- Email: [email protected]
Energy use and sensitivity analysis of energy inputs for alfalfa production in Iran
Energy for Sustainable Development 16 (2012) 84–89
Contents lists available at SciVerse ScienceDirect
Energy for Sustainable Development
Energy use and sensitivity analysis of energy inputs for alfalfa production in Iran Hassan Ghasemi Mobtaker ⁎, Asadollah Akram, Alireza Keyhani Department of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, Iran
a r t i c l e
i n f o
Article history: Received 12 January 2011 Revised 28 October 2011 Accepted 30 October 2011 Available online 25 November 2011 Keywords: Energy Alfalfa Modeling Cobb–Douglas Sensitivity of energy
a b s t r a c t The aims of this study were to investigate influences of energy inputs and energy forms on output levels and evaluation of inputs sensitivity for alfalfa production in Hamedan province, Iran. The sensitivity of energy inputs was estimated using the marginal physical productivity (MPP) method and partial regression coefficients on alfalfa yield. Data were collected from 80 alfalfa farms in August and September 2009. The sample volume was determined by random sampling method. The total energy of 810.57 GJ ha− 1 was calculated for establishment and 7 years production life. The electricity used in pumping system was the highest energy inputs for alfalfa production (75.79%). Econometric model evaluation showed that the machinery energy was the most significant input affecting the output level. Sensitivity analysis results indicate that with an additional use of 1 MJ of each machinery and seeds energy, would lead to an additional increase in yield by 5.094 and 4.986 kg, respectively. The MPP of human labor, farmyard manure and biocides was negative. It can be because of applying the inputs more than required or improperly applying. © 2011 International Energy Initiative. Published by Elsevier Inc. All rights reserved.
Introduction Alfalfa (Medicago sativa L.), a perennial legume is used primarily as hay, silage and pasture for animal fodder and also as a source of protein. It is grown in monoculture, in rotation with grain crops and in mixtures with various species of forage grasses and is entered into the rotation after winter cereals or before spring crops such as corn or cotton. Alfalfa is sown, managed and harvested for hay production using a variety of cultural practices that greatly influence the energetics of its production. It is established preferably in the autumn or in the spring. During the four or five years after sowing, the crop is harvested two to three times per year in dryland areas and five to seven times in irrigated lands. Various systems are used for irrigation. After harvest, the hay is preserved mainly as small to medium rectangular bales (25–45 kg per bale) (Tsatsarelis and Koundouras, 1994). The production of alfalfa in 2007 was about 5,132,500 t/yr in Iran and the cultivation land area was about 50,000 ha. Hamedan province with 11% share is ranked second in alfalfa production in Iran (Anonymous, 2007). In Hamedan province alfalfa is established preferably in autumn (September and October) and it is harvested three or four times per year (in June, July, August and also in September if it is harvested four times a year) and up to 6 or 7 years after sowing. Efficient use of energy is one of the principal requirements of sustainable agriculture. Energy use in agriculture has been increasing
⁎ Corresponding author. Tel.: + 98 2612801011; fax: + 98 2612808138. E-mail address: [email protected] (H.G. Mobtaker).
