You are on page 1of 9

Maximizing Farm Income in Agroforestry Sector

Using Linear Programming Model

P. Ahmad Fauzi1, O. Suzyana2, M. Huda Farhana1 & A. R. Rohana1


1
Forest Research Institute Malaysia, 52109 Kepong, Selangor Darul Ehsan
2
International Islamic University Malaysia, P.O. Box 10, 50728 Gombak, Kuala Lumpur

Linear programming was used to determine the optimal net farm income
of several agroforestry cropping system options in Malaysia. Many practical
problems in operational research can be expressed in the form of linear
programming problems. A number of algorithms for other types of optimization
problems could be solved simultaneously by using linear programming. Linear
programming model (LINDO) was applied in this study in order to find out
the best options to optimize the net farm income. The results from this study
will be used to recommend the most feasible and optimal cropping system
suitable in Malaysia.

INTRODUCTION

Linear programming is a problem-solving approach that helps managers make


decisions in the allocation of their limited resources to meet certain criteria.
The objective is usually to minimize cost or maximize profit, revenues, etc. The
reason it is called linear programming is because the objective and the restrictions
(constraints) involved in making the decisions are formulated mathematically,
which enable managers to derive optimal solutions using known mathematical
techniques. Linear programming has found practical application in almost all
facets of business, from advertising to production planning. Transportation and
aggregate production planning problems are the most typical objects of linear
programming analysis. A number of assumptions of the model are made, namely,
in optimization, fixedness, finiteness, determinism, continuity, homogeneity,
addictively and proportionality (Hazell & Norton 1986). The addictively and
proportionality together define linearity in the activities, thereby, giving rise to
the name linear programming.
Linear programming is a mathematical procedure for determining optimal
allocation of scarce resources (Schrage 1986). Further, linear programming is a
method of determining an optimum program of interdependent activities within
available resources and arises whenever two or more activities compete for limited
resources. The term linear assumes that all relationships involved in the particular
problem to be solved by this method are linear. It is not only capable of allocating
resources in order to achieve optimal return but also capable of selecting the
most profitable activities in the business operation. In agricultural projects, linear

1
programming model has been used as investment decision model in regional land
evaluation for maximizing profit by utilizing the available resources (Ahmad Fauzi
1994).
The general objective of this study was to select combine crops which are
expected to best optimize the constraints imposed by available resources. We
estimated the expected land and labour used in order to maximize the net
farm income and to compute the quantity of crops to be produced. Finally we
recommend the optimal cropping system to be adopted in Malaysia.

MATERIALS AND METHODS

The study on optimization model will employ linear programming model in


selecting the best combination of cropping pattern due to topographic land
area constraints as well as other resources available to the farmers. This linear
programming model is expected to maximize the farm income of farmers. The
objective function of the linear programming model could also minimize the
production cost of a given cropping pattern. The general model is as shown
below:
n
Maximize income ∑ cj xj
J 1

n
Subject to ∑ aij xj ≤ ri (i = 1, 2,…m)
J 1

and xj ≤ 0 (j = 1, 2, ….n)

The reverse of this model, i.e. minimizing of income model, uses primal and
dual for maximizing and minimizing concepts respectively. In this paper, we looked
at net farm income for several combination crops, namely, (a) oil palm + sentang
(Azadirachta excelsa) + banana, (b) rubber + sentang, (c) banana + oil palm, (d) oil
palm + sentang, and (e) banana + rubber. The amount of crops produced is very
much dependent on the use of land and labour. Generally, the linear programming
model can be written as follows:

Maximize {Net Farm Income}


Subject to {Input Production Constraints}
{Policy Constraints}

This model would ultimately maximize the net farm income of combined
crops. We used Linear Integer and Quadratic Programming Program (LINDO)
software for this analysis.

2
Data for study

The most important data for the linear programming matrix is the cost per unit
of each input requirement (Table 1). This was the basis used for selecting the best
combination of crops which maximized the net income.
The estimated linear programming model is as shown in Appendix 1. The
model was based on the objectives of study which were to maximize the net farm
income in agroforestry sector and at the same time optimize the constraints
imposed by available resources. The description of variables used in the linear
programming is shown in Appendix 2.

