Increases in decomposition of soil organic matter resulting from global warming
or from land use change could significantly increase the atmospheric burden
of CO2, which would further enhance the greenhouse effect. Inventories
of Soil Organic Carbon (SOC) stocks at national scale are needed in the context
of the Framework Convention on Climate Change (UNFCCC) (Smith,
2004; Smith et al., 2008). In this study,
the Kyoto Protocol allows carbon dioxide emissions to be offset by demonstrable
removal of carbon from the atmosphere, by improved management of agricultural
soils. In order to use this possibility, the first step is the knowledge of
the SOC and how to calculate this SOC stocks. A good estimation of carbon pools
in the soils has been suggested as a means to help mitigate atmospheric CO2
increases and anticipated changes in climate (Batjes, 1999;
Lal et al., 1998, 2000;
Bernoux et al., 2002).
The soil is a key component of the global carbon cycle. In the world, soils
compartment represent a large reservoir of carbon, with estimates ranging from
1500 to 2000 Pg C in the upper 100 cm (Post et al.,
1982; Eswaran et al., 1993; Batjes,
1996; IPCC, 2001). SOC stocks may be very sensitive
to climate change, having a negative feedback which could enhance global warming.
The soils of the world are thought to store three times more organic carbon
than is held in the plant biomass of terrestrial ecosystems (650 Pg) and about
twice as much than is current in the atmosphere (750 Pg) (Eswaran
et al., 1993; Kimble et al., 1990;
Post et al., 1982; Batjes
and Sombroek, 1997). Regional and global estimates of soil carbon stocks
had to be made by extrapolating means of soil carbon content for broad categories
of types of soils or vegetation across the areas occupied by those categories
(Kimble et al., 1990; Batjes,
1996; Bernoux et al., 2002). The soil compartment,
global carbon pools are difficult to estimate because of still limited knowledge
about specific properties of soil types (Sombroek et
al., 1993; Batjes, 1996), the high spatial variability
of soil carbon even within one soil map unit (Cerri et
al., 2000) and the different effects of the factors controlling the
soil organic carbon cycle (Pastor and Post, 1986; Parton
et al., 1987). Thus, regional studies are necessary to refine global
estimates, mainly at country scale (Bernoux et al.,
Organic carbon storage in Tunisian soils reflects capacity that arid and semi-arid
regions to sequester carbon (Brahim et al., 2010).
The importance of an understanding of the national organic carbon pool levels
is reinforced by the statements of the United Nations Framework Convention on
Climate Change (UNFCCC) signed at Rio de Janeiro in 1992. In fact, the UNFCCC
aims to stabilize greenhouse gas concentrations in the atmosphere at a level
that limits adverse impacts on the global warming. Potential mechanisms cover
emission reductions and activities that increase carbon sinks, including terrestrial
sinks (Smith, 2004).
The objective of this study is to assess and given consistent values and a distribution maps, for the 0 to 30 cm and 0 to 100 cm depth of the organic carbon stocks in the soils of Tunisia, by governorate and by delegation. The aim of this study is to provide a valuable baseline data for evaluating the effect of soil occupation and climatic region for Tunisian SOC stocks.
MATERIALS AND METHODS
Study area: This study was conducted for all Tunisian soils during the
period 2007 to 2009. Tunisia (31°38°N; 7°12°E), situated in
north of Africa and south of Mediterranean Sea (Fig. 1) and
it covered an area of 164,000 km2, a wide range of natural regions.
Three dominant climatic zones illustrate the country and reflect influence by
Sea and Sahara desert: (1) Northern region is humid (600-1200 mm year-1)
occupied by rainforest still; (2) Central region is semi-arid (200-600 mm year-1)
steppe is here dominant vegetation; (3) Southern region is arid its a
Desert (<200 mm year-1). From north to south Tunisians region,
soils vary widely, due to variability in climate, vegetation, parent material,
texture, structure and anthropic effect (Brahim et al.,
2010). Nine big orders of soils have been inventoried (Table
Database: A database was built from previous analytical results from soil profile information for soils pits surveyed by Tunisian research groups and the IRD (ex-ORSTOM) project, the Ministry of Agriculture of Tunisia and Tunisian thesis reports. The data contained information on organic carbon in soil (fraction <2 mm; walkley-black method), pH (measured in water 1:1), bulk density (Db) (Cylindre method; Mg m-3), granulometric fraction (after dispersion with sodium hexametaphosphate of soils), Clay (particle 0-2 μm), Silt (fine and coarse 2-50 μm), Sand (fine and coarse; 50-2000 μm) and CaCO3 (Carbonate of calcium measured with Bernard calcimeter method).
