39
software. Model for predicting the yield of organic rice farming production could be
developed using Verhulst growth model Burghes Borrie 1981. The following Equation 3 was used as the model for the prediction of the yield of organic rice
production through time.
3
where: Yt = yield with respect to time tonha Y
= initial yield tonha Y
∞
= maximum sustainable yield tonha = coefficient
t = time year
Equation 3 implies that the yield will level at its maximum value through time. How long the leveling condition will be reached depends upon the limitations
on soil fertility and land area. Some historical data is needed to run the model in order to predict the yield.
4.3. Results and Discussion
4.3.1. Production Yield
Rice production data was collected randomly from farmers practicing SRI organic rice farming in the study area of Sukabumi District. In general, farmers in
this area are small land holders less than 0.5 ha. However, their lands are of technical irrigation paddy fields so that there is no problem for them to apply the
SRI method. The production data including some of the production factors is presented in Table 4.1.
Most farmers in the areas do not have good farm management practice. They do not have well recorded production data, particularly in water requirement.
Consequently, the water requirement had to be calculated using the following Equation 4 Doorenbos Kassam in Tarjuelo de Juan 1999:
1 1
ETm ETa
Ym Ya
4
1
1 1
t o
e Y
Y Y
t Y
40
where: Ya = actual yield kg
ETa = actual seasonal evapotranspiration ET mm Ym
= maximum yield kg ETm = seasonal ET for maximum yield mm
= water yield sensitivity coefficient. Table 4.1. Rice production data of farmers practicing SRI organic rice farming in
Sukabumi District
No. Name of Farmer
Area Water
1
Seed Fertilizer
Labor Production
Yield
ha 1000 m
3
ha kgha
tonha m-daysha
2
tons tonha
1. H. Baban
0.200 3.220
7.500 7.500
225.000 1.600
8.000 2.
H. Uk ar 0.250
3.220 6.400
7.200 220.000
2.000 8.000
3. Jalal
0.200 3.310
7.500 8.000
225.000 1.640
8.200 4.
Maman 0.170
3.070 7.100
7.600 265.000
1.300 7.647
5. Baen
0.150 3.220
6.700 6.700
240.000 1.200
8.000 6.
Oben 0.120
3.190 7.100
7.000 300.000
0.950 7.917
7. Barnas
0.050 3.220
10.000 16.000
160.000 0.400
8.000 8.
Suhendi 0.200
3.650 7.500
10.000 225.000
1.800 9.000
9. Udin
0.150 3.220
3.300 6.700
240.000 1.200
8.000 10.
Suparman 0.300
2.970 5.000
3.300 183.000
2.220 7.400
11. F. Sarifudin
0.070 4.230
7.100 18.600
243.000 0.727
10.386 12.
Endang 0.100
3.220 10.000
8.000 170.000
0.800 8.000
13. Pahrudin
0.060 2.660
16.700 33.300
283.000 0.400
6.667 14.
Neneng Hamidah 0.070
2.250 14.300
0.400 243.000
0.400 5.714
15. Nanang
0.075 3.360
6.700 0.700
227.000 0.625
8.333 16.
Usup Pakot 0.120
2.040 8.300
0.400 300.000
1.200 5.208
17. Iep
0.070 1.530
7.100 1.000
243.000 0.280
4.000 18.
Romi 0.060
1.840 8.300
8.300 283.000
0.285 4.750
19. Hasan
0.040 2.800
12.500 12.500
200.000 0.280
7.000 20.
Ujang Zaenal 0.070
3.550 7.100
8.600 243.000
0.615 8.786
1
Water requirement was calculated using formula of Equation 4
2
Labor required per season
Cobb-Douglas production function was used to develop model for predicting the yield by conducting optimization for the constant values of Equation 2. The
optimization resulted in the following model:
94 .
002 .
00019 .
002 .
664 .
2 W
L F
S Y
LD
4
where the constant and parameters of A, , , , are 2.664, -0.002, 0.00019, 0.002,
and 0.94, respectively. The model is valid only in the range of minimum and maximum values of S, F, L and W variables as presented in Table 4.2.
41
Based on the data of the production factors S, F, L, and W and values of the constants, the model was used to predict the production. The results are presented in
Table 4.1. The values of the model agreed nicely with the data values as depicted in the following regression curve with an error of 0.002 Fig. 4.1. Data is presented in
Appendix 4.1. For the purpose of financial farm analysis, which is not discussed in detail in this analysis, labor should be expressed in m-days with its unit price.
Labors are needed in land preparation, transplanting, fertilizing, weeding, harvesting, etc. The total labor needed per hectare varies from 200 to 300
m-daysha DISIMP NTT 2008; Fitriadi Nurmalina 2009; Rachmiyanti 2009.
Production Regression Curve
y = 1.0012x - 0.0095 R
2
= 0.9990 2.00
3.00 4.00
5.00 6.00
7.00 8.00
9.00 10.00
11.00
2.00 4.00
6.00 8.00
10.00 12.00
Field Data toha Mo
d e
l to
n h
a
Model vs. Field Data
Figure 4.1. Regression of observation data vs. model for rice production yield
Furthermore, optimization of parameters of the production factors used in this model resulted in the following values of variables as presented in Table 4.2. The
yield components of the production factors was also calculated and presented in Table 4.3.
4.3.2. Prediction of Yield