Table 1 Data from agroforestry sectors


Cost (RM) X1* X2* X3* X4* X5*

Total gross income (TGI) 624994.634 380288.32 363294.6 620970.3 122612.7


Land clearing 0 800 0 0 0
Farm road 0 799.2 0 0 399.6
Planting material 2379.2 616 2379.2 2156 839.2
Replanting material 1248.65 225 1308.17 1248.65 225
Cover crop 300.00 1351.39 300 300 44.14
Lining 830.00 220.72 0 830 220.72
Seedling# 1307.25 336 0 1307.25 336
Holing 451.70 952.18 0 392.18 560
Total fix cost 7816.79 5855.49 3987.37 7534.07 3239.18
TSP 26.04 0 26.04 0 26.04
Other chemical fertilizer 9510 9148.38 0 9510 9148.38
NPK 15:15:15 460.33 0 460.33 0 460.33
NPK 12:12:17:2 553.35 0 553.35 0 553.35
Organic fertilizer 527.56 0 527.56 0 527.56
General weedicide 6385.86 5278.38 5887.86 6362.86 1101.37
General insecticide 4463.78 1576.58 4312.43 4200 1840.36
Plastic beg 21.35 0 21.35 0 21.35
Total variable cost 21948.28 16003.34 11788.93 20072.86 13678.76
L planting 0 555 0 0 614.52
L fertilizer application 549.60 158.92 549.60 544.86 163.65
L weedicide application 446.46 219.87 413.26 434.2813 198.84
L insecicide application 288.86 126.13 288.86 283.784 131.20
L pruning 417.33 37.692 404.86 417.3313 25.23
L wrapping of fruit 3.04 0 3.04 0 5.07
L plant maintenance 10.15 0 0 0 10.15
L harvesting 4470.15 2702.52 4290.15 4438.02 2554.65
Total cost of labour per MD 6185.58 3245.13 5949.77 6118.28 3088.79
Total labour cost 92783.67 48676.8919 89246.48 91774.19 46331.8
Land tax 9600 1500 1500 9600 1500
Contingency (10%) 13214.87 7203.57192 10652.28 12898.11 6474.974
TOTAL COST 121532.21 81200.59 93982.99 118355.6 63959.85

x1*- oil palm + sentang + banana; x2*- rubber + sentang; x3*- banana + oil palm; x4*- oil palm + sentang; x5*-
banana + rubber

3
RESULTS AND DISCUSSION

The net income showed that a 0.2 ha plantation crop of oil palm + sentang gave a
high return of about RM12724.95 (Appendices 3 and 4). The combine crops of oil
palm + sentang were more profitable than other combine crops. The combination
of oil palm + sentang gave a total gross income of RM15739.95 and for planting
materials, the combination gave a value of only RM54.65.
For the second run of linear programming, the combination of agroforestry
data for the five plantations needed a little refinement. This showed that linear
programming was able to allocate the variable costs to others matter such as the
cost of labour and maintenance.
The total variable cost of oil palm + sentang can be minimized to RM508.79
in order to maximize the net income to RM12724.95. The result showed that the
combination needed a value of RM243.33 to run this project using land tax. Results
of the total cost of labour per man day can be reduced to give RM155.08. Total
labour cost included the costs for planting, application of fertilizer, weedicide and
insecticide, wrapping of fruits, plant maintenance and also cost for harvesting.
Total fixed cost for the combination of oil palm + sentang was RM190.97 and the
total variable cost was RM508.79.

CONCLUSION

Linear programming application is a very useful tool for solving optimization


problems. This study emphasized on maximizing the net income of each crop
combination, i.e. oil palm + sentang + banana, rubber + sentang, banana + oil
palm, oil palm +sentang and banana + rubber. Linear programming showed
that the intention could optimize the input requirement cost especially land and
labour. Based on the results obtained, we conclude that combination of oil palm
+ sentang increased net income by RM12724.95 annually and gave more profit
than other combine crops.
Besides that, this study also indicated how linear programming played
important roles in the agroforestry system to achieve the goals of optimization
imposed by available resources.

ACKNOWLEDGEMENTS

We are extremely grateful to the Ministry of Science, Technology and Innovation


(MOSTI) for the financial support (Project Number: 05-04-01-0099 EA001). The
Director-General of FRIM, Dr. Joseph Jawa Kendawang and Mr. Dawend Jiwan
from Forest Department Sarawak are thanked for their continuous guidance and
encouragement throughout the project. Our thanks are also due to Dr. Woon Weng
Chuen and Dr. Norini Haron for their leadership and collaboration.

