2 FOOD AND AGRICULTURE ORGANIZATION’S ACTIVITIES IN CARBON SEQUESTRATION IN DRYLAND FARMING SYSTEMS

Soil Carbon Sequestration in Dryland Farming Systems



519



21.3 MAIN CHARACTERISTICS OF FARMING

SYSTEMS IN DRYLANDS

Drylands have particular characteristics that will affect their

capacity to sequester carbon. Drylands often experience high

temperatures, low and erratic rainfall, minimal cloud cover,

and small amounts of plant residues to act as surface cover

to minimize radiation impact. As a result, soils in the drylands

are, generally, both inherently low in organic matter and

nutrients and rapidly lose large proportions of those small

quantities as CO2 when exposed by tillage and other conventional practices. Exposed and loosened soils are also highly

prone to soil erosion, particularly rainfall patterns that

include intense, storm precipitation after long dry periods.

The key issue in drylands is therefore to maximize the capture, infiltration, and storage of rainfall water into soils by

promoting conditions that accumulate organic matter and

increase soil biodiversity. Drylands are particularly prone to

soil degradation and desertification, with 70% of the agricultural land degraded. This means that soils have lost considerable amounts of carbon. As a consequence, the C stock of

most dryland soils is less than 1%, and in many cases less

than 0.5% (Lal, 2002). Increasing soil quality is therefore the

main strategy for CS in drylands. Because drylands cover

approximately 43% of the Earth’s land surface (FAO, 2000),

and dryland soils have lost carbon as a result of land degradation, they offer a great potential to sequester carbon (Scurlock and Hall, 1998; Rosenberg et al., 1999). Furthermore,

soil carbon decomposition is also dependent on soil moisture,

so dry soils are less likely to lose carbon (Glenn et al., 1993),

and consequently the residence time of carbon in drylands is

much longer than, for instance, in forest (Gifford et al., 1992).

Whereas forest and intensive farming systems may be important carbon sinks, increasing carbon in degraded agricultural

soils of dryland regions would also have direct environmental,

economic, and social benefits to the local people and smallholders that depend on them.



© 2005 by Taylor & Francis Group, LLC



520



Koohafkan, Rey, and Antoine



Although most of the research on soil organic matter

dynamics and processes has been conducted in temperate

zones, several reviews have highlighted the potential offered

by drylands and degraded lands to sequester carbon (Izaurralde et al., 2001; Lal, 2001). Agricultural productivity in

drylands is not only limited by natural constraints, but also

by low input management as a result of limited resources and

technologies. The depletion in soil carbon in agricultural soils

as a result of land misuse and soil mismanagement, can be

reversed. Important strategies to improve productivity

include (1) growing adapted species, (2) enhancing water-use

efficiency and water retention in soils, (3) managing and

enhancing soil fertility, and (4) adopting improved cropping

systems (Lal, 2001). Improved cropping systems include crop

rotations, planted fallows, residue mulch, conservation of

trees, and growing leguminous species. Recommended practices involve soil water conservation and management, irrigation management, soil fertility management either by

adding inorganic fertilizers or organic inputs, residues management, and reduced or zero tillage (Lal, 2003). Some of these

practices are the main principle of conservation agriculture,

which has been proven to be effective in increasing productivity and CS. Furthermore, the fact that conservation agriculture requires much less external inputs makes it more

attractive to poor farmers. The case studies presented here

analyse the effect of such practices on soil carbon stocks in

various dryland systems.

21.4 CASE STUDIES IN DRYLANDS

Some global estimates have been made about the potential

for CS in drylands. The total amount of C loss as a consequence of desertification may be 18 to 28 Pg (1015 g, or 1

gigaton) C (Lal, 2001). Assuming that two-thirds of the C lost

(18 to 28 Pg) can be resequestered (IPCC, 1995) through soil

and vegetation restoration, the potential of C sequestration

through desertification control is 12 to 18 Pg C (Lal, 2002).

The case studies presented here assess the effect of different



© 2005 by Taylor & Francis Group, LLC



Soil Carbon Sequestration in Dryland Farming Systems



521



management practices on soil carbon stocks in various dryland ecosystems.

The effect of climate and/or land use change can be predicted only through the use of accurate dynamic models.