in response to increasing population, limited supply of arable land, and a desire for higher standards of living. Continuous demand in increasing food production resulted in intensive use of chemical fertilizers, pesticides, agricultural machinery, and other natural resources. However, intensive use of energy causes problems threatening public health and environment. Efficient use of energy in agriculture will minimize environmental problems, prevent destruction of natural resources, and promote sustainable agriculture as an economical production system (Erdal et al., 2007). Modern farming has become very energy intensive; therefore there is a great need to balance the use and availability of energy (Singh et al., 2000). Production functions are central to the determination of the efficient allocation of resources. Many researchers have studied energy analysis and relationship between inputs and yield to determine the energy efficiency of plant production. Singh et al. (2000) applied different mathematical functions to establish relationship between the yield and total energy input for cotton. Their result showed that the average yield of cotton can be increased by 6 ± 8% with an additional energy input of 1 ± 3%, mainly through tillage, irrigation and spraying. Singh et al. (2004) studied energy use for wheat crop in India. They applied Cobb–Douglas function to establish relationship between energy inputs and yield and accounted regression coefficients and marginal physical productivity for each of energy inputs. Mohammadi et al. (2010) estimated the production function of kiwifruit. They investigated different functions in order to analyze the relationship between energy inputs and energy output and at last the Cobb–Douglas function was selected as the suitable function. Results revealed that energy inputs of human labor, water for irrigation, total fertilizer and machinery contributed
0973-0826/$ – see front matter © 2011 International Energy Initiative. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.esd.2011.10.009
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
Nomenclature n N s d t Yi X1 X2 X3 X4 X5 X6 X7 X8 ei αi βi γi DE IDE RE NRE MPPxj αj GM(Y) GM(Xj)
required sample size number of holdings in target population standard deviation acceptable error (permissible error was chosen as 5%) confidence limit (1.96 in the case of 95% reliability) yield level of the ith farmer labor energy machinery energy diesel fuel energy chemical fertilizer energy farmyard manure energy biocides energy electricity energy seed energy error term coefficients of the exogenous variables coefficients of the exogenous variables coefficients of the exogenous variables direct energy indirect energy renewable energy non-renewable energy marginal physical productivity of jth input regression coefficient of jth input geometric mean of yield geometric mean of jth input energy
significantly to the yield. The impact of human labor energy was found to be the highest among the other inputs in kiwifruit production. Mohammadi and Omid (2010) studied energy inputs–yield relationship for greenhouse cucumber production in Iran. The regression results revealed that the contribution of energy inputs on yield (except for fertilizers and seeds energies) was significant. Mobtaker et al. (2010) investigated the energy consumption and inputs sensitivity for barley production in Iran. Econometric model evaluation showed that machinery energy was the most significant input which affects the output level. As well, sensitivity analysis indicates that MPP of biocides energy was negative. Rafiee et al. (2010) determined the energy balance between the energy inputs and yield for apple production in Tehran, Iran. Results showed the energy input of diesel fuel had the biggest share within the total energy inputs. The impact of human labor, farmyard manure, water for irrigation, electricity and chemical fertilizer energy inputs was significantly positive on yield. Although many studies have conducted on energy use in agricultural crops (Tsatsarelis, 1993; Mandal et al., 2002; Yilmaz et al., 2005; Jianbo, 2006; Strapatsa et al., 2006; Cetin and Vardar, 2008; Mohammadi et al., 2008), there is no any study on the energy consumption and sensitivity analysis for alfalfa production in Iran. The aims of this study were to determine the amount of energy use and the relationship between energy inputs and yield of alfalfa production in Iran. Also sensitivity analysis of energy inputs for alfalfa crop production was investigated and the Marginal Physical Product technique was utilized to analyze the sensitivity of energy inputs on alfalfa yield.
85
1,008,038 ha. The central region of Hamedan province is an important alfalfa production center. According to Ministry of Jihad-e-Agriculture (MAJ) experts and local officers, there were 1500 alfalfa farms in 2007 which were distributed homogeneously in this region (Anonymous, 2007). Therefore, a simple random sampling method was used to determine survey volume and the farms were chosen randomly from study region. This method is expressed as below (Kizilaslan, 2009; Mobtaker et al., 2010): n¼
Nðs t Þ2 ðN−1Þd2 þ ðs t Þ2
ð1Þ
