and mineral N contents
For each measurement, data are presented as the mean values ± were impacted by the interaction of depth and location, while only the
standard deviation of 3 replicates. Whenever the location (far from and mineral N and C:N ratios were significantly affected by soil depth
under the tree) had no significant effect according to the methods (Supplementary Table 1). In the first 20 cm, the C:N ratios increased
described below, the average value of 6 replicates was calculated from 12.7 to 14.0 far from the tree and then decreased to 8.7 at 180 cm,
instead. All statistical analyses were processed with R Software (version while under the tree, the C:N ratios increased from 11.0 at the surface to
4.0.2) (R Core Team 2020). 14.3 at a depth of 1 m (Fig. 4c). Soil pH presented values ranging be­
To analyse the effect of depth and location on soil characteristics, tween 6 and 7 in the topsoil. Below 40 cm, soil under the tree presented
linear mixed models were applied to each soil variable, with soil depth higher pH values (6.9 ± 0.6 between 40 and 180 cm) than those far from
and location as fixed factors and the 3 replicated profiles as random the tree (5.7 ± 0.3) (Fig. 4d). As often occurs in tropical soils, available
factors. Data from the same soil profile were considered dependent on phosphorus was very low (less than 3 mg kg− 1) and significantly higher
each other. Post hoc Tukey tests allowed us to determine the significance under than that far from the tree (F = 3.77, p-value = 5.6 × 10− 3,
of the differences between each category of soil depth and location. C Supplementary Table 1). The available phosphorus decreased with
stocks in both locations were compared for each soil layer with Wilcoxon depth to less than 1 mg kg− 1 at 180 cm for both locations (Fig. 4f).
rank sum tests, as required for comparisons between 2 populations with Mineral N presented similar patterns, with average values of 5.5 ± 2.8

5
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Fig. 3. Comparison of soil texture (%) variations in clay (a), silt (b), fine sand (c) and coarse sand (d) in the soil profile from topsoil to a depth of 180 cm in the pits
under (dark) and far from the Faidherbia tree (white). Data are mean values from 3 pit walls, and error bars represent the standard deviation (n = 3).

Fig. 4. Total C content (a), total N content (b), C:N ratio (c), soil pH (d), mineral N content (e) and available phosphorus content (f) in the soil profile from topsoil to
a depth of 180 cm in the pits under (dark) and far from the tree (white). Data are mean values from 3 pit walls, and error bars represent the standard deviation (n
= 3).

and 9.4 ± 3.1 mg kg− 1 in the topsoil far from and under the tree and Despite important differences in total C stocks within the whole
decreasing to 5.0 ± 3.6 and 3.3 ± 0.1 mg kg− 1 at 180 cm, respectively profile under the tree (7761 ± 346 g m-2, n = 3) compared to far from the
(Fig. 4e). tree (5425 ± 558 g m-2, n = 3), the samples at each soil depth did not

6
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Table 1 below 40 cm depth. Far from the tree, tree roots were rare between 0 and
Total soil carbon stocks (g m− 2) of equivalent soil mass according to location 200 cm depth (Fig. 6).
(under or far from the tree) at different soil depths down to 200 cm. Data are
mean values ± standard deviation (n = 3). Significant differences between both
locations for each soil layer were tested with Wilcoxon tests. 3.3. Root litter quality, soil moisture and decomposition rates
Carbon stocks on equivalent soil mass (g Results of the tests of
m-2) Wilcoxon Faidherbia litter was significantly enriched in lignin compared to
cowpea and pearl millet litter (Table 2). Pearl millet roots presented a
Soil depth Under tree Far from tree W p-value
similar amount of lignin as in cowpea, while their soluble fraction was
0–10 cm 883 ± 295 634 ± 44 6 0.66 lower. However, the cellulose content was higher in pearl millet than in
10–20 cm 712 ± 164 543 ± 145 8 0.2
20–40 cm 907 ± 73 634 ± 25 6 0.2
cowpea, while hemicellulose was not significantly different between the
40–70 cm 1272 ± 147 1064 ± 30 9 0.1 two crops. Large differences in the N content explained the important
70–100 cm 1083 ± 353 684 ± 52 9 0.1 variations in C:N ratios, which varied from 13.0 for Faidherbia fine roots
100–130 cm 769 ± 23 536 ± 88 9 0.1 to 30.2 and 32.5 for cowpea and pearl millet roots, respectively.
130–160 cm 920 ± 57 623 ± 34 9 0.1
The relative humidity of the soil in contact with the litterbags
160–200 cm 1214 ± 101 919 ± 97 9 0.1
significantly increased with soil depth (F = 337.9, p-value < 2.2 ×
Total stock 7761 ± 346 5425 ± 558 9 0.1 10− 16, Table 3), and it was higher under the tree than far from the tree
(F = 24.9, p-value = 7.3 × 10− 3, Table 3). The soil humidity was still
differ significantly between locations under and far from the tree high on d1 (9.0 ± 5.9 m3H2O m− 3soil) from the previous wet season and
(Table 1). decreased significantly from d1 to d3 (F = 100.5, p-value = < 2.2 ×
10− 16, Table 3). The wet season that started immediately before d4
increased the humidity of the soil in contact with the litterbags on d4
3.2. Above- and belowground biomass
and d5.
Regarding root litter decomposition, no significant effect of location
Pearl millet biomass was higher under the tree than far from the tree
was observed on any date (Supplementary Table 2); thus, data were
(Fig. 5). The difference was mainly noteworthy for the aboveground
compiled for both locations.
parts, resulting in a lower R:S ratio (0.03 compared to 0.05 far from the
After 15 months of the experiment, neither the crop nor the tree fine
tree). At both locations, millet roots were concentrated in the first 40 cm
roots reached an asymptote; therefore, we described decomposition
depth whereas under the tree, Faidherbia fine roots were concentrated
with 2 slopes k1 and k2 (linear fitting) rather than with one extinction
coefficient (exponential fitting). After 1.5 months of fine root decom­
position, i.e., d1, the root litterbags had lost almost half of their initial
dry mass; then, with the dry season, the remaining fine root mass
decreased more slowly from d1 to d5 and reached approximately 25% of
the initial mass at the end of the experiment (Fig. 7). The remaining fine
root mass on d2 (pearl millet only) was significantly impacted by depth
(F = 3.8, p-value = 4.8 × 10− 2), with a lower fine root remaining mass at
a depth of 20 cm than at 40 cm and a lower fine root remaining mass at a
depth of 90 cm than at 180 cm (Fig. 8a). On d5, the remaining mass was
significantly higher for Faidherbia than for cowpea fine roots (F = 3.9, p-
value = 3.5 × 10− 2, Fig. 8b).
The k1 coefficient of the first decomposition stage was significantly
impacted by the plant species (F = 3.9, p-value = 3.5 × 10− 2, data not
shown), with higher coefficients for cowpea and lower coefficients for
Faidherbia. The rate of fine root decomposition was also significantly
impacted by soil depth in the case of pearl millet (F = 7.4, p-value =
4.54 × 10− 3, Table 4), with lower values at 180 cm than at 20 cm depth.
The cowpea fine root decomposition rate also decreased with soil depth
but to a lesser extent than that of pearl millet (F = 7.7, p-value = 2.13 ×
10− 2, Table 4), while the fine root decomposition rate of the Faidherbia
tree was only slightly impacted by depth (Table 4).