Given the difficulty of measuring changes in soil carbon

stocks, modeling is a useful tool and has been used as an

effective methodology for analysing and predicting the effect

of land management practices on soil carbon stocks. A number

of process-based models have been developed over the last

two decades and are available (as reviewed by Smith et al.,

1997). FAO has developed a model in collaboration with the

University of Trent (Canada) as a methodological framework

for the assessment of carbon stocks and prediction of CS

scenarios that links SOC turnover simulation models (particularly CENTURY and RothC-26.3) to geographical information systems and field measurement procedures (FAO,

2004a).

For the case studies, the CENTURY 4.0 (Parton et al.,

1987, 1988) model was used. It has been tested against a

variety of long-term agricultural field trials and has also been

used in a variety of climatic zones, including dryland regions.

The ability of any model to predict accurately into the future

depends on the accuracy and quality of the data used to

parametize the model (climatic, soil, and land management

data). Few studies contain sufficiently detailed information

and the complete data set required for modeling purposes,

particularly in dryland regions where such studies are scarce.

Data from four distinctly different dryland systems in

Nigeria (Kano region), Kenya (Makueni district), India

(Andhra Pradesh and Karnataka States), and Argentina

(Monte Redondo and Santa Maria provinces) were used in

investigations carried out by Essex University. Table 21.1

summarizes the main characteristics of these systems.

21.4.1



Case Study 1: Nigeria



Nigeria comprises some of the most densely inhabited areas

of semi-arid West Africa (Harris, 2000). As a result, the soils

of this region have been cultivated for long periods. Plant



© 2005 by Taylor & Francis Group, LLC



522



Table 21.1 Main Characteristics of Study Sites



1

Kano Region



Case Study

Country

Soil type



Farming

systems



Study sites



Futchimiram, Borno State:

Low-intensity agropastoral

CP: 5-year cycle of grazing and

millet cropping



© 2005 by Taylor & Francis Group, LLC



2

Andrha Pradesh and

Karnataka States



Kenya

India

Ferralsols naturally low Alfisols and vertisols

in P



Argentina

Haplic

Phaeozems



Annual or multiple

cropping



Smallholder farming

Integration among

livestock, crops, and

trees



Grazed prairie

Row cropping



Short in livestock

Very little fertilizer

Crop residue management

Darjani



Cattle manure

Tillage

Soil fertility treatment

Lingampally

Large mixed dryland

farming



Inorganic

fertilizers

No tillage

Tucuman

province:

graze prairie

and row

cropping



Koohafkan, Rey, and Antoine



Farming

practice



Nigeria

Ferrugineous tropical soils, sandy,

poor water holding capacity

(WHC) and low-nutrient organic

matter content

Smallholder farming:

Intensive: permanent annual or

biannual cultivation (cropping

intensity >60%)

Less intensive: shrub/short-bush

fallow regime (30% to 60%)

Extensive: long-bush fallow and

uncultivated areas (<30%)

Low-input systems

Cattle manure



3

Makueni District



4

Tucuma,

Catamarca, and

Cordoba

Provinces



Main crops



Kaiani



Yedakulapaly

Large farmers using

irrigation



Dagaceri, Jigawa State:

Intensive agropastoral (legumes

and grains)

CP: shrub with short-bush

fallowing

Tumbau, Kano close-settlement

zone (CSZ):

Highly intensive agricultural

CP: crop and livestock production

system with intercropping of

legumes, intensive manuring,

and inorganic fertilizer

Millet, sorghum, groundnut,

sesame, cowpea



Kymausoi



Metalkunta

Small mixed dryland

farming

Mixed crop and livestock



Athi Kamunyuni



Malligere

Small, mixed dryland

farming



Natural

Open forest savannahs

vegetation Grasslands



Tucuman

province:

graze prairie

and row

cropping

Catamarca

province



Cordoba

province



Maize,

Maize and pulses

Large agrodiversity,

sunflower,

Millet, cowpea, sorghum

between 8 to 10 crops:

wheat, and

paddy, sorghum, maize,

soybean

millet, groundnut,

coconut, cotton, etc.