where n is the required sample size; s, the standard deviation; t, the t value at 95% confidence limit (1.96); N, the number of holding in target population and d, the acceptable error (permissible error 5%). Consequently calculated sample size in this study was 68, but it was considered to be 80 to ensure the accuracy. The data were collected using a face to face questionnaire in August and September 2009. The questionnaires included total energy inputs from different sources and yield weight. The inputs used in the production of alfalfa were specified in order to calculate the energy equivalences in the study. Inputs in alfalfa production were: human labor, machinery, diesel fuel, chemical fertilizers, farmyard manure, biocides, electricity and seeds. It should be mentioned that the free sources of energy (solar energy input for photosynthesis) are not accounted for. The output was dry hay (15% w.b.). To calculate the energy equivalent of inputs and output, units in Table 1 were used. In order to obtain a relationship between inputs and yield, a mathematical function needs to be specified. For this purpose different functions were investigated and finally Cobb–Douglas production function was selected; because it produced better results (yielded better estimates in terms of statistical significance and expected signs of parameters). The Cobb–Douglas production function is frequently used in both energy and economics studies to show the relationship between input factors and the level of production (Singh et al., 2000; Singh et al., 2004; Mohammadi and Omid, 2010; Mobtaker et al., 2010). This function is nothing but the logarithmic function. Logarithmic functions are used where changes of variables in the model show as many folds compared to one another. The coefficient of variables in this function which is in log form also represents elasticities (Mohammadi and Omid, 2010). Also, it is easy to analyze, and it seems to be a good approximation for actual productions (Singh et al., 2000). The Cobb–Douglas production function is expressed as: Y ¼ f ðxÞ expðuÞ:
ð2Þ
Eq. (2) can be further re-written as: n lnY i ¼ a þ ∑j¼1 α j ln X ij þ ei i ¼ 1; 2; 3;…; n
ð3Þ
where Yi denotes the yield of the ith farmer; Xij, the vector of inputs used in the production process; a, the constant term; αj, represent coefficients of inputs which are estimated from the model and ei, the error term. Assuming that when the energy input is zero, the crop production is zero too, Eq. (3) is reformed to (Singh et al., 2003; Hatirli et al., 2006; Mohammadi and Omid, 2010): n lnY i ¼ ∑j¼1 α j ln X ij þ ei :
ð4Þ
Materials and methods The study was performed in central region of Hamedan province which is located in the west of Iran; within 33° 59′ and 35° 48′ north latitude and 47° 34′ and 49° 36′ east longitude. Hamedan region has an area of 1,949,400 ha; and the farming area, with a share of 51.7%, is
In the present case, n = 8; therefore Eq. (4) can be expressed in the following form: lnY i ¼ α 1 lnX 1 þ α 2 lnX 2 þ α 3 lnX 3 þ α 4 lnX 4 þ α 5 lnX 5 þ α 6 lnX 6 þ α 7 lnX 7 þ α 8 lnX 8 þ ei ð5Þ
86
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
Table 1 Energy equivalent of inputs and output in agricultural production. Inputs (unit)
Unit
A. Inputs 1. Human labor
h
1.96
2. Machinery
h
64.80
3. Diesel fuel
l
56.31
4. Chemical fertilizers (a) Nitrogen
kg
66.14
(b) Phosphate (P2O5) (c) Potassium (K2O) 5. Farmyard manure 6. Biocides 7. Electricity
kg
12.44
kg
11.115
kg
0.30
8. Seed
kg
28.1
B. Output 1. Dry hay (15% w.b.)
kg
15.8
kg kWh
Energy equivalent (MJ unit− 1)
120 11.93a
Reference
Yilmaz et al. (2005); Unakitan et al. (2010) Singh (2002); Hatirli et al. (2006) Rafiee et al. (2010); Unakitan et al. (2010)
Esengun et al. (2007); Mobtaker et al. (2010) Esengun et al. (2007); Mobtaker et al. (2010) Shrestha (1998); Mobtaker et al. (2010) Singh (2002); Bojaca and Schrevens (2010) Mobtaker et al. (2010) Ozkan et al. (2004); Hatirli et al. (2005) Tsatsarelis and Koundouras (1994)
Tsatsarelis and Koundouras (1994)
where MPPxj is marginal physical productivity of jth input; αj, regression coefficient of jth input; GM(Y), geometric mean of yield; and GM(Xj), geometric mean of jth input energy on per hectare basis. A positive value of MPP of any factor indicates that with an increase in input, production is increased and a negative value of MPP of any factor input indicates that additional units of inputs contribute negatively to production. Hence, it is better to keep the variable resource in surplus rather than utilizing it as a fixed resource (Singh et al., 2004). In last part of the study the returns to scale of production was calculated. In production, returns to scale refer to changes in output, subsequent to a proportional change in all inputs (where all inputs increase by a constant factor). In the Cobb–Douglas production function, n P it is indicated by the sum of the elasticities ( α i ) derived in the form of i¼1 regression coefficients. If the sum of the coefficients is greater than unity n P ( α i > 1), then it could be concluded as the increasing returns to scale, n i¼1 P on the other hand if the latter parameter is less than unity ( α i b1), i¼1
then it is indicated as the decreasing returns to scale; and, if the result n P is unity ( α i ¼ 1), it shows as the constant returns to scale (Singh et al., 2004).i¼1 Basic information on energy inputs and alfalfa yields were entered into Excel's spreadsheet and SPSS 17.0 software. Modeling carried out using linear regression technique. Results and discussion
a
This coefficient used according to the efficiency of power plants and power loss of distribution networks reported in references for Iran.