||Location of Tunisia in the Mediterranean Sea and semi-arid
||Soil categories and their relation to the original soil classes
of soil map
Elaboration of soil-department association map: The soil-department association maps are obtained by association soil map and the map of departmental divisions in the country. In this study, the mean departmental maps, the map of governorates and the map of delegations. Therefore, you find soil-governorate association (SGA) map and soil-delegation association (SDA) map.
Soil map: The original soil map 1/500 000 in Fig. 2a
(Belkhodja et al., 1973) was built up from 35007
map units. We used it in this study the term map unit: MU definition given by
Bernoux et al. (2002), that is a single-part
polygon of a digital map. These MU were split into nine soil groups. We made
S10 code when the MU is water and urban soil. The Tunisian soil map showed that
Luvisols is the lowest soil order area within the country it covers alone 0.38%
of the total area, although, Lithosols and Regosols are the biggest soil orders
covering 25.63 and 24.50% from total area, respectively. The list of available
soil groups and here codes used for analysis are described in Table
Departmental maps: This is the map of governorates and the map of delegations
which was built from. (1) Governorate map with 27 MUs and (2) Delegation
map which was built from 264 MUs.
||Soil, governorate and delegation maps, with designation (a),
(b) and (c), respectively, used in elaboration of soil-department association
|| Tunisian governorates and their codes
The soil governorate and delegation maps are illustrated in Fig.
Soil-Governorate Association map (SGA): The SGA map was derived by intersection of the soil and governorate maps. In Tunisia, a governorate is an administrative division. It is the equivalent of a state or province in other parts of the World. Each MU of the output map was characterized by combining the information derived from the soil (9 categories in Table 1) and governorate (24 governorates in Table 2) maps. The MUs of the SGA map that corresponded to a MU characterized as lagoon or sebkha or urban zone in the soil map, were classified as water and urban soil (S10).
Soil-Delegation Association map (SDA): The SDA map was derived by intersection of the soil and delegation maps. The Delegations of Tunisia are the second level administrative division. The delegation is the equivalent of sector, it is the smallest division. The 24 Tunisian governorates were divided into 264 delegations. Every MUs of the output map was characterized by combining the information derived from the soil (9 categories in Table 1) and delegation maps (262 delegations, Fig. 3). The MUs of the SDA map that corresponded to a MU characterization as a lagoon or urban zone in the soil map were classified as water and urban soil.
Bulk density: Bulk density is not determined in most routine analyses,
so values have to be determined using Pedotransfer Functions (PTF) (Batjes,
1996; Bernoux et al., 2002; Brahim
et al., 2010).
Calculation of the individual organic carbon stocks
Classical way: profile by profile: In most studies (Eswaran
et al., 1993; Batjes, 1996) SOC stocks has
been calculated to a depth of 30 and 100 cm. To estimate SOC stocks, requires
knowledge of the vertical distribution of organic carbon in profiles (Bernoux
et al., 1998). Calculation of SOC stocks traditionally report results
to 1 m depth (Eswaran et al., 1993; Batjes,
1996; Brahim et al., 2010) or the surface
layer (0-30 cm) (Bernoux et al., 2002; Brahim
et al., 2010) level in direct contact (exchange) with atmospheric
gas. Calculations of the SOC 0-30 cm in the classical way for a given depth
consists of addition of SOC Stocks by horizon and multiplied by Db
and the organic carbon concentration and horizon thickness.
|| Tunisian governorates and their delegations area
For an individual profile with n layers, we estimated the organic carbon stock
by the following equation:
where, SOC is the organic carbon stock (kg C m-2), Dbi is the bulk density (Mg m-3) of layer i, Ci is the proportion of organic carbon (g C g-1) in layer i, Di is the thickness of this layer (cm).