Grass, woodland system



Soil Carbon Sequestration in Dryland Farming Systems



Kaska, Yobe State:

Low-intensity agropastoral

CP: 7-year cycle of grazing and

millet-cowpea cropping



Note: CP = current practices.



523



© 2005 by Taylor & Francis Group, LLC



524



Koohafkan, Rey, and Antoine



production is limited by rainfall and nutrients (Breman and

De Wit, 1983). The economy and infrastructure of northern

Nigeria are not suited to high external inputs or fertilizers,

and thus smallholder farming units operate as low-input systems. Legumes such as cowpeas are used to provide nitrogen

inputs.

CENTURY was run for several practices at four sites:

Futchimiram, Kaska, Dagaceri, and Tumbau for the last 50

to 60 years with alternate cycles of grazing and cropping

(Table 21.1). Land degradation is a problem at all sites. Current practices (CP) were compared with continuous cultivation (CC), additions of inorganic fertilizer (IF), farmyard

manure (FYM), plant residues (PR), and retained plant residues and grazing (NG). The predicted annual change in soil

carbon for the various scenarios is presented in Figure 21.1.

The modeling exercise of the farming systems in the

Nigerian case studies shows that soil carbon stocks can be

increased from a low base with a variety of technologies and

practices already available to farmers (Figure 21.1). The total

amount of carbon that can be sequestered with the use of

legumes, fallow periods, farmyard manure, and retention of

plant residues varied between 0.1 to 0.3 metric tons C ha −1

year−1. Figures for CS were slightly higher when trees were

introduced. The use of inorganic fertilizers caused no change

or loss of soil carbon. Continuous cultivation (reduced fallows)

caused small carbon losses each year when no additional

organic inputs were provided. However, when a cropping practice is accumulating significant amounts of carbon, fallowing

will decrease the CS potential. Despite the intensification of

the current systems, the levels of carbon were maintained.

The main conclusion from these systems is that CS can only

be achieved by increasing organic inputs into the soil.

21.4.2



Case Study 2: Kenya



Arid and semi-arid lands occupy two-thirds of Kenya (Nandwa

et al., 1999). Erratic rainfall and poor fertility as a result of

intensive cultivation are the main limiting factors of plant

productivity. Droughts have also affected farming livelihoods.



© 2005 by Taylor & Francis Group, LLC



Soil Carbon Sequestration in Dryland Farming Systems



Λ soil carbon

(t C ha-1 y-1)



0.4



Futchimiram



CP

CC

2+ NG, harvest only grain

IF (100 kg ha-1 urea), NG

PR (0.5 t ha-1 y-1), NG

5y F, 5 y C, 2 applications FYM 3 t ha-1, GR

CC, FYM 1.5 t ha-1 y-1, GR

CC, FYM 1.5 t ha-1 y-1, PR0.5 t ha-1 y-1, NG



0.3

0.2

0.1

0.0

-0.1

1



2



3



4



525



5



6



7



8



Λ soil carbon

(t C ha-1 y-1)



0.4

0.3



Kaska



CP

CC: millet-cowpea

C-F, FYM 3 t ha-1, to millet

CC, FYM 3 t ha-1, to millet



0.2

0.1

0.0

-0.1

1



Λ soil carbon

(t C ha-1 y-1)



0.4



2



3



4



Dagaceri



CP

remove trees

NG, harvest aboveground

CC, millet-cowpea

FYM 1.29 t ha-1 y-1,F, GR, harvest only grain

NGRs, harvest only grain



0.3

0.2

0.1

0.0

-0.1

1



Λ soil carbon

(t C ha-1 y-1)



0.4

0.3



2



3



4



5



6



Tumbau



CP

IF (110 kg ha-1 urea)

FYM 3.75 t ha-1 y-1

FYM 6.75 t ha-1 y-1

FYM 3.75 t ha-1 y-1, add nitrogen-fixing trees

FYM 6.75 t ha-1 y-1, plant residues 2 t ha-1

FYM 6.75 t ha-1 y-1, harvest only grain



0.2

0.1

0.0

-0.1

1



2



3



4



5



6



7



SCENARIOS

CP:

CC:

NG:

IF:

GR:



Current practice

Continuous cultivation

No grazing residues

Inorganic fertilizer

Grazing residues



PR:

F:

C:

FYM:



Plant residues

Fallow

Cultivation

Farmyard manure



Figure 21.1 Average annual change in soil carbon stock under

various practice scenarios (CENTURY) for the case study 1: Nigeria.