where Xi stands for corresponding energies as X1, human labor; X2, machinery; X3, diesel fuel; X4, chemical fertilizers; X5, farmyard manure; X6, biocides; X7, electricity; and X8, seed. Energy demand in agriculture can be divided into direct energy (DE), indirect energy (IDE), renewable energy (RE) and non-renewable energy (NRE). Direct energy is directly used at farms and on fields. Indirect energy, on the other hand, consists of the energy used in the manufacture, packaging and transport of fertilizers, pesticides and farm machinery (Ozkan et al., 2004). The DE includes human labor, diesel fuel and electricity while the IDE covers chemical fertilizers, farmyard manure, biocides, machinery and seeds used in the alfalfa production. Renewable energy includes human labor, seeds and farmyard manure and non-renewable energy consists of machinery, diesel fuel, chemical fertilizers, biocides and electricity. The effect of direct, indirect, renewable and non-renewable energies on production was also studied. For this purpose, Cobb–Douglas function was determined as Eqs. (6) and (7): ln Y i ¼ β1 ln DE þ β2 ln IDE þ ei
ð6Þ
ln Y i ¼ γ1 ln RE þ γ2 ln NRE þ ei
ð7Þ
where Yi is the ith farm's yield; DE, IDE, RE and NRE are direct, indirect, renewable and non-renewable energies used for alfalfa production, respectively. βi and γi are coefficient of exogenous variables. Eqs. (5)–(7) were estimated using ordinary least square technique. The Marginal Physical Product (MPP) technique, based on the response coefficients of the inputs, was utilized to analyze the sensitivity of energy inputs on alfalfa yield. The MPP which is a factor indicates the change in the output with a unit change in the factor input in question, keeping all other factors constant at their geometric mean level. The MPP of the various inputs was computed using the αj of the various energy inputs as (Singh et al., 2004): MPP xj ¼
GM ðY Þ αj GM X j
ð8Þ
Table 2 shows the quantities of inputs used in whole production life (7 years) and establishment of baled alfalfa hay production and their energy equivalences. Also Fig. 1 shows the distribution percent of the energy associated with the inputs. The results revealed that around 576 h of human labor and 255 h of machinery power per hectare were required to produce alfalfa in the research area. The total energy input for various processes in the alfalfa production was calculated to be 810.57 GJ ha − 1. Tsatsarelis and Koundouras (1994) calculated the energy inputs for alfalfa whole production life and establishment in Greece as 116 GJ ha − 1. The highest average energy consumption of inputs was for electricity (614.37 GJ ha − 1) which was accounted for about 76% of the total energy input (Fig. 1), followed by chemical fertilizers (105.45 GJ ha − 1, 13%). The electricity consumption in alfalfa production is for irrigation purposes. Having deep wells in the region and not using modern efficient irrigation methods are among the reasons of high consumption of electrical energy in the studied region. In order to reduce the electricity consumption, using of modern methods of irrigation with high efficiency (which leads in saving water consumption) can be suggested. Table 2 Amounts of inputs, outputs and energy inputs and output in alfalfa production. Inputs (unit)
Quantity per unit area (ha)
Inputs 1. Human labor (h) 2. Machinery (h) 3. Diesel fuel (L) 4. Chemical fertilizers (kg) (a) Nitrogen (kg) (b) Phosphate (P2O5) (kg) (c) Potassium (K2O) 5. Farmyard manure (ton) 6. Biocide (kg) 7. Electricity (MWh) 8. Seeds (kg) The total energy input (GJ)
575.65 254.53 1144.46 2878.61 1296.90 1575.62 6.09 12.78 18.04 51.50 95.87
Output 1. Dry hay (15% w.b.) (ton) Total energy output (GJ)
96.53
Total energy equivalent (GJ ha− 1) 1.13 16.49 64.44 105.45 85.78 19.60 0.067 3.83 2.16 614.37 2.69 810.57
1525.21 1525.21
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
80.0
614.37
Percentage (%)
70.0 60.0 50.0 40.0 30.0
105.44
20.0 10.0
1.12
16.49
64.44
3.83
2.69
2.16
0.0
Fig. 1. The share and amount (GJ ha − 1 ) of energy inputs for alfalfa production in Hamedan, Iran.