3.4. Relationships between fine root decomposition, soil characteristics

and litter quality

The contribution of the main soil variables and the fine root
decomposition rate to differences among soil depths is represented by
the PCA (Fig. 9), which explained 50.5% of the dataset’s variability.
Individuals at each soil depth were well separated with no overlap be­
tween 95% confidence ellipses of three distinguished groups: 0–40 cm,
90 cm and 180 cm (Fig. 9, p-value = 2.20 × 10− 9 for Wilk’s test). The
variables that best explained the separation between soil depths were C:
N, sand, Olsen P and clay. The k1 coefficient increased with these var­
Fig. 5. Above- (a) and belowground (b) biomasses of pearl millet under (right) iables. These variables were not correlated (orthogonal) with k2, the soil
and far (left) from the Faidherbia tree according to organs (ears and grains, pH or silt content (Fig. 9). Therefore, the first axis of the PCA best
stems and leaves, stump) and soil depths (0–40, 40–100 and 100–200 cm). For described variables that correlated with k1, and the second axis vari­
each location, R:S ratios are indicated in italics. ables correlated with k2. Importantly, k1 and k2 were not correlated.

7
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Fig. 6. Comparison of fine (a) and medium (b) root biomasses of Faidherbia under (right) and far (left) from the tree according to soil depth (0–40, 40–100, 100–150
and 150–200 cm).

(Barthès et al., 2006; Tounkara et al., 2020). The C and N contents

Table 2
decreased with soil depth and were higher under than far from the tree,
Initial values of biochemical qualities of the fine root litter added to the litter­
as expected according to other agroforestry studies (Félix et al., 2018;
bags. Data are mean values ± standard deviation (n = 3). Significant differences
between root litter types were tested with one-way analyses of variance. ***, ** Nair et al., 2009b), especially with Faidherbia albida (Dilla et al., 2019),
and * indicate the significance of the impact of the studied effects on litter but these differences were surprisingly not significant (considering
quality with p-values < 0.001, 0.01, and 0.05, respectively. Letters indicate p-value > 0.05). In the 0–40 cm soil layer, the soil C:N ratio was higher
differences between the 3 types of litter. far from than that under the tree, whereas in the deeper layers, the
Initial values Statistics
opposite was true. These changes were associated with a large standard
deviation, probably due to the low N concentration. Because Faidherbia
Faidherbia Pearl Cowpea F- p-value
leaf litter was shown to release high amounts of nutrients, especially N,
millet value
during decomposition (Mubarak et al., 2008), the long-term effects of
Carbon fractions (% DM)
this litter may have contributed to a decrease in the surface soil C:N
Soluble 14.5 ± 2.7a 13.9 ± 27.1 ± 2.2 29.3 8.1 ×
fraction 2.2a b
10− 4
*** ratio, likely by increasing the bacterial pathway of decomposition
Cellulose 19.4 ± 2.4a 37.9 ± 21.3 ± 29.9 7.6 × (Rousk and Bååth 2007).
2.4 b 4.4a 10− 4
*** At a depth of 40–180 cm, the higher C:N ratio under the tree than far
Hemicellulose 14.2 ± 1.9a 27.4 ± 34.3 ± 9.8 8.9 1.6 × from it was also related to a lower soil mineral N content, while the yield
2.2 ab b
10− 2
*
of the pearl millet was almost 3 times higher under tree (Roupsard et al.,
Lignine 51.9 ± 4.1 b 20.8 ± 17.3 ± 82.6 4.3 ×
2.1a 4.4a 10− 5
*** 2020; Leroux et al., 2020). Due to higher pearl millet above- and
Elemental composition (% DM) belowground production, more mineral N could be taken up by the crop
C 45.4 ± 0.1c 40.9 ± 42.9 ± 0.8 37.9 4.0 × under the tree. However, with respect to pearl millet root distribution,
0.7a b
10− 4 ***
the main difference between the locations under and far from the tree
N 3.5 ± 0.1c 1.3 ± 1.4 ± 0.0 b 1779 4.8 ×
0.0a 10− 9 ***
occurred in the topsoil (0–40 cm), and millet did not invest biomass at
P 0.09 ± 0.0 0.07 ± 0.11 ± 12.0 8 .0 × great depth under the tree, which is in agreement with the higher root:
ab
0.01a 0.01 b 10− 3 ** total biomass ratio far from the tree. Another possible explanation is that
a
C:N 13.0 ± 0.3 32.5 ± 30.2 ± 0.3 981 2.8 × microorganisms may immobilize soil mineral N following an N-mining
0.9c b
10− 8 ***
strategy (Chen et al., 2014) to mineralize soil organic matter or plant
litter with a high C:N ratio. The total amounts of soil C and N were not
4. Discussion influenced by tree presence, suggesting that the nature of the litter
entering the soil instead of the soil organic matter (C and N) differed
4.1. Impact of Faidherbia albida trees on soil characteristics under and far from the tree and would be responsible for the hypo­
thetical N-mining strategy. We did not separate roots or aboveground
As expected for this type of soil, the total C and N contents were quite plant parts according to their location (under or far from the tree) before
low (less than 0.5 and 0.05%, respectively), as previously described assessing their C and N contents. However, the fine and medium roots of

8
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Table 3A
Differences in the volumetric humidity (m3H2O m− 3soil) of the soil in contact with the litterbags among soil depths (20, 40, 90 and 180 cm), plant species (pearl millet,
Faidherbia tree, cowpea), locations (far from and under the tree) and sampling dates (d1 to d5). Data are mean values ± standard deviation. The different lowercase
letters indicate significant differences between the modalities, and ns indicates the absence of a significant effect.
Soil volumetric humidity (m3H2O m− 3
soil) Statistics

F-value P-value

Soil depth (cm) 20 40 90 180

0.020 ± 0.018a 0.027 ± 0.024 b 0.045 ± 0.036 b 0.110 ± 0.051c 337.9 <2.2 × 10− 16