© 2005 by Taylor & Francis Group, LLC



526



Koohafkan, Rey, and Antoine



Annual or multiple cropping is practiced with occasional fallow periods. CENTURY was run to equilibrium using a grassland tree scenario with grass fires every 10 years and major

fires every 30 years at four farmland sites: Darjani, Kaiani,

Kymausoi, and Athi Kamunyuni. CPs were compared with

CC, burn residues (BR), inorganic fertilizers (IF), and FYM

applications (Figure 21.2).

The practices that led to increased CS were the addition

of organic material in the form of farmyard manure and plant

residues, particularly when systems are at steady state or

near. Removal of fallows resulted in losses of 0.1 metric tons

C ha−1y−1. In this case, the use of inorganic fertilizers was also

inadequate as a sole source of plant nutrients. The combination of legumes in rotations, 2 to 4 metric tons ha−1 year−1 of

farmyard manure in addition to 0.6 metric tons ha−1 year−1 of

plant residues results in the highest rate of CS in all dryland

cases: 0.7 metric tons C ha−1 year−1.

21.4.3



Case Study 3: India



Over half of the farming population of India live in semi-arid

regions. Crop yields have recently increased as a result of the

green revolution. In this case, irrigation and inorganic fertilizers are expensive and inaccessible for the rural poor. Soil

fertility treatments using legume cultivation and vermicompost production are increasingly used, leading generally to

increased organic matter. Meanwhile, practices such as the

use of inorganic fertilizers and the continuous cultivation of

cereals lead to a substantial decline in soil carbon levels.

CENTURY was run for several practices at three sites:

Lingampally, Metalkunta, and Malligere. The CP is 1 year of

fallow and 4 years of cropping. The use of IF, FYM, green

manure, vermicompost (V) and PR, trees, and legumes inclusion were compared with CP (Figure 21.3).

The use of farmyard manure, green manure, and vermicompost and plant residues produced increases in soil carbon

of 0.2 to 0.4 ha−1 year−1. Increases in soil carbon can also be

obtained by leaving crop residues in the soil. Agroforestry

practices substantially increased soil carbon by 0.9 metric



© 2005 by Taylor & Francis Group, LLC



Soil Carbon Sequestration in Dryland Farming Systems



527



Λ soil carbon

(t C ha-1 y-1)



1.0

0.8



CP

CC

only IF, BR,CC

only IF,BR, F

FYM 4.5 t ha-1 y-1, BR, F

FYM 4.5 t ha-1 y-1, do not BR,F

FYM 6.75 t ha-1 y-1, do not BR,F



Darjani



0.6

0.4

0.2

0.0

-0.2



1



2



3



4



5



6



7



Λ soil carbon

(t C ha-1 y-1)



1.0

0.8



Kaiani



CP

FYM 2t ha-1 y-1

FYM 2t ha-1 y-1, PR0.3 tha-1 y-1



0.6

0.4

0.2

0.0

-0.2



1



Λ soil carbon

(t C ha-1 y-1)



1.0

0.8



2



3



Kymausoi



CP

PR0.3 t ha-1 y-1

FYM 1.5 t ha-1 y-1, PR 0.6 tha-1 y-1

FYM 1.5 t ha-1 y-1, PR 0.6t ha-1 y-1, L(cowpea)

FYM 2t ha-1 y-1, PR 0.6t ha-1 y-1, L(cowpea)

FYM 4t ha-1 y-1, PR 0.6t ha-1 y-1, L(cowpea)



0.6

0.4

0.2

0.0

-0.2



1



Λ soil carbon

(t C ha-1 y-1)



0.8



2



3



4



5



6



Athi Kamunyuni



CP

G

FYM 1.25 t ha-1 y-1

FYM 2.25 t ha-1 y-1

FYM 2.27 t ha-1 y-1,F reduced to1year

FYM 3.3 t ha-1 y-1

FYM 3.9 t ha-1 y-1, PR0.3 t ha-1



0.6

0.4

0.2

0.0



1



2



3



4



5



6



7



SCENARIOS

CP:

CC:

NG:

IF:



Current practice

Continuous cultivation

No grazing residues

Inorganic fertilizer



PR:

Plant residues

F:

Fallow

BR:

Burn residues

FYM: Farm yard manure



Figure 21.2 Average annual change in soil carbon stock under

various practice scenarios (CENTURY) for the case study 2: Kenya.