The energy share of nitrogen fertilizer was about 81% of total energy of used chemical fertilizer. As alfalfa is a leguminous and nitrogen fixing plant, the nitrogen fertilizer can be substituted by other fertilizers such as potassium, phosphorus and even farmyard manures. This can lead to less consumption of energy. The energy consumption was minimum for potassium (67.90 MJ ha− 1) which accounted for about 0.01% of the total energy consumption. The consumption of human labor was only 0.14% (1.13 GJ ha − 1). Similar results have been reported in the literature implying that the energy input of human labor has a little share of total energy input in agricultural production (Sartori et al., 2005; Strapatsa et al., 2006; Kizilaslan, 2009). The average yield of alfalfa (for 7 years) was obtained to be 96.53 t ha − 1, accordingly, the total energy output per hectare was calculated as 1525.21 GJ ha − 1 (Table 2). The share of energy input as direct, indirect, renewable and nonrenewable forms is illustrated in Fig. 2. The total consumed energy input could be classified as direct energy (83.88%), and indirect energy (16.12%) or renewable energy (00.95%) and non-renewable energy (99.05%). This indicates that alfalfa production depends mainly on non-renewable energy (electricity, diesel fuel and chemical fertilizers) in the studied area. Therefore, it is clear that non-renewable energy consumption was higher than that of renewable in alfalfa production, which is in agreement with the literatures for different crops (Yilmaz et al., 2005; Erdal et al., 2007; Kizilaslan, 2009; Mobtaker et al., 2010). Since the main non-renewable inputs were electricity and chemical fertilizers, efficient use of water and management of plant nutrients
1.2
802.90
Percentage (%)
1
using renewable resources like farmyard manure would increase the rate of renewable energy. Regression results for Eq. (5) are shown in Table 3. For data used in this study, autocorrelation was tested using Durbin–Watson method (Hatirli et al., 2005; Mohammadi and Omid, 2010). The Durbin–Watson value was found to be 2.671 for Eq. (5) which indicates that there was no autocorrelation at the 5% significance level in the estimated model. The R2 value was determined as 0.99 for this equation, implying that around 0.99 of the variability in the energy inputs was explained by this model. The results of assessment of Cobb–Douglass function on each inputs in alfalfa production indicates that the impact of each inputs differ in constitution of yield. The results revealed that the impact of energy inputs could have positive effect on yield (except for human labor, farmyard manure and biocides). As it can be seen from Table 3, Machinery had the highest impact (0.875) among the other inputs and significantly contributed on the yield at 1% level. It indicates that a 1% increase in the energy machinery input led to 0.875% increase in yield in these circumstances. The second important input was found to be the diesel fuel with 0.193 elasticity, followed by seeds and electricity, with elasticity of 0.135 and 0.032 respectively. Mobtaker et al. (2010) concluded that machinery energy was the most significant input affecting the barley output level in Iran. Hatirli et al. (2006) developed an econometric model for greenhouse tomato production in Antalya province of Turkey and reported that the human labor, chemical fertilizers, biocides, machinery and water energy were important inputs, which significantly contributed to yield. The human labor, farmyard manure and biocides had a negative impact on alfalfa yield. The value of return to scale for the model (1) was calculated by gathering the regression coefficients as 1.136. The higher value of return to scale than unity implies increasing return to scale. The sensitivity of energy inputs on output level was analyzed using the marginal physical productivity method and partial regression coefficients and the results are provided in Table 3. As it can be seen, the major MPP was drown for machinery energy (5.094), followed by seeds energy (4.986). It indicates that using an additional of 1 MJ either for machinery or seeds energy would result in increasing the yield by 5.094 and 4.986 kg, respectively. As a consequence, parameters with a large sensitivity coefficient have a strong influence on the state variable. This identifies which factors should be identified and measured most carefully to assess the state of the environmental system, and which environmental factors should be managed preferentially (Drechsler, 1998). The MPP of human labor, farmyard manure and biocides energy were found to be −9.462, − 1.391 and −2.635; a negative value of MPP implies that additional units of inputs are contributing negatively