Root species Faidherbia Pearl millet Cowpea

0.060 ± 0.054 0.049 ± 0.046 0.051 ± 0.057 ns
Location Far Under
0.039 ± 0.036a 0.069 ± 0.061 b 103.5 4.4 × 10− 4

Sampling date d1 d2 d3 d4 d5
0.090 ± 0.059 d 0.058 ± 0.042c 0.026 ± 0.028a 0.042 ± 0.040 b 0.051 ± 0.053 bc
100.5 <2.2 × 10− 16

Table 3B
Differences in the volumetric humidity (m3H2O m− 3soil) of the soil in contact with the litterbags between each location (far from and under the tree) on each sampling
date (d1 to d5). Data are mean values ± standard deviation. The different lowercase letters indicate significant differences between the modalities, and ns indicates the
absence of a significant effect.
d1 d2 d3 d4 d5

Far Under Far Under Far Under Far Under Far Under

0.069 ± 0.113 ± 0.069 0.043 ± 0.070 ± 0.048 0.018 ± 0.035 ± 0.032 0.030 ± 0.019 0.051 ± 0.049 0.032 ± 0.072 ±
0.039a b
0.029a b
0.021a b
a b
0.033 0.063
F = 16.9 p-value = 1.41 × 10− 2 F = 12.6 p-value = 2.22 × 10− 2 F = 15.2 p-value = 1.72 × 10− 2 F = 14.6 p-value = 1.68 × 10− 2 ns

Fig. 7. Dynamics of root litter decomposition (after 1.5, 3, 6, 9 and 15 months) for the three plant species (Faidherbia tree (a), pearl millet (b) and cowpea (c)) at four
soil depths (20, 40, 90, 180 cm). The wet season is represented in blue, and the dry season is represented in yellow. Data are mean values, and error bars are standard
deviations (n = 6). Coefficients k1 and k2 are shown only on the left plot but were calculated for each regression. (For interpretation of the references to color in this
figure legend, the reader is referred to the Web version of this article.)

Faidherbia were particularly abundant under the tree below a depth of silvo-agro-pastoral system removing the leaves, twigs and fruits from the
40 cm and had a low C:N ratio of 13. The low root C:N ratio can be ground during the litterfall season (April to July) due to the reverse
explained by the high N availability in this N-fixing species, which phenology of this tree (Roupsard et al., 1999). However, ruminants tend
would prevent a lack of N and thus hamper the N-mining strategy. to stand under trees, where excrement is deposited, which enriches the
However, the presence of N-binding materials such as lignin and poly­ topsoil nutrient content under trees. At a depth of 50 cm, the lack of
phenols could restrict N accessibility and lead to a microbial N immo­ nutrients under trees compared to that far from trees could come from
bilization phase, as described during the leaf decomposition of a N-fixing the increase in Faidherbia root biomass at the same depth and thus the
tree by Teklay and Malmer (2004). Furthermore, the slow decomposi­ increase in nutrient uptake at depth compared to the topsoil.
tion of Faidherbia roots due to the high lignin content could favour the A significant interaction between the soil depth and location
development of K-strategy microorganisms dominated by fungi (Chen impacted the soil pH. Indeed, the higher soil pH under the tree compared
et al., 2014). However, without more information on the importance with that far from the tree occurred mostly below a depth of 40 cm,
and nature of the soil microbial communities in comparison to total soil while no significant differences were observed in the topsoil, as previ­
C, we cannot conclude the origin of the soil C:N changes. ously reported by Félix et al. (2018) for Piliostigma shrubs in Burkina
Several soil fertility indicators, such as mineral N and Olsen P, were Faso. In the 40–90 cm horizon, soil pH tended to increase under the tree
higher in the topsoil (0–40 cm) than at depth (40–200 cm) under the (from 6.7 to 7.3), as found by Rao et al. (1997), while acidification (from
tree. The enrichment of nutrients in the topsoil under the Faidherbia 6.5 to 5.7) was recorded far from the tree. Sandy soils are poorly buff­
albida tree was in agreement with the results found by Yengwe et al. ered (Wezel et al., 2000), and pH is sensitive to small variations in
(2018) and explained the higher crop yield under this tree. No remaining acid-basic reactions. Although our study did not allow us to conclude the
detritus of Faidherbia leaves was observed on the soil surface during the mechanisms to explain the increase in pH under the tree, acidification of
sampling period. This was due to the active livestock in this the soil profile far from the tree is an indicator of fertility degradation.

9
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Fig. 8. Final root litter remaining mass (a) on d2 (01/17/2019) for pearl millet at four soil depths (20, 40, 90 and 180 cm) and (b) on d5 (01/22/2020) for three
plant species (Faidherbia tree, pearl millet and cowpea) at all soil depths. Data are mean values, and error bars are standard deviations (n = 3). The different letters
indicate significant differences in remaining dry mass for each soil depth (a) or for each plant species (b).

