A farm-level analysis for carbon sequestration in Ghana using

IMPACT linked with the DSSAT, Household and Ruminant models

E. González-Estrada1, V.K. Walen2, J. Naab3, P.K. Thornton1 and M. Herrero1

1International Livestock Research Institute, PO Box 30709, Nairobi, Kenya2University of Florida, PO Box 110570, Gainesville, FL, USA3Savanna Agricultural Research Institute, PO Box 494, Wa, Upper West Region, Ghana 

Case study:

CRSP-SM field trials in Wa, Upper West region of Ghana

Objective:

Explore the role of crop-livestock interactions in the carbon dynamics at the farm level

IMPACT

Integrated Modelling Platform for Mixed Animal-Crop Systems

Household Optimization Model

Multiple-Criteria linear programming

DSSAT

Crops

Ruminant

Livestock

Framework

Case studyPiisi, Upper West Region, Ghana

Household size: 12 (5 adults & 7 children)

Plot ID Size Crop

1 0.8 ha Groundnuts

2 1.6 ha Rice

3 1.2 haMilletCowpeaSorghum

4 0.3 ha Bambara nuts

5 0.6 ha Yams

6 0.4 ha Maize

Case study –2-

Livestock Number

Cattle 2

Sheep 3

Goats 2

Pigs 3

Chicken 15

Guinea fowl 8

IMPACT’s base line scenario analysis:

Net income: 160 USD

Total carbon balance in agriculturalland (Inputs-Outputs): -3219 kg

Food security status: Low energy and protein intake during June-October

IMPACT

Integrated Modelling Platform for Mixed Animal-Crop Systems

Household Optimization Model

Multiple-Criteria linear programming

DSSAT

Crops

Ruminant

Livestock

Framework

ILRI’s household model:

• Integrates biological, social and economic aspects of smallholder farming systems

• Linear programming model– An objective function (i.e. maximize net income)– A set of production activities– A set of constraints

• Runs for a period of 1 year in monthly timesteps

• The model maximises gross margins subject to:

– Land constraints (fixed farm size and number of plots)

– Satisfying food security for the household

– Labour constraints

– Cash constraints (cannot spend more than what is generated)

– Seasonal production constraints for crop and livestock activities

• Other objectives can be maximised or minimised (i.e minimise costs, nutrient losses)

ILRI’s household model - constraints -

HH model Output 1. Food security

- Role of livestock within the system -

Base-line analysis:

•Only chicken, pigs, guinea fowls and a goat are sold or consumed by the household

•Ruminants are kept for dowry, sacrifices, savings (no cash expenses for their maintenance other than 7,000 cedis a year for health)

Optimised management (food security achieved throughout the year, no land-use changes):

Two tropical livestock units (TLU) are sold to generate cash to “buy food security”.

HH model Output 2.Maximize net income

Change of land-use pattern:

• Increase the number of TLU that are sold!!!

• Increase the proportion of fodder-yielding crops (e.g. maize)

No practical solution!!!

• Farmer’s attitude towards risk.

HH model Output 3.Maximize carbon fixation

Net income: 96 USD (-40%)

Total carbon balance in agriculturalLand (Inputs-Outputs): 1930 kg

Food security: Achieved

• Ground nuts• Rice• Bambara nuts• Millet• Sorghum• Yams• Maize

Current land use

• Ground nuts• Rice• Bambara nuts• Maize

Optimized land use

“Where is my Fufu???”

Back to: HH model Output 1. - Role of livestock within the system -

Base-line analysis:

•Only chicken, pigs, guinea fowls and a goat are sold or consumed by the household

•Ruminants are kept for dowry, sacrifices, savings (no cash expenses for their maintenance other than 7,000 cedis a year for health)

Optimised scenario by the household model (food security achieved throughout the year, no land-use changes):

Two tropical livestock units (TLU) are sold to generate cash to “buy food security”.

Some more on the role of livestock within the system and carbon sequestration

Field trials. CRSP plots in Ghana.Improved MAIZE management

• Crop residue incorporated to the soil

Simulate grain yield in response to maize stover incorporated to the soil

DSSAT-Century

Simulate TLU maintenance in response to maize stover availability

Ruminant

0

50

100

150

200

250

300

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

rain

fall (m

m)

Rainfall distribution in Wa, and its effect on fodder resources utilization

Homestead native fodder

Maize stover

0

1

2

3

4

5

6

0 10 20 30 40 50 60 70 80 90 100Stover incorporated to the soil (%)

Num

ber o

f TLU

2600

2650

2700

2750

2800

2850

2900

2950

Gra

in y

ield

(kg/

ha)

Trade-off analysis by simulated output of number of TLU vs grain yield

Grain yield Number of TLU

Conclusion

Is the household of this case study ready for adopting technologies to improve carbon sequestration?

Household level analyses for priority setting

Produce a household typology for a given ecoregion

Identify farm types that are capable to modify management strategies towards a better use of carbon

La fin

Merci!