Predicted organic carbon: However, when using large database at continental
scales, problems of interpolation appear, if not all horizons are sampled or
analyzed and the sampling depth crosses a horizon (Bernoux
et al., 1998). In present study, distribution from vertical organic
carbon profile was calculated using Arrouays and Pelissier
(1994) model. With this model, we have obtained several missing values from
organic carbon in profiles. Usually, organic carbon content in soils decline
progressively with depth over the entire profile and was fit by an exponential
where, X and C(X) are the depth and the C the carbon content, X1 and C1 the depth and the carbon content of a fixed upper position and X2 and C2 depth and carbon content for a fixed deeper position.
The carbon content used in these equations can be expressed in different ways: weight percentage of the <2 mm fraction, or weight per volume. However, this way of expressing results requires knowledge of the bulk density (Db) and the Db of most soil samples is usually not determined.
Large amounts of organic carbon, which are not yet considered in most global
carbon budget (Eswaran et al., 1993; Batjes,
1996), are stored between depth of 100 and 200 cm. Much of this deeper carbon
occurs in fairly stable forms and therefore will not contribute much to current
gaseous emissions. In most studies, a SOC stock has been calculated to a depth
of 100 cm. The 30 cm depth that is most directly involved in interaction with
the atmosphere and that is most sensitive to land use and environmental changes.
A metre depth was the lower limit of biological activity in arid and semi-arid
regions (Singh et al., 2007).
SGA and SDA maps organization: SGA map: We elaborate SGA map after intersection
of the soil and governorate maps. Theoretically the new SGA map contains 264
possible cases from MUs. In the map there is the total of 23160 MUs.
The surface covered by each SGA category is given in Table 3.
The largest SGA category (18551 km2) corresponded to the S1 G24 association,
but the smallest (<1 km2), we make zero (0) value, when SGA category
it inferior at 1 km2, for example; S4 G2, S6 G2, S9 G19, in totality
we have 10 cases SGA <1 km2.
|| Number (N) of map units and corresponding area (km2)
of the SGA categories
|*Categories not represented in the digital SGA map are indicated
More than 84% of the SGA categories (223 of the total) had an area smaller
than 1000 km2 and covered 41 171 km2 (26.5% from Tunisian
area), but 16% of the SGA categories (41 of the total) had an area >1000
km2 covering 114 131 km2 (73.5% from Tunisian area).
SDA map: We find this new map with intersection of soil by delegation maps. It totalled 41759 MU. The MU was spread into 2882 theoretically possible cases, 11 soil categories or groups crossed with 262 delegations.
Soil profile database representatives: Only 1572 soil horizons of the total (5024 soil horizons) were sufficiently documented to permit the organic carbon stocks calculations to 0-30 cm and 0-100 cm. First, the representativeness of the information contained in the soil profile database was analyzed. For that, the number of individual organic carbon stocks calculated by soil groups (Table 4), SGA and SDA categories was examined.
Table 4 showed that in 0-30 cm depth, Luvisols (S7) have
the biggest organic carbon stocks 7.16 kg C m-2 and Lithosols (S1)
has the lowest organic carbon stocks 1.84 kg C m-2. Mean SOC content
in the upper 100 cm of the various Tunisian soils ranged from 4.04 kg C m-2
for Lithosols to 15.92 kg C m-2 for Luvisols.
|| Organic carbon stock by soil
|*Average values are expressed in kg C m-2
The large values for the latter are due to the abundance of organic matter in north-western Tunisian zone. Similarly large amounts 13.88 kg C m-2 is encountered in Podzoluvisols, because they are situated under forest with an environment rich of organic matter. Small amounts of organic carbon are encountered in Lithosols and Solonchaks from the semi-arid and arid regions where vegetation is limited.
Changes in the relative distribution of soil organic carbon stocks with depth have been observed in Table 4, the ratio of SOC of 0-30 cm divided by that in the 0-100 cm zone. On average, 37.54 to 45.54% of the total SOC in the upper 100 cm of mineral soil is held in the first 30 cm. these figures illustrate the potentially large amounts of CO2 that can be released when soils are deforested or with changes in land use.