4
REFERENCES

AHMAD FAUZI, P. 1994. Investment Decision Model in Regional Land Evaluation in Recent
Developments in Land Evaluation. Malaysian Society of Soil Science, Kuala Lumpur.
AHMAD FAUZI, P., MOHD SHAHWAHID, O. & AMIR ABDUL NASIR, S. 2002. Optimizing log supply
from timber concession complex, Dungun, Terengganu to their subsidiaries down
stream processing mills using linear programming model. Pp. 197–209 in Jamaluddin,
K., Ahmad, S., Mohd Noor, R., Hafizah, K. & Muzammil, M. (Eds.) The Proceedings of
the Science, Techonology & Social Science National Seminar. 27−28 May 2002, Kuantan.
HAZELL, P. B. R & NORTON, R. D. 1986. Mathematical Programming for Economic Analysis in
Agriculture. Macmillan Publishers Company, New York.
SCHRAGE, L. 1986. Linear, Integer and Quadratic Programming with LINDO. Third edition. The
Scientific Press, University Avenue, Palo Alto.

5
Appendix 1 Estimated linear programming model

MODEL:- MAX = -15*L1 -15*L2 -15*L3 -15*L4 -15*L5 -15 + 503462.42 *X1 + 299087.73 *X2 +
269311.61 *X3 + 502614.73 *X4 + 58652.84 *X5 ;
[LABOUR1] L1 >= 0;
[LABOUR2] L2 >= 0;
[LABOUR3] L3 >= 0;
[LABOUR4] L4 >= 0;
[LABOUR5] L5 >= 0;
[LAND1] LD1 -X1 >= 0;
[LAND2] LD2 -X2 >= 0;
[LAND3] LD3 -X3 >= 0;
[LAND4] LD4 -X4 >= 0;
[LAND5] LD5 -X5>= 0;
[SYNLAND] X1 + X2 + X3 + X4 +X5 <= 10;
[TOTAL_GROSS_INCOME] 624994.634*X1 + 380288.32*X2 + 363294.6*X3 +620970.3*X4 +
122612.7*X5 >= 0;
[LAND_CLEARING] 800*X2 + 47300*X6 + 82800*X7 + 356160*X8 + 50500*X9 + 33950*X10 +
18000*X11 + 20700*X12 + 12320*X13 + 7000*X14 + 100000*X15 + 96600*X16 + 170000*X17 >= 0;
[FARM_ROAD] 799.2*X2 + 399.6*X5 + 1400000*X6 + 138000*X7 + 592000*X8 + 15700*X10 +
60000*X11 +140000*X12 +70000*X13 + 72000*X14 >= 0;
[PLANTING_MATERIAL] 2379.2*X1 + 616*X2 + 2379.2*X3 + 2156*X4 + 839.2*X5 + 119524.68*X6
+ 24242.8*X7 + 277260.02*X8 + 187069.08*X9 + 53115.04*X10 >= 0;
[REPLANTING_MATERIAL] 1248.65*X1 + 225*X2 + 1308.17*X3 + 1248.65*X4 + 225*X5 >= 0;
[COVER_CORP] 300*X1 + 1351.39*X2 + 300*X3 + 300*X4 + 44.14*X5 >= 0;
[LINING] 830.00*X1 + 220.72*X2 + 830*X4 + 220.72*X5 >= 0;
[SEEDLING] 1307.25*X1 + 336*X2 + 1307.25*X4 + 336*X5 >= 0;
[HOLING] 451.70*X1 + 952.18*X2 + 392.18*X4 + 560*X5 >= 0;
[TOTAL_FIXED_COST] 7816.79*X1 + 5855.49*X2 + 3987.37*X3 + 7534.07*X4 + 3239.18*X5 >= 0;
[TSP] 26.04*X1 + 26.04*X3 + 26.04*X5 >= 0;
[OTHER_CHMCL_FERTLZR] 9510*X1 + 9148.38*X2 + 9510*X4 + 9148.38*X5 >= 0;
[NPK_15_15_15] 460.33*X1 + 460.33*X3 + 460.33*X5 >= 0;
[NPK_12_12_17_2] 553.35*X1 + 553.35*X3 + 553.35*X5 >= 0;
[ORG_FERTLZR] 527.56*X1 + 527.56*X3 + 527.56*X5 >= 0;
[GNRL_WEEDICIDE] 6385.86*X1 + 5278.38*X2 + 5887.86*X3 + 6362.86*X4 + 1101.37*X5 >= 0;
[GNRL_INSECTICIDE] 4463.78*X1 + 1576.58*X2 + 4312.43*X3 + 4200*X4 + 1840.36*X5 >= 0;
[PLASTIC_BAG] 21.35*X1 + 21.35*X3 + 21.35*X5 >= 0;
[TOTAL_VAR_COST] 21948.28*X1 + 16003.34*X2 + 11788.93*X3 + 20072.86*X4 + 13678.8*X5
>= 0;
[L_PLANTING] 555*X2 + 614.52*X5 >= 0;
[L_FERTLZR_APPL] -L1 + 549.60*X1 + 158.92*X2 + 549.60*X3 + 544.86*X4 + 163.65*X5 - TL >= 0;
[L_WEEDICIDE_APPL] -L2 + 446.46*X1 + 219.87*X2 + 413.26*X3 + 434.2813*X4 + 198.84*X5
- TL >= 0;
[L_INSECTICIDE_APPL] -L3 + 288.86*X1 + 126.13*X2 + 288.86*X3 + 283.784*X4 + 131.20*X5
+ - TL >= 0;
[L_PRUNNING] -L4 + 417.33*x1 + 37.692*x2 + 404.86*x3 + 417.3313*x4 + 25.23*x5 - TL >= 0;
[L_WRAPPG_OF_FRUITS] 3.04*x1 + 3.04*x3 + 5.07*x5 - TL >= 0;
[L_PLANT_MAINTENANCE] 10.15*x1 + 10.15*x5 + 71100.00*x6 - TL >= 0;
[L_HARVESTING] 4470.15*x1 + 2702.52*x2 + 4290.15*x3 + 4438.02*x4 + 2554.65*x5 - TL >= 0;
[TOTAL_COST_OF_LBR_PER_MD] 6185.58*x1 + 3245.13*x2 + 5949.77*x3 + 6118.28*x4 +
3088.79*x5 >= 0;
[TOTAL_LBR_COST] 92783.66*x1 + 48676.89*x2 + 89246.48*x3 + 91774.19*x4 + 46331.8*x5 >= 0;
[LAND_TAX] 9600*x1 + 1500*x2 + 1500*x3 + 9600*x4 + 1500*x5 >= 0;
[CONTINGENCY] 13214.87*x1 + 7203.57*x2 + 10652.28*x3 + 12898.11*x4 + 6474.974*x5 >= 0;
[TOTAL_COST] 121532.21*x1 + 81200.59*x2 + 93983*x3 + 118355.6*x4 + 63959.8*x5 <= 3000;
[TOTAL_LABOUR] TL <= 20;
[PO1] LD4 >= 5;