(Faye et al., 2020), the reduction of water runoff under the tree crown
Table 4
(Lal 1989) and the potential benefit of hydraulic redistribution through
The k1 coefficient for each plant species (Faidherbia tree, pearl millet and
the Faidherbia root system (Bayala and Prieto 2020), we expected a
cowpea) according to soil depth (20, 40, 90, 180 cm). Data are mean values (n =
6). The different lowercase letters indicate significant differences between the slower fine root decomposition rate far from the tree than under the tree.
soil depths accompanied by their p-values for each plant species, and ns in­ This was not confirmed here. However, the lack of tree replicates may
dicates the absence of a significant effect of soil depth. bias our results, and the study would need to be extended to a wider area
of the park, including different tree sizes representing the local diversity
Soil depths (cm) Statistics
of the parkland.
20 40 90 180 F- P-value
Root litter quality was the main factor controlling the rate of
value
decomposition. Faidherbia albida roots decomposed more slowly than
Faidherbia 5.08 × – – 4.44 × ns cowpea roots due to less soluble compounds and high lignin contents, as
10− 3 10− 3
reported in Mubarak et al. (2012), while the root N content (higher in
Pearl 5.93 × 5.17 × 4.55 × 4.16 × 7.4 4.54 ×
millet 10− 3 b 10− 3a 10− 3a 10− 3a 10− 3 Faidherbia fine roots) did not seem to influence k1. Over a short period
Cowpea 6.53 × 5.75 × – 5.37 × 7.7 2.13 × of time, soluble C drives the decomposition of plant residues (Bertrand
10− 3 b 10− 3a 10− 3a 10− 2 et al. 2006, 2009; Moorhead et al., 2016; Liang et al., 2018), while the
litter N content (C:N ratio) has no impact unless N limits decomposition
(Recous et al., 1995; Bertrand et al., 2006), which does not seem to be
Acidic soils are indeed known for their relatively low microbial abun­
the case here.
dance and diversity and low cation exchange capacity (Robson 2012)
The remaining C after 15 months of decomposition accounted for
and could affect millet productivity.
root C inputs (Fig. 10) for 2 plant species (Faidherbia fine roots and pearl
millet), at both locations (under and far from the tree) and at 3 soil layers
4.2. Impact of soil depth on root litter decomposition (0–40, 40–100 and 100–200 cm of depth). The effect of all combined
factors was significant (F = 10.3, p-value = 5.4 × 10− 3, Supplementary
Due to the climatic conditions in the study area, root decomposition Table 2). Under the tree, Faidherbia root biomass was higher than far
occurred rapidly after crop harvest (end of October 2018), while the soil from the tree, and low decomposition rates of this perennial root litter
was still moist from the previous wet season, and lasted for two months were observed in the litterbags. Both of these factors resulted in higher C
thereafter (January 2019). Then, the soil dried progressively from 9.0 ± inputs from Faidherbia root litter under (0.62 gC m− 2 between 0 and
5.9 m3H2O m− 3soil on d1 to 2.6 ± 2.8 m3H2O m− 3soil on d3, as no rain 200 cm of depth, Fig. 10) than far from the tree (0.02 gC m− 2), which
occurred until the next wet season, which started in July 2019. Faster could explain the tendency of higher soil C stocks under than far from
decomposition in wetter soils confirmed a previous report by Duthoit the tree (Fig. 10, Table 1). Furthermore, the root C input was higher at
et al. (2020) regarding soil respiration. This moisture regime leads to depth than at the surface; between 100 and 200 cm depths, the amount
two contrasting kinetics of decomposition, with a relatively rapid first of remaining C after 15 months of decomposition originating from
phase (k1) and a slower second phase (k2), following the same time Faidherbia root litter under the tree was 7 times higher than that at 20
scale, similar to the few other studies conducted under similar envi­ cm. No significant C inputs from the Faidherbia root litter were note­
ronmental conditions (Mubarak et al. 2008, 2012). This result suggested worthy far from the tree. Due to a different root distribution, pearl millet
that the labile part of the root litter decomposed quickly during the first presented the opposite trend. Pearl millet root C inputs were signifi­
phase of decomposition (k1) when the soil was very wet. Then, the cantly higher at the soil surface than at depth, with no difference be­
decomposition slowed (k2) as the soil dried. tween the 2 locations (Fig. 10). The pearl millet crop provided 2.34 to
Soil moisture is a key factor controlling root decomposition and 2.44 gC m− 3 at 0–40 cm of depth through its roots. This amount is very
seems to be the main driver of decomposition kinetics after litter species, low compared to the soil C stocks at the same depth (4708 gC m− 3 far
i.e., quality (Arrouays et al., 2002; Butenschoen et al., 2011). Because from the tree and 6429 gC m− 3 under the tree), but it is repeated every
the humidity of the root litterbags was significantly higher for the in­ growing season. The role in the soil carbon stocks of pearl millet in
dividuals located under the tree due to tree shading, which reduced soil topsoil and of Faidherbia fine roots at depth was in agreement with that
evaporation (Hasselquist et al., 2018), greater soil water infiltration described by Jackson et al. (2017), attesting that fine roots contribute

10
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Fig. 9. Relationships between soil characteristics (clay, silt, sand, available P (as POlsen) and mineral N (as Nmin) contents, pH and the C:N ratio) and root
decomposition (k1 and k2), according to the soil depth (20, 40, 90, 180 cm).

Fig. 10. Potential root C contribution to soil C stocks in one cultural season for Faidherbia and pearl millet according to the soil depth and at two locations: under
(left) and far (right) from the tree. For each location, the different letters indicate significant differences in soil C stocks and in remaining C between the soil depths.