When, SGA/SDA categories exist in the SGA/SDA maps and is not a Water and urban soil category, it was decided to use the mean of organic carbon stocks with SGA/SDA categories adjacent. In case those SGA/SDA categories dont exist we make the value of the means by soil groups (Table 4).
Soil organic carbon stocks values: It could be noted that for some SGA/SDA
categories, which are many associations are not represented in the digital SGA/SDA
maps. From SGA categories, we have 216 possible association cases (24 governorates
multiply with 9 soil groups), however from the SGA map we have only 96 association
cases. As well as SGA, we have 2358 possible association cases from SDA categories
(262 delegations multiply with 9 soil groups), but from the SDA map we have
only 213 association cases. Generally, at routine study fewer samples were taken
from the deeper layers (0-100 cm) than from the superficial layers (0-30 cm).
This difference in the number of samples at the various depths must be kept
in mind, as it implies that the results are less reliable for the deeper layers.
Additionally, sampling soil profiles is technically difficult and sample preparation
for analysis is time consuming.
||The average stock for each soil categories according to its
distribution by governorates
Thus most studies on soil carbon were restricted to the upper 30 to 50 cm of
the soil and only few include deeper sections of the soil cover.
In the level 0-30 cm, the organic carbon stocks ranged from 0.77 kg C m-2 (S2, Regosols-G21, Kebili governorate) to 13.61 kg C m-2 (S7, Luvisols-G5, Nabeul governorate). More than 34% of all SGA categories was associated with organic carbon stocks from 0.77 to 2.99 kg C m-2, 46% was associated with 3.05 to 5.75 kg C m-2 and 19% of the extent covered by SGA showed organic carbon stocks ranging from 6.02 to 13.61 kg C m-2. In 0-100 cm, the biggest organic carbon stocks in SGA categories it 27.29 kg C m-2 and corresponded to the S7, Luvisols-G5, Nabeul governorate, but the smallest it 2.56 kg C m-2 and corresponded to the SGA categories S5, Kastanozems-G18, Sfax governorate association. In Fig. 4, we note the average stock for each soil categories according to its distribution by governorates.
These stocks are consistent with data for the world level (Batjes,
1996) derived from the WISE (World Inventory of Soil Emission Potentials)
soil database. Batjes (1996) reported worldwide mean
carbon stock values for the 0 to 30 cm layer of 3.1, 4.5 and 5 kg C m-2
for Regosols, Vertisols and Cambisols, respectively. It accounted for 0 to 100
cm depth of 9.6, 11.1 and 9.6 kg C m-2 for Kastanozems, Vertisols
and Cambisols, respectively. But Batjes (1996) calculated
for the soils of arid zone slightly higher values for Lithosols and Gleysols,
(3.6 and 7.7 kg C m-2, respectively for 0-30 cm and 13.1 kg C m-2
for 0 to 100 cm for Gleysols) and lower values for Solonchaks, Luvisols and
Podzoluvisols (1.8, 3.1 and 5.6 kg C m-2, respectively).
||The average stock for each soil categories according to its
distribution by delegations
When the international database of Batjes (1996) derived
from the WISE data is used for Gleysols, the estimated total carbon for this
group is high, presumably because the SCD international database includes several
Gleysols form other regions that contain more carbon than the Tunisian soils.
The regions with the highest organic carbon stocks are located in northern part of Tunisia. On the other hand, Southern part of the country has governorates incorporated the lowest organic carbon stocks. This geographic organic carbon stocks repartition is influenced by regional climatic conditions, when north of Tunisia is rainfall but south is deserted. This consequence is detailed with delegations.
At the superficial layer (0-30 cm), the organic carbon stocks ranged from 0.12 kg C m-2 (S1, Lithosols-1G20, Degueche delegation) to 19.98 kg C m-2 (S7, Luvisols-13G5, Menzel Temime delegation). More than 69% of all SDA categories were associated with organic carbon stocks from 0.12 to 4.99 kg C m-2 and 30% of the extent covered by SDA showed organic carbon stocks varying from 5 to 19.98 kg C m-2.
In 0-100 cm, the biggest value of organic carbon stocks in SDA categories is 35.54 kg C m-2 and corresponded to the S7, Luvisols-13G5, Menzel Temime delegation, but the smallest it 1.03 kg C m-2 and corresponded to the SDA categories S3, Cambisols-1G21, Douz delegation. In Fig. 5 we note the average stock for each soil categories according to its distribution by delegations.