© 2005 by Taylor & Francis Group, LLC



528



Koohafkan, Rey, and Antoine



Λ soil carbon

(t C ha-1 y-1)



1.0

0.8



Lingampally



CP

IF

replace IF with FYM

FYM 6 t ha-1 y-1

FYM 6 t ha-1 y-1, GM 250 kg ha-1 y-1,V 100 kg ha-1 y-1

FYM 6 t ha-1 y-1, PR 2 t ha-1y -1

FYM 6 t ha-1 y-1, PR 2 t ha-1 y -1, trees



0.6

0.4

0.2

0.0

-0.2

1



Λ soil carbon

(t C ha-1 y-1)



1.0

0.8



2



3



4



5



6



7



Metalkunta



CP

PR



0.6

0.4

0.2

0.0

1



Λ soil carbon

(t C ha-1 y-1)



1.0



2



Malligere



CP

replace IF with FYM

add trees, Glyricidia



0.8

0.6

0.4

0.2

0.0

1



2



3



SCENARIOS

CP:

IF:

V:



Current practice

Inorganic fertilizer

Vermicompost



PR:

FYM:

GM:



Plant residues

Farmyard manure

Green manure



Figure 21.3 Average annual change in soil carbon stock under

various practice scenarios (CENTURY) for the case study 3: India.



tons C ha−1 year−1. Leguminous crops had also clear beneficial

effects when included in rotations. The use of inorganic fertilizers resulted in carbon loss. Inorganic fertilizers and irrigation both have a carbon cost, reducing the amount of

sequestered carbon. A full carbon accounting in these farming

systems should consider the high-energy cost of nitrogen fertilizers (Pretty et al., 2002), the use of mechanized operations,

and the transfer of carbon from one farm to another by livestock.



© 2005 by Taylor & Francis Group, LLC



Soil Carbon Sequestration in Dryland Farming Systems



21.4.4



529



Case Study 4: Argentina



In recent years, Argentina has adopted reduced or zero-tillage practices especially in dryland regions, as a result of soil

degradation. Two case studies under various conventional

and zero-tillage systems were used: Monte Redondo and

Santa Maria provinces (Figure 21.4). The use of CT, zerotillage (ZT), IF, FYM, and green manure addition scenarios

were compared.

Adopting zero tillage will halt the decline in soil carbon.

However, to induce CS, organic inputs are needed (green and

farmyard manures), and can be used to replace the inorganic

fertilizer applications.



Λ soil carbon

(t C ha-1 y-1)



0.4



Monte Redondo



CT, IF

ZT, IF

ZT, fym 1.5 t ha-1 y-1, IF

ZT, GM 10 t ha-1 crop-1, IF

ZT, FYM 1.5 t ha-1 y-1, GM 10 t ha-1 crop-1, no IF

ZT: FYM 3.3 t ha-1 crop-1, no IF



0.3

0.2

0.1

0.0

-0.1



1



2



3



4



5



6



Λ soil carbon

(t C ha-1 y-1)



0.4

0.3



Santa Maria



CT, IF

ZT, IF

ZT, FYM 1.5 t ha-1 y-1, IF

ZTl, FYM 3.3t ha-1 crop-1, no IF

ZT, GM 10 t ha-1 crop-1, IF

ZT, FYM 1.5 t ha-1 y-1, GM10 t ha-1 crop-1, no IF



0.2

0.1

0.0

-0.1



1



2



3



4



5



6



SCENARIOS

CT: Conventional tillage

IF:

Inorganic fertilizer

ZF: Zero tillage



GM: Green manure

FYM: Farm yard manure



Figure 21.4 Average annual change in soil carbon stock under

various practice scenarios (CENTURY) for the case study 4:

Argentina.



© 2005 by Taylor & Francis Group, LLC