11
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

substantially to soil organic carbon storage. Future studies should Fondation TOTAL, as part of a sponsorship agreement. Among the
prospect deeper soil depths to take into account a more representative technical staff from Senegal, we are extremely indebted to Ibou Diouf
cross-section of the tree root system. We hypothesized that soil depth from Sob village for extensive help in the field and Patricia Moulin from
would slow the fine root decomposition rate due to reduced microbial LAMA/IESOL for conducting all soil analyses.
activity and moisture and temperature buffering, which was confirmed
for the first phase of decomposition (k1 was higher at 20 cm than at soil Appendix A. Supplementary data
depths of 40, 90 and 180 cm for the three species). We did not measure
microbial biomass C; however, several studies have reported its close Supplementary data to this article can be found online at https://doi.
relationships with the organic C content in soils (Insam and Domsch org/10.1016/j.jaridenv.2021.104696.
1988; Webster et al., 2001; Ng et al., 2014). In the topsoil, more abun­
dant microbial biomass and activity as well as drying/rewetting cycles References
that create a flush of C and microbial activity may explain the quicker
decomposition rates (Miller et al., 2005; Sun et al., 2013). According to Arrouays, D., Balesdent, J., Germon, J.-C., Jayet, P.-A., Soussana, J.-F., Stengel, P., 2002.
Contribution à la lutte contre l’effet de serre. Stocker du carbone dans les sols
the PCA, the soil characteristics that best explained k1 were sand, the agricoles de France?.
Olsen P content and the soil C:N ratio, suggesting that the very low Balesdent, J., Basile-Doelsch, I., Chadoeuf, J., Cornu, S., Derrien, D., Fekiacova, Z.,
amount of P may have decreased the microbial activity at depth. Hatté, C., 2018. Atmosphere–soil carbon transfer as a function of soil depth. Nature
559, 599–602. https://doi.org/10.1038/s41586-018-0328-3.
Balesdent, J., Recous, S., 1997. Les temps de résidence du carbone et le potentiel de
5. Conclusion stockage de carbone dans quelques sols cultivés français. Can. J. Soil Sci. 77 (2),
187–193. https://doi.org/10.4141/S96-109.
Barthès, B.G., Brunet, D., Ferrer, H., Chotte, J.-L., Feller, C., 2006. Determination of total
Root litter decomposition varied mostly according to soil depth, with carbon and nitrogen content in a range of tropical soils using near infrared
litter quality and soil moisture being the main factors related to the spectroscopy: influence of replication and sample grinding and drying. J. Near
decomposition coefficient k1 in the first 1.5 months. Organic C origi­ Infrared Spectrosc. 14 (5), 341–348. https://doi.org/10.1255/jnirs.686.
Battie-Laclau, P., Taschen, E., Plassard, C., Dezette, D., Abadie, J., Arnal, D.,
nating from roots would be stored for a longer time period at depth than
Benezech, P., Duthoit, M., Pablo, A.-L., Jourdan, C., Laclau, J.-P., Bertrand, I.,
in the topsoil. Furthermore, tree root litter tended to be more recalci­ Taudière, A., Hinsinger, P., 2020. Role of trees and herbaceous vegetation beneath
trant than annual crop root litter and was more abundant below 40 cm, trees in maintaining arbuscular mycorrhizal communities in temperate alley
while annual roots were concentrated in the topsoil. Therefore, slow tree cropping systems. Plant Soil 453 (1–2), 153–171. https://doi.org/10.1007/s11104-
019-04181-z.
root decomposition at depth could play a role in increasing belowground Bayala, J., Prieto, I., 2020. Water acquisition, sharing and redistribution by roots:
C inputs and sequestration. In contrast, pearl millet induced root C in­ applications to agroforestry systems. Plant Soil 453 (1–2), 17–28. https://doi.org/
puts mainly in the topsoil and it did not depend on the location. The root 10.1007/s11104-019-04173-z.
Bertrand, I., Chabbert, B., Kurek, B., Recous, S., 2006. Can the biochemical features and
decomposition rate was not affected by the location, but the tree fine histology of wheat residues explain their decomposition in soil? Plant Soil 281 (1–2),
root biomass and pearl millet vegetative production were higher under 291–307. https://doi.org/10.1007/s11104-005-4628-7.
the tree than far from the tree. This difference resulted in higher soil Bertrand, I., Prevot, M., Chabbert, B., 2009. Soil decomposition of wheat internodes of
different maturity stages: relative impact of the soluble and structural fractions.
carbon stocks under the tree than far from it. Bioresour. Technol. 100 (1), 155–163. https://doi.org/10.1016/j.
In agroforestry systems, the diversity of plant species induces a great biortech.2008.06.019.
diversity of root qualities and thus various decomposition kinetics. Blake, G.R., Hartge, K.H., 1986. Bulk density. In: Klute, A. (Ed.), SSSA Book Series. Soil
Science Society of America, American Society of Agronomy, Madison, WI, USA,
Introducing trees in arable lands would globally increase root litter in­ pp. 363–375.
puts while slowing root decomposition, especially at depth, and would Butenschoen, O., Scheu, S., Eisenhauer, N., 2011. Interactive effects of warming, soil
thus increase the soil carbon storage potential of the system. Further humidity and plant diversity on litter decomposition and microbial activity. Soil
Biol. Biochem. 43 (9), 1902–1907. https://doi.org/10.1016/j.soilbio.2011.05.011.
research should focus on this aspect with replicated trees and different
Cardinael, R., 2015. Stockage de carbone et dynamique des matières organiques des sols
distances from the trees according to a gradient to confirm the influence en agroforesterie sous climat méditerranéen et tempéré. Université Paris-Saclay
of tree presence on root decomposition kinetics. (ComUE). Thesis.
Cardinael, R., Mao, Z., Prieto, I., Stokes, A., Dupraz, C., Kim, J.H., Jourdan, C., 2015.
Competition with winter crops induces deeper rooting of walnut trees in a
Credit authorship contribution statement Mediterranean alley cropping agroforestry system. Plant Soil 391 (1–2), 219–235.
https://doi.org/10.1007/s11104-015-2422-8.
Siegwart Lorène: Data curation, Formal analysis, Writing – original Chander, K., Goyal, S., Nandal, D.P., Kapoor, K.K., 1998. Soil organic matter, microbial
biomass and enzyme activities in a tropical agroforestry system. Biol. Fertil. Soils 27
draft. Bertrand Isabelle: Conceptualization, Validation, Writing – re­ (2), 168–172. https://doi.org/10.1007/s003740050416.
view & editing, Supervision, Project administration, Funding acquisi­ Chen, R., Senbayram, M., Blagodatsky, S., Myachina, O., Dittert, K., Lin, X.,
tion. Roupsard Olivier: Writing – review & editing. Duthoit Maxime: Blagodatskaya, E., Kuzyakov, Y., 2014. Soil C and N availability determine the
priming effect: microbial N mining and stoichiometric decomposition theories.
Technical help. Jourdan Christophe: Conceptualization, Methodology, Global Change Biol. 20 (7), 2356–2367. https://doi.org/10.1111/gcb.12475.
Validation, Investigation, Writing – review & editing, Supervision. Chenu, C., Angers, D.A., Barré, P., Derrien, D., Arrouays, D., Balesdent, J., 2019.
Increasing organic stocks in agricultural soils: knowledge gaps and potential
innovations. Soil Tillage Res. 188, 41–52. https://doi.org/10.1016/j.
Declaration of competing interest still.2018.04.011.
Cotrufo, M.F., Wallenstein, M.D., Boot, C.M., Denef, K., Paul, E., 2013. The Microbial
The authors declare that they have no known competing financial Efficiency-Matrix Stabilization (MEMS) framework integrates plant litter
decomposition with soil organic matter stabilization: do labile plant inputs form
interests or personal relationships that could have appeared to influence stable soil organic matter? Global Change Biol. 19 (4), 988–995. https://doi.org/
the work reported in this paper. 10.1111/gcb.12113.
Diatta, S., 2021. Estimation du stock et du stockage annuel de carbone dans la biomasse
aérienne de Faidherbia albida du parc agroforestier de Sob (Observatoire de
Acknowledgements
Niakhar, Sénégal): approches par inventaires et dendrochronologie.
Dilla, A.M., Smethurst, P.J., Barry, K., Parsons, D., 2019. Preliminary estimate of carbon
We would like to thank both the EU-LEAP-Agri RAMSES II (ANR18- sequestration potential of Faidherbia albida (Delile) A.Chev in an agroforestry
LEAP-0003-02) and the DSCATT (N◦ AF1802-01, N◦ FTC002181) pro­ parkland in the Central Rift Valley of Ethiopia. For. Trees Livelihoods 28 (2), 79–89.
https://doi.org/10.1080/14728028.2018.1564146.
jects for the funding of the experimental setup at the “Faidherbia-Flux” Duthoit, M., Roupsard, O., Crequy, N., Sauze, J., Van Den Meersche, K., 2020.
platform of Sob, Niakhar district, Senegal (https://lped.info/wikiObs Conception d’un dispositif automatisé de chambres de mesures d’échanges gazeux
SN/?Faidherbia-Flux). The DSCATT project is financed by Agropolis du sol à fermeture horizontale. Cahier des Techniques de l’INRA.