Total organic carbon stored at the country level (0-30 and 0-100 cm): In the following text, three estimates are given for the soil carbon pool of Tunisia, for 0 to 30 cm and 0 to 1 m depth. The first value is based on the soils map and database, the second is founded of soils-governorates maps and database and the third is for soils-delegations maps and database.
Organic carbon stock by soils: The potential total organic carbon stocks of Tunisian soils in different soil groups for the 0 to 30 cm and 0 to 1 m layer was obtained by combining the table of the representative organic carbon stocks in the database with soils map. Using this way, we calculated that the soils of Tunisia store 0.455 Pg C in the superficial layer (0-30 cm) (Fig. 6a) and 1.131 Pg C in 1 m depth (Fig. 7a). Maps of organic carbon stocks by soils showed that Tunisian north have a highest stock, its influenced by vegetation and geographical relief. Organic carbon stocks by soils have an average value 3.36 kg C m-2, but the minimum and the maximum values are 1.84 and 7.16 kg C m-2, respectively.
Organic carbon stock by soils and governorates: The potential total organic carbon stocks of Tunisian soils by governorates for the 0 to 30 cm layer (Fig. 6b) was obtained by SGA map, after combining the soils map with governorates map. We calculated that a total of 0.417 Pg C (417 Tg C). From 0 to 100 cm we obtained 1.031 Pg C (Fig. 7b). By this way, we remarked a decrease in organic carbon stock minimum value from the country level (0.55 kg C m-2 in 0-30 cm and 2.56 kg C m-2 in 0-100 cm) and increase from maximum value (13.06 kg C m-2 in 0-30 cm and 27.29 kg C m-2 in 0-100 cm).
Organic carbon stock by soils and delegations: We calculated Tunisian SOC stocks using SDA map following combining the soils map with delegations map. It was observed that a total of 0.433 Pg C was stored in 0-30 cm layer (Fig. 6c) and 1.084 Pg C was stored in 0 to 100 cm depth (Fig. 7c). By this method, minimum and maximum of organic carbon stock decrease, but value of means by the three ways have in global a few variation between 3.36 and 3.85 kg C m-2 in 0-30 cm layer and between 8.35 and 9.17 kg C m-2 in 0-100 cm layer.
This analysis gives as a clear picture about the characteristic of regions
where climate is arid and similar soils are common, which includes much of Maghreb
countries in North Africa and South Mediterranean sea. Total soil organic carbon
stocks of these regions may have been underestimated because of insufficient
studies and sampling of soils at many depths, by previous approaches based on
soils types and by providing data on spatially referenced estimates of inclusions
within map units. Figure 6 and 7 showed
that an area dominated by forests and mountainous zone contains significant
amounts of organic carbon in North and Tunisian centre.
|| Organic carbon stocks in 0-30 cm depth by (a) soils, (b)
soils-governorates and (c) soils-delegations
|| Organic carbon stocks in 0-100 cm depth by (a) soils, (b)
soils-governorates and (c) soils-delegations
SOC stocks at the clay rich Tunisian soils (Vertisols) were almost twice as
high as at the sandy soils (Lithosols). This is most likely due to effective
stabilisation mechanisms of clay (Bernoux et al.,
2002). Inaccessibility of organic carbon in aggregates and micropores and
adsorption on clay surfaces are acknowledged as major stabilisation mechanisms
(Six et al., 2002).
Figure 6 shows SOC stocks by three methods. In most studies
in database, soils where sampled in extreme southern Tunisia only in Oasis soils.
Irrigation and fertilization used in oasis have an effect on SOC stocks repartition
on SGA and SDA maps. Both Fig. 6b and 6c
showed that soils and governorates or delegations have different influences
on the SOC distribution. For instance, in sud west of Tunisia in Tozeur governorate
(Fig. 6b) we estimate 3 to 6 kg C m-2, but in this
region SOC stock has a values varying between 0.1 and 3 kg C m-2
in 0-30 cm layer. We explain this result that sampled soils have collected with
Oasis and all values from our database have the same origin and influenced SOC
stocks in their soils at this governorate. Equal observations of estimation
of soils and delegations illustrated in Fig. 6c and 7c
at the same sector Tunisian sud west, precisely in Douz and El-Faouar delegations
in Kebili governorate. This result is not surprising in view of the processes
of database elaboration in which punctual sites of sampling has an immense influence.