6
Appendix 2 Description of variables
Variable name Description

LABOUR[1] L1 Labour for combination 1 (oil palm + sentang + banana)


LABOUR[2] Labour for combination 2 (rubber + sentang)
LABOUR[3] Labour for combination 3 (banana + oil palm)
LABOUR[4] Labour for combination 4 (banana + oil palm)
LABOUR[5] Labour for combination 5 (banana + rubber)
LAND1 Land use for combination 1 (oil palm + sentang + banana)
LAND2 Land use for combination 2 (rubber + sentang)
LAND3 Land use for combination 3 (banana + oil palm)
LAND4 Land use for combination 4 (banana + oil palm)
LAND5 Land use for combination 5 (banana + rubber)
(TGI) Total gross income
[LAND_CLEARING] Land clearing
[FARM_ROAD] Farm road
[PLANTING_MATERIAL] Planting material
[REPLANTING_MATERIAL] Replanting material
[COVER_CORP] Cover crop
[LINING] Lining
[SEEDLING] Seedling
[HOLING] Holing
[TOTAL_FIXED_COST] Total fix cost
[TSP] TSP
[OTHER_CHMCL_FERTLZR] Other chemical fertilizer
[NPK_15_15_15] NPK 15:15:15
[NPK_12_12_17_2] NPK 12:12:17:2
[ORG_FERTLZR] Organic fertilizer
[GNRL_WEEDICIDE] General weedicide
[GNRL_INSECTICIDE] General insecticide
[PLASTIC_BAG] Plastic beg
[TOTAL_VAR_COST] Total variable cost
[L_PLANTING] Labour in planting
[L_FERTLZR_APPL] Labour in fertilizer application
[L_WEEDICIDE_APPL] Labour weedicide application
[L_INSECTICIDE_APPL] Labour insecicide application
[L_PRUNNING] Labour in pruning
[L_WRAPPG_OF_FRUITS] Labour in wrapping of fruit
[L_PLANT_MAINTENANCE] Labour in plant maintenance
[L_HARVESTING] Labour inharvesting
[TOTAL_COST_OF_LBR_PER_MD] Total cost of labour per MD
[TOTAL_LBR_COST] Total labour cost
[LAND_TAX] Land tax
[CONTINGENCY] Contingency (10%)
[TOTAL_COST] Total coat for all combination
[TOTAL_LABOUR] Total of labour used
[PO1] Policy used