Fondation on "Programme Investissement d’Avenir" Labex Agro (ANR-

10-LABX-0001-01) and, for activities led on African study sites, by

12
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

Ellert, B.H., Bettany, J.R., 1995. Calculation of organic matter and nutrients stored in sensing to assess the effect of trees on millet yield in complex parklands of Central
soils under contrasting management regimes. Can. J. Soil Sci. 75 (4), 529–538. Senegal. Agric. Syst. 184, 102918. https://doi.org/10.1016/j.agsy.2020.102918.
https://doi.org/10.4141/cjss95-075. Liang, Z., Elsgaard, L., Nicolaisen, M.H., Lyhne-Kjærbye, A., Olesen, J.E., 2018. Carbon
Faye, W., Fall, A.N., Orange, D., Do, F., Roupsard, O., Kane, A., 2020. Climatic variability mineralization and microbial activity in agricultural topsoil and subsoil as regulated
in the Sine-Saloum basin and its impacts on water resources: case of the Sob and by root nitrogen and recalcitrant carbon concentrations. Plant Soil 433 (1), 65–82.
Diohine watersheds in the region of Niakhar. Proc IAHS 383, 391–399. https://doi. https://doi.org/10.1007/s11104-018-3826-z.
org/10.5194/piahs-383-391-2020. Lin, B.B., 2007. Agroforestry management as an adaptive strategy against potential
Félix, G.F., Clermont-Dauphin, C., Hien, E., Groot, J.C.J., Penche, A., Barthès, B.G., microclimate extremes in coffee agriculture. Agric. For. Meteorol. 144 (1), 85–94.
Manlay, R.J., Tittonell, P., Cournac, L., 2018. Ramial wood amendments ( Piliostigma https://doi.org/10.1016/j.agrformet.2006.12.009.
reticulatum ) mitigate degradation of tropical soils but do not replenish nutrient Liu, C., Jin, Y., Hu, Y., Tang, J., Xiong, Q., Xu, M., Bibi, F., Beng, K.C., 2019. Drivers of
exports. Land Degrad. Dev. 29 (8), 2694–2706. https://doi.org/10.1002/ldr.3033. soil bacterial community structure and diversity in tropical agroforestry systems.
Feller, C., Beare, M.H., 1997. Physical control of soil organic matter dynamics in the Agric. Ecosyst. Environ. 278, 24–34. https://doi.org/10.1016/j.agee.2019.03.015.
tropics. Geoderma 79 (1), 69–116. https://doi.org/10.1016/S0016-7061(97)00039- Machinet, G.E., Bertrand, I., Chabbert, B., Recous, S., 2009. Decomposition in soil and
6. chemical changes of maize roots with genetic variations affecting cell wall quality.
Freschet, G.T., Cornwell, W.K., Wardle, D.A., Elumeeva, T.G., Liu, W., Jackson, B.G., Eur. J. Soil Sci. 60 (2), 176–185. https://doi.org/10.1111/j.1365-2389.2008.01109.
Onipchenko, V.G., Soudzilovskaia, N.A., Tao, J., Cornelissen, J.H.C., 2013. Linking x.
litter decomposition of above- and below-ground organs to plant-soil feedbacks Maignien, R., 1965. Notice explicative: carte pédologique du Sénégal au 1/1.000.000,
worldwide. J. Ecol. 101 (4), 943–952. https://doi.org/10.1111/1365-2745.12092. p. 71.
Fujisaki, K., Chevallier, T., Chapuis-Lardy, L., Albrecht, A., Razafimbelo, T., Masse, D., Makkonen, M., Berg, M.P., Handa, I.T., Hättenschwiler, S., Ruijven, J van, Bodegom, PM
Ndour, Y.B., Chotte, J.-L., 2018. Soil carbon stock changes in tropical croplands are van, Aerts, R., 2012. Highly consistent effects of plant litter identity and functional
mainly driven by carbon inputs: a synthesis. Agric. Ecosyst. Environ. 259, 147–158. traits on decomposition across a latitudinal gradient. Ecol. Lett. 15 (9), 1033–1041.
https://doi.org/10.1016/j.agee.2017.12.008. https://doi.org/10.1111/j.1461-0248.2012.01826.x.
Germon, A., Cardinael, R., Prieto, I., Mao, Z., Kim, J., Stokes, A., Dupraz, C., Laclau, J.-P., Mary, B., Beaudoin, N., Justes, E., Machet, J.M., 1999. Calculation of nitrogen
Jourdan, C., 2016. Unexpected phenology and lifespan of shallow and deep fine mineralization and leaching in fallow soil using a simple dynamic model. Eur. J. Soil
roots of walnut trees grown in a silvoarable Mediterranean agroforestry system. Sci. 50 (4), 549–566. https://doi.org/10.1046/j.1365-2389.1999.00264.x.
Plant Soil 401 (1–2), 409–426. https://doi.org/10.1007/s11104-015-2753-5. Matejovic, I., 1997. Determination of carbon and nitrogen in samples of various soils by
Goering, H.K., Van Soest, P.J., 1970. Forage Fiber Analyses: Apparatus, Reagents, the dry combustion. Commun. Soil Sci. Plant Anal. 28 (17–18), 1499–1511. https://
Procedures, and Some Applications. doi.org/10.1080/00103629709369892.
Guillot, E, Bertrand, I, Rumpel, C, Gomez, C, Arnal, D, Abadie, J, Hinsinger, P, 2021. Miller, A., Schimel, J., Meixner, T., Sickman, J., Melack, J., 2005. Episodic rewetting
Spatial heterogeneity of soil quality within a Mediterranean alley cropping enhances carbon and nitrogen release from chaparral soils. Soil Biol. Biochem. 37
agroforestry system: comparison with a monocropping system. Eur. J. Soil Biol. 105, (12), 2195–2204. https://doi.org/10.1016/j.soilbio.2005.03.021.
103330. https://doi.org/10.1016/j.ejsobi.2021.103330. Monteith, J.L., Ong, C.K., Corlett, J.E., 1991. Microclimatic interactions in agroforestry
Guenet, B, Eglin, T, Vasilyeva, N, Peylin, P, Ciais, P, Chenu, C, 2013. The relative systems. For. Ecol. Manag. 45 (1), 31–44. https://doi.org/10.1016/0378-1127(91)
importance of decomposition and transport mechanisms in accounting for soil 90204-9.
organic carbon profiles. Biogeosciences 10, 2379–2392. https://doi.org/10.5194/ Moorhead, D.L., Sinsabaugh, R.L., Hill, B.H., Weintraub, M.N., 2016. Vector analysis of
bg-10-2379-2013. ecoenzyme activities reveal constraints on coupled C, N and P dynamics. Soil Biol.
Handa, I.T., Aerts, R., Berendse, F., Berg, M.P., Bruder, A., Butenschoen, O., Chauvet, E., Biochem. 93, 1–7. https://doi.org/10.1016/j.soilbio.2015.10.019.
Gessner, M.O., Jabiol, J., Makkonen, M., McKie, B.G., Malmqvist, B., Peeters, E.T.H. Mubarak, A.R., Abdalla, M.H., Nortcliff, S., 2012. Millet (Pennisetum typhoides) yield
M., Scheu, S., Schmid, B., van Ruijven, J., Vos, V.C.A., Hättenschwiler, S., 2014. and selected soil attributes as influenced by some tree types of the semi-arid tropics
Consequences of biodiversity loss for litter decomposition across biomes. Nature of Sudan. J. Arid Environ. 77, 96–102. https://doi.org/10.1016/j.
509, 218–221. https://doi.org/10.1038/nature13247. jaridenv.2011.10.008.