The result is significant; however, it illustrates the danger of extrapolating
understanding of process from one region.
Figure 6a and 7a showed that soils have
different influences on the organic carbon stocks distribution, depending of
the geographical localization. For example, the regions with the highest organic
carbon stocks has a soil influence marked by the presence of mountainous zones.
On the other hand, North-western Tunisian region had high organic carbon stock
mostly because of the colder climatic influence, which influences soils directly
by forest and those organic matters.
Generally, South country regions are characterized by low SOC stocks and sandy
soils which showed an important climatic influence. This sector surround semiarid
and arid zones and SOC values have ranged between 0.1 and 3 kg C m-2
in superficial layer (0-30 cm) the same as between 1 and 8 kg C m-2
at 1 m depth. Clay with high surface area protects organic carbon from decomposition
on developing stable clay-organic carbon complexes (Singh
et al., 2007). Organic carbon associated with sand particles was
readily decomposable as compared to that in silt and clay. Singh
et al. (2007) confirmed that intensive agriculture without proper
management in the semi-arid region was the cause of rapid SOC depletion in cropland
as compared untilled soils under scrub vegetation.
In similar conditions in Jordan if climate change and/or human land uses alter
these lands, then soil carbon storage could decline (Batjes,
2006). Changes in soil carbon storage, either positive or negative, are
unlikely to be uniform throughout the globe, because the distribution of soil
carbon stocks, the factors that stabilize soil carbon and the forces that contribute
to change vary widely among regions. Moreover, mean annual rainfall, tillage,
period of canopy cover, clay content, land use history and productivity have
pronounced effects on SOC stocks (Bouajila and Gallali,
2008; Brahim et al. 2009). Bernoux
et al. (2002) showed several sources of uncertainties with national
SOC stocks estimation, because the information from soil database stem different
sources and the methodology used for the analyses of organic carbon content
and Db may be varied among different laboratories.
Present results are in support of previous work and provide more details for
aridisols. Indeed, related to previous findings from Tunisian soil carbon stocks,
the calculated SOC stocks to 0-30 cm using the FAO world soils database were
closed to the amount 0.498 PgC reported by Henry et al.
(2009), however the estimated stocks to 1 m were lesser than the result
(0.727 Pg C). These stocks are comparable from stock by soils with estimation
established by Brahim et al. (2010) using 1483
soil profiles for Tunisia it estimate 0.405 and 1.006 Pg C from 0-30 and 0-100
cm, respectively. In this study, we used a larger database (1576 soil profiles)
and estimates stocks by soil type and the administrative division, it provides
The total mass of organic carbon stored in the first 30 cm of the Tunisian soils is comprised between 0.417 and 0.455 Pg C. The estimates of SOC stocks for the entire country are comprised between 1.031 and 1.131 Pg C for the upper 100 cm. The spatial distribution of SOC stocks was mainly determined by the distribution of the organic carbon concentration. Variability of SOC stocks is caused by different factors, like clay contents and climatic zones. Generally, organic carbon stocks by Tunisian soils has a smallest values, low and erratic rainfall inputs of organic carbon into the system while warm conditions facilitate the decomposition of organic matter during the short growing season.
This study provides useful baseline data for future studies dealing with land use changes, the impact on carbon dynamics at regional scale and the first step for calculated CO2 fluxes from soils in Tunisia.
This research was co-financed by AFD (Agence Française pour le Développement), the French Ministry of Foreign Affairs (MAEE), the Fond Français pour l'Environnement Mondial (FFEM), the IRD (Institut de Recherche pour le Développement) through the CORUS-2 project number 6112 Séquestration du carbone et biodiversité dans les sols africains méditerranéens et leurs vulnérabilité aux changements climatiques, the RIME-PAMPA project No. CZZ 3076 PAMPA and ARUB of Pedology Research Unit: 04/UR/10-02.