Table 3 Econometric estimation results of inputs. Endogenous variable: yield
679.86
87
Coefficient
t-ratio
MPP
0.8
Exogenous variables
0.6
Model 1: ln Yi = α1 ln X1 + α2 ln X2 + α3 ln X3 + α4 ln X4 + α5 ln X5 + α6 ln X6 + α7 ln X7 + α8 ln X8 + ei Human labor − 0.110 − 1.475 − 9.462 Machinery 0.875 8.268⁎ 5.094 Diesel fuel 0.193 2.633⁎⁎ 0.289 Chemical fertilizers 0.025 1.677⁎⁎⁎ 0.028 Farmyard manure − 0.002 − 0.699 − 1.391 Biocides − 0.012 − 3.258⁎ − 2.635 Electricity 0.032 1.276 0.006 Seeds 0.135 3.175⁎ 4.986 Durbin–Watson 1.761 R2 0.998 n P Return to scale ( α i ) 1.136
0.4 0.2
130.61 7.64
0
Fig. 2. The share and amount (GJ ha− 1) of total energy input in the form of direct, indirect, renewable and non-renewable for alfalfa production in Hamedan, Iran.
i¼1
⁎, ⁎⁎, ⁎⁎⁎ indicates significance at 1%, 5% and 10% levels, respectively.
88
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
to production, i.e. less production with more input. The other important variables that affect alfalfa yield are diesel fuel and chemical fertilizer energy with MPP of 0.289 and 0.028, respectively. Singh et al. (2004) estimated the sensitivity of energy inputs on wheat productivity for five agro-climate zones in India. They reported that MPP of chemicals in zones 1–5, were calculated to be 0.385, 2.816, − 0.211, 0.610 and 0.624, respectively. Rafiee et al. (2010) estimated the sensitivity of energy inputs on apple production in Iran. The results showed that the MPP value of water for irrigation was the highest, followed by human labor and chemicals energy inputs, respectively. The energy obtained from existing inputs was divided into two direct and indirect forms. The assessed trends of both forms of energy were positive, indicating the positive impacts on the output level. Impact of direct energy (0.566) was more than that of indirect energy (0.340). This impact was significant at 1% level. The regression coefficient for renewable energy (0.322) and non-renewable energy (0.644) was significant at 1% level. It is concluded that impact of non-renewable energy was higher than that of renewable energy in Alfalfa production. Similar results have been reported in the literatures (Mohammadi and Omid, 2010; Unakitan et al., 2010). Durbin–Watson values were calculated as 2.290 and 2.371 for Eqs. (6) and (7), respectively; indicating that there is no autocorrelation at the 5% significance level in the estimated models. The R 2 values were found to be 0.99 and 0.96, respectively. As it can be seen from Table 4 the MPP of direct, indirect, renewable and non-renewable energy was found to be 0.087, 0.293, 0.188 and 0.012, respectively. This indicates that an additional use of 1 MJ of each of direct, indirect, renewable and non-renewable energy would lead to an additional increase in yield by 0.087, 0.293, 0.188 and 0.012 kg, respectively. It is concluded that impact of non-renewable energy was higher than that of renewable energy in alfalfa production. Mohammadi and Omid (2010) and Hatirli et al. (2005) have also reported similar results for greenhouse cucumber production in Iran and energy use in Turkish agriculture, respectively. The return to scale values for the models (2) and (3) were 0.906 and 0.966, respectively, implying the decreasing return to scale. Energy management is an important issue in terms of efficient, sustainable and economic use of energy. Optimization is an important way to maximize the amount of productivity, which can significantly impact the energy consumption and production costs. Linear programming is among the ways that can be used to optimize the energy inputs. So, the present study can be extended to determine wasteful uses of energy inputs by inefficient farms and suggest necessary quantities of various inputs should be used by them. More studies in this direction are currently underway.