Hasselquist, N.J., Benegas, L., Roupsard, O., Malmer, A., Ilstedt, U., 2018. Canopy cover Mubarak, A.R., Elbashir, A.A., Elamin, L.A., Daldoum, D.M.A., Steffens, D., Benckiser, G.,
effects on local soil water dynamics in a tropical agroforestry system: evaporation 2008. Decomposition and nutrient release from litter fall in the semi-arid tropics of
drives soil water isotopic enrichment. Hydrol. Process. 32 (8), 994–1004. https:// Sudan. Commun. Soil Sci. Plant Anal. 39 (15–16), 2359–2377. https://doi.org/
doi.org/10.1002/hyp.11482. 10.1080/00103620802292574.
Herman, D.J., Firestone, M.K., Nuccio, E., Hodge, A., 2012. Interactions between an Nair, P.K.R., Kumar, B.M., Nair, V.D., 2009a. Agroforestry as a strategy for carbon
arbuscular mycorrhizal fungus and a soil microbial community mediating litter sequestration. J. Plant Nutr. Soil Sci. 172 (1), 10–23. https://doi.org/10.1002/
decomposition. FEMS Microbiol. Ecol. 80 (1), 236–247. https://doi.org/10.1111/ jpln.200800030.
j.1574-6941.2011.01292.x. Nair, P.K.R., Nair, V.D., Kumar, B.M., Haile, S.G., 2009b. Soil carbon sequestration in
Hicks Pries, C.E., Sulman, B.N., West, C., O’Neill, C., Poppleton, E., Porras, R.C., tropical agroforestry systems: a feasibility appraisal. Environ. Sci. Pol. 12 (8),
Castanha, C., Zhu, B., Wiedemeier, D.B., Torn, M.S., 2018. Root litter decomposition 1099–1111. https://doi.org/10.1016/j.envsci.2009.01.010.
slows with soil depth. Soil Biol. Biochem. 125, 103–114. https://doi.org/10.1016/j. Ndiaye, O., Le Hesran, J.-Y., Etard, J.-F., Diallo, A., Simondon, F., Neil Ward, M.,
soilbio.2018.07.002. Robert, V., 2001. Variations climatiques et mortalité attribuée au paludisme dans
Insam, H., Domsch, K.H., 1988. Relationship between soil organic carbon and microbial zone de Niakhar, Sénégal, de 1984 à 1996. Cahiers santé 11, 25–33.
biomass on chronosequences of reclamation sites. Microb. Ecol. 15 (2), 177–188. Ng, E.-L., Patti, A.F., Rose, M.T., Schefe, C.R., Wilkinson, K., Smernik, R.J., Cavagnaro, T.
https://doi.org/10.1007/BF02011711. R., 2014. Does the chemical nature of soil carbon drive the structure and functioning
IUSS Working group WRB, 2014. World Reference Base for Soil Resources 2014: of soil microbial communities? Soil Biol. Biochem. 70, 54–61. https://doi.org/
International Soil Classification System for Naming Soils and Creating Legends for 10.1016/j.soilbio.2013.12.004.
Soil Maps. FAO, Rome. Oelbermann, M., Paul Voroney, R., Gordon, A.M., 2004. Carbon sequestration in tropical
Jackson, R.B., Lajtha, K., Crow, S.E., Hugelius, G., Kramer, M.G., Piñeiro, G., 2017. The and temperate agroforestry systems: a review with examples from Costa Rica and
ecology of soil carbon: pools, vulnerabilities, and biotic and abiotic controls. Annu. southern Canada. Agric. Ecosyst. Environ. 104 (3), 359–377. https://doi.org/
Rev. Ecol. Evol. Syst. 48 (1), 419–445. https://doi.org/10.1146/annurev-ecolsys- 10.1016/j.agee.2004.04.001.
112414-054234. Oelbermann, M., Voroney, R.P., 2007. Carbon and nitrogen in a temperate agroforestry
Jha, P., Prasad Mohapatra, K., 2010. Leaf litterfall, fine root production and turnover in system: using stable isotopes as a tool to understand soil dynamics. Ecol. Eng. 29 (4),
four major tree species of the semi-arid region of India. Plant Soil 326 (1–2), 342–349. https://doi.org/10.1016/j.ecoleng.2006.09.014.
481–491. https://doi.org/10.1007/s11104-009-0027-9. Ohno, T., Zibilske, L.M., 1991. Determination of low concentrations of phosphorus in soil
Koenig, R., Johnson, C., 1942. Colorimetric determination of phosphorus in biological extracts using malachite green. Soil Sci. Soc. Am. J. 55 (3), 892–895. https://doi.
materials. Ind. Eng. Chem. Anal. Ed. 14 (2), 155–156. https://doi.org/10.1021/ org/10.2136/sssaj1991.03615995005500030046x.
i560102a026. Paustian, K., Lehmann, J., Ogle, S., Reay, D., Robertson, G.P., Smith, P., 2016. Climate-
Kuzyakov, Y., Domanski, G., 2000. Carbon input by plants into the soil. Review. J. Plant smart soils. Nature 532 (7597), 49–57. https://doi.org/10.1038/nature17174.
Nutr. Soil Sci. 163 (4), 421–431. https://doi.org/10.1002/1522-2624, 200008)163: Poeplau, C., Vos, C., Don, A., 2017. Soil organic carbon stocks are systematically
4<421::AID-JPLN421>3.0.CO;2-R. overestimated by misuse of the parameters bulk density and rock fragment content.
Lagerlöf, J., Adolfsson, L., Börjesson, G., Ehlers, K., Vinyoles, G.P., Sundh, I., 2014. Land- Soil 3 (1), 61–66. https://doi.org/10.5194/soil-3-61-2017.
use intensification and agroforestry in the Kenyan highland: impacts on soil Pries, C.E.H., Sulman, B.N., West, C., O’Neill, C., Poppleton, E., Porras, R.C.,
microbial community composition and functional capacity. Appl. Soil Ecol. 82, Castanha, C., Zhu, B., Wiedemeier, D.B., Torn, M.S., 2018. Root litter decomposition
93–99. https://doi.org/10.1016/j.apsoil.2014.05.015. slows with soil depth. Soil Biol. Biochem. 125, 103–114. https://doi.org/10.1016/j.
Lal, R., 1989. Agroforestry systems and soil surface management of a tropical alfisol: I: soilbio.2018.07.002.
soil moisture and crop yields. Agrofor. Syst. 8 (1), 7–29. https://doi.org/10.1007/ Prieto, I., Stokes, A., Roumet, C., 2016. Root functional parameters predict fine root
BF00159066. decomposability at the community level. J. Ecol. 104 (3), 725–733. https://doi.org/
Lalou, R., Sultan, B., Muller, B., Ndonky, A., 2019. Does climate opportunity facilitate 10.1111/1365-2745.12537.
smallholder farmers’ adaptive capacity in the Sahel? Palgrave Commun. 5 (1), 1–11. R Core Team, 2020. R: A Language and Environment for Statistical Computing. R
https://doi.org/10.1057/s41599-019-0288-8. Foundation for Statistical Computing, Vienna, Austria.
Leaky, R., 1996. Definition of agroforestry revisited. Agroforest. Today 8, 5–7. Rahimi, J., Ago, E.E., Ayantunde, A., Berger, S., Bogaert, J., Butterbach-Bahl, K.,
Leroux, L., Falconnier, G.N., Diouf, A.A., Ndao, B., Gbodjo, J.E., Tall, L., Balde, A.A., Cappelaere, B., Demarty, J., Diouf, A.A., Falk, U., Haas, E., Hiernaux, P., Kraus, D.,
Clermont-Dauphin, C., Bégué, A., Affholder, F., Roupsard, O., 2020. Using remote Roupsard, O., Scheer, C., Srivastava, A.K., Tagesson, T., Grote, R., 2021. Modelling
gas exchange and biomass production in west African Sahelian and Sudanian