7
Appendix 3 Otimization linear programming matrix for agroforestry

Appendix 4 Optimization LINDO output

Global optimal solution found


Objective value: 12724.95
Total solver iterations: 1

Variable Value Reduced Cost


L1 0.000000 15.00000
L2 0.000000 15.00000
L3 0.000000 15.00000
L4 0.000000 15.00000
L5 0.000000 15.00000
X1 0.000000 12642.26
X2 0.000000 45742.70
X3 0.000000 129801.2
X4 0.2534734E-01 0.000000
X5 0.000000 212962.0
LD1 0.000000 0.000000
LD2 0.000000 0.000000
LD3 0.000000 0.000000
LD4 5.000000 0.000000
LD5 0.000000 0.000000
X6 0.000000 0.000000
X7 0.000000 0.000000
X8 0.000000 0.000000
X9 0.000000 0.000000
X10 0.000000 0.000000

8
X11 0.000000 0.000000
X12 0.000000 0.000000
X13 0.000000 0.000000
X14 0.000000 0.000000
X15 0.000000 0.000000
X16 0.000000 0.000000
X17 0.000000 0.000000
Row Slack@ surplus Dual Price
LABOUR1 0.000000 0.000000
LABOUR2 0.000000 0.000000
LABOUR3 0.000000 0.000000
LABOUR4 0.000000 0.000000
LABOUR5 0.000000 0.000000
LAND1 0.000000 0.000000
LAND2 0.000000 0.000000
LAND3 0.000000 0.000000
LAND4 4.9764653 0.000000
LAND5 0.000000 0.000000
SYNLAND 9.974653 0.000000
TOTAL_GROSS_INCOME 15739.95 0.000000
LAND_CLEARING 0.000000 0.000000
FARM_ROAD 0.000000 0.000000
PLANTING_MATERIAL 54.64887 0.000000
REPLANTING_MATERIAL 31.64996 0.000000
COVER_CORP 7.604203 0.000000
LINING 21.03829 0.000000
SEEDLING 33.13531 0.000000
HOLING 9.940721 0.000000
TOTAL_FIXED_COST 190.9687 0.000000
TSP 0.000000 0.000000
OTHER_CHMCL_FERTLZR 241.0532 0.000000
NPK_15_15_15 0.000000 0.000000
NPK_12_12_17_2 0.000000 0.000000
ORG_FERTLZR 0.000000 0.000000
GNRL_WEEDICIDE 161.2816 0.000000
GNRL_INSECTICIDE 106.4588 0.000000
PLASTIC_BAG 0.000000 0.000000
TOTAL_VAR_COST 508.7937 0.000000
L_PLANTING 0.000000 0.000000
L_FERTLZR_APPL 13.81075 0.000000
L_WEEDICIDE_APPL 11.00788 0.000000
L_INSECTICIDE_APPL 7.193170 0.000000
L_PRUNNING 10.57824 0.000000
L_WRAPPG_OF_FRUITS 0.000000 0.000000
L_PLANT_MAINTENANCE 0.000000 0.000000
L_HARVESTING 112.4920 0.000000
TOTAL_COST_OF_LBR_PER_MD 155.0821 0.000000
TOTAL_LBR_COST 2326.232 0.000000
LAND_TAX 243.3345 0.000000
CONTINGENCY 326.9328 0.000000
TOTAL_COST 0.000000 4.246649
TOTAL_LABOUR 20.00000 0.000000
PO1 0.000000 0.000000

You might also like