Table 4 Econometric estimation results of direct, indirect, renewable and non-renewable energies. Endogenous variable: energy output
Coefficient
t-ratio
MPP
Exogenous variables Model 2: ln Yi = β1 ln DE + β2 ln IDE + ei Direct energy Indirect energy Durbin–Watson R2 n P Return to scale ( β i )
0.566 0.340 2.290 0.99 0.906
8.198⁎ 4.324⁎
0.087 0.293
0.322 0.644 2.371 0.96 0.966
6.930⁎ 21.278⁎
0.188 0.012
i¼1
Model 3: ln Yi = γ1 ln RE + γ2 ln NRE + ei Renewable energy Non-renewable energy Durbin–Watson R2 n P Return to scale ( γ i ) i¼1
⁎ Indicates significance at 1%.
Conclusion The purpose of this study was to investigate influences of energy inputs and energy forms on output levels and evaluation of inputs sensitivity for alfalfa production in Hamedan province, Iran. Data were collected from 80 alfalfa farms and the sample volume was determined by random sampling method. Total energy input was found to be 810.57 GJ ha − 1 and total energy output was calculated as 1525.21 GJ ha − 1. It was founded that the electricity was the most energy consuming input followed by chemical fertilizers. The share of non-renewable energy was higher than that of renewable energy consumption. Econometric estimation results revealed that machinery had the highest impact (0.875) among other inputs and significantly contributed on yield at 1% level. The estimated MPP for machinery energy was the biggest among inputs of energy and MPP of Human labor, farmyard manure and biocides energy were negative. The MPP of direct, indirect, renewable and nonrenewable energy were found to be 0.087, 0.293, 0.188 and 0.012, respectively. Optimal consumptions of electricity, chemical fertilizers and other major inputs would be useful not only in reducing negative effects to environment, but also in maintaining sustainability. Lack of soil analysis in the area leads to unconscious usage of chemical fertilizer. In order to reduce the electricity consumption, using of modern methods of irrigation with high efficiency (which leads in saving water consumption) can be suggested. Also it is suggested that new policies are to be taken to reduce the negative effects of energy inputs such as plant, soil and climate pollution. Therefore, analysis of energy consumption is an important task.