13
L. Siegwart et al. Journal of Arid Environments 198 (2022) 104696

ecological Zones. Geosci. Model Dev. Discuss. (GMDD) 1–39. https://doi.org/ Smith, P., Bustamante, M., Ahammad, H., Clark, H., Dong, H., Elsiddig, E.A., Haberl, H.,
10.5194/gmd-2020-417. Harper, R., House, J., Jafari, M., Masera, M., 2014. Climate Change 2014 Mitigation
Rao, M.R., Nair, P.K.R., Ong, C.K., 1997. Biophysical interactions in tropical agroforestry of Climate Change: Working Group III Contribution to the Fifth Assessment Report of
systems. Agrofor. Syst. 38 (1/3), 3–50. https://doi.org/10.1023/A:1005971525590. the Intergovernmental Panel on Climate Change. Cambridge University Press,
Rasse, D.P., Rumpel, C., Dignac, M.-F., 2005. Is soil carbon mostly root carbon? Cambridge, United Kingdom and New York, NY, USA.
Mechanisms for a specific stabilisation. Plant Soil 269 (1), 341–356. https://doi.org/ Sun, Y., Yan, F., Liu, F., 2013. Drying/rewetting cycles of the soil under alternate partial
10.1007/s11104-004-0907-y. root-zone drying irrigation reduce carbon and nitrogen retention in the soil–plant
Recous, S., Robin, D., Darwis, D., Mary, B., 1995. Soil inorganic N availability: effect on systems of potato. Agric. Water Manag. 128, 85–91. https://doi.org/10.1016/j.
maize residue decomposition. Soil Biol. Biochem. 27 (12), 1529–1538. https://doi. agwat.2013.06.015.
org/10.1016/0038-0717(95)00096-W. Teklay, T., Malmer, A., 2004. Decomposition of leaves from two indigenous trees of
Robson, A., 2012. Soil Acidity and Plant Growth. Elsevier. contrasting qualities under shaded-coffee and agricultural land-uses during the dry
Roupsard, O., Audebert, A., Ndour, A.P., Clermont-Dauphin, C., Agbohessou, Y., season at Wondo Genet, Ethiopia. Soil Biol. Biochem. 36 (5), 777–786. https://doi.
Sanou, J., Koala, J., Faye, E., Sambakhe, D., Jourdan, C., le Maire, G., Tall, L., org/10.1016/j.soilbio.2003.12.013.
Sanogo, D., Seghieri, J., Cournac, L., Leroux, L., 2020. How far does the tree affect Tounkara, A, Clermont-Dauphin, C, Affholder, F, Ndiaye, S, Masse, D, Cournac, L, 2020.
the crop in agroforestry? New spatial analysis methods in a Faidherbia parkland. Inorganic fertilizer use efficiency of millet crop increased with organic fertilizer
Agric. Ecosyst. Environ. 296, 106928. https://doi.org/10.1016/j.agee.2020.106928. application in rainfed agriculture on smallholdings in central Senegal. Agric. Ecosyst.
Roupsard, O., Ferhi, A., Granier, A., Pallo, F., Depommier, D., Mallet, B., Joly, H.I., Environ. 294, 106878. https://doi.org/10.1016/j.agee.2020.106878.
Dreyer, E., 1999. Reverse phenology and dry-season water uptake by Faidherbia Webster, E.A., Hopkins, D.W., Chudek, J.A., Haslam, S.F.I., Šimek, M., Pîcek, T., 2001.
albida (Del.) A. Chev. in an agroforestry parkland of Sudanese west Africa. Funct. The relationship between microbial carbon and the resource quality of soil carbon.
Ecol. 13 (4), 460–472. https://doi.org/10.1046/j.1365-2435.1999.00345.x. J. Environ. Qual. 30 (1), 147–150. https://doi.org/10.2134/jeq2001.301147x.
Rousk, J., Bååth, E., 2007. Fungal and bacterial growth in soil with plant materials of Wezel, A., Rajot, J.-L., Herbrig, C., 2000. Influence of shrubs on soil characteristics and
different C/N ratios. FEMS Microbiol. Ecol. 62 (3), 258–267. https://doi.org/ their function in Sahelian agro-ecosystems in semi-arid Niger. J. Arid Environ. 44
10.1111/j.1574-6941.2007.00398.x. (4), 383–398. https://doi.org/10.1006/jare.1999.0609.
Six, J., Feller, C., Denef, K., Ogle, S., , Joao Carlos De Moraes Sa, Albrecht, A., 2002. Soil Yengwe, J., Gebremikael, M.T., Buchan, D., Lungu, O., De Neve, S., 2018. Effects of
organic matter, biota and aggregation in temperate and tropical soils - effects of no- Faidherbia albida canopy and leaf litter on soil microbial communities and nitrogen
tillage. Agronomie 22 (7–8), 755–775. https://doi.org/10.1051/agro:2002043. mineralization in selected Zambian soils. Agrofor. Syst. 92 (2), 349–363. https://doi.
org/10.1007/s10457-016-0063-4.

14