References Anonymous. Annual agricultural statistics. Ministry of Jihad-e-Agriculture of Iran; 2007http://www.maj.ir. Bojaca CR, Schrevens E. Energy assessment of peri-urban horticulture and its uncertainty: case study for Bogota, Colombia. Energy 2010;35(5):2109–18. Cetin B, Vardar A. An economic analysis of energy requirements and input costs for tomato production in Turkey. Renew Energy 2008;33:428–33. Drechsler M. Sensitivity analysis of complex models. Biol Conserv 1998;86(3): 401–12. Erdal G, Esengun K, Erdal H, Gunduz O. Energy use and economical analysis of sugar beet production in Tokat province of Turkey. Energy 2007;32(1):35–41. Esengun K, Gunduz O, Erdal G. Input–output energy analysis in dry apricot production of Turkey. Energy Convers Manage 2007;48:592–8. Hatirli SA, Ozkan B, Fert C. An econometric analysis of energy input–output in Turkish agriculture. Renewable Sustainable Energy Rev. 2005;9(6):608–23. Hatirli SA, Ozkan B, Fert C. Energy inputs and crop yield relationship in greenhouse tomato production. Renew Energy 2006;31(4):427–38. Jianbo L. Energy balance and economic benefits of two agroforestry systems in northern and southern China. Agric Ecosyst Environ 2006;116:255–62. Kizilaslan H. Input–output energy analysis of cherries production in Tokat Province of Turkey. Appl Energy 2009;86(7–8):1354–8. Mandal KG, Saha KP, Ghosh PK, Hati KM, Bandyopadhyay KK. Bioenergy and economic analysis of soybean-based crop production systems in central India. Biomass Bioenergy 2002;23(5):337–45. Mobtaker HG, Keyhani A, Mohammadi A, Rafiee S, Akram A. Sensitivity analysis of energy inputs for barley production in Hamedan Province of Iran. Agric Ecosyst Environ 2010;137(3–4):367–72. Mohammadi A, Omid M. Economical analysis and relation between energy inputs and yield of greenhouse cucumber production in Iran. Appl Energy 2010;87(1):191–6. Mohammadi A, Tabatabaeefar A, Shahan S, Rafiee S, Keyhani A. Energy use and economical analysis of potato production in Iran a case study: Ardabil province. Energy Convers Manage 2008;49(12):3566–70. Mohammadi A, Rafiee Sh, Mohtasebi SS, Rafiee H. Energy inputs–yield relationship and cost analysis of kiwifruit production in Iran. Renew Energy 2010;35(5): 1071–5. Ozkan B, Akcaoz H, Fert C. Energy input–output analysis in Turkish agriculture. Renew Energy 2004;29(1):39–51. Rafiee Sh, Mousavi Avval SH, Mohammadi A. Modeling and sensitivity analysis of energy inputs for apple production in Iran. Energy 2010;35:3301–6. Sartori L, Basso B, Bertocco M, Oliviero G. Energy use and economic evaluation of a three year crop rotation for conservation and organic farming in NE Italy. Biosyst Eng 2005;91(2):245–56. Shrestha DS. Energy use efficiency indicator for agriculture; 1998. Available from: http://www.usaskca/agriculture/caedac/PDF/mcrae.
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89 Singh JM. On farm energy use pattern in different cropping systems in Haryana, India. Master of science. Germany: International Institute of Management, University of Flensburg; 2002. Singh S, Singh S, Pannu CJS, Singh J. Optimization of energy input for raising cotton crop in Punjab. Energy Convers Manage 2000;41(17):1851–61. Singh H, Mishra D, Nahar NM, Ranjan M. Energy use pattern in production agriculture of a typical village in arid zone India: part II. Energy Convers Manage 2003;44(7): 1053–67. Singh G, Singh S, Singh J. Optimization of energy inputs for wheat crop in Punjab. Energy Convers Manage 2004;45(3):453–65.
89
Strapatsa AV, Nanos GD, Tsatsarelis CA. Energy flow for integrated apple production in Greece. Agric Ecosyst Environ 2006;116(3–4):176–80. Tsatsarelis CA. Energy inputs and outputs for soft winter wheat production in Greece. Agric Ecosyst Environ 1993;43:109–18. Tsatsarelis CA, Koundouras DS. Energetics of baled alfalfa hay production in northern Greece. Agric Ecosyst Environ 1994;49(2):123–30. Unakitan G, Hurma H, Yilmaz F. An analysis of energy use efficiency of canola production in Turkey. Energy 2010;35:3623–7. Yilmaz I, Akcaoz H, Ozkan B. An analysis of energy use and input costs for cotton production in Turkey. Renew Energy 2005;30(2):145–55.
Contents lists available at SciVerse ScienceDirect
Energy for Sustainable Development
Energy use and sensitivity analysis of energy inputs for alfalfa production in Iran Hassan Ghasemi Mobtaker ⁎, Asadollah Akram, Alireza Keyhani Department of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, Iran
a r t i c l e
i n f o
Article history: Received 12 January 2011 Revised 28 October 2011 Accepted 30 October 2011 Available online 25 November 2011 Keywords: Energy Alfalfa Modeling Cobb–Douglas Sensitivity of energy
a b s t r a c t The aims of this study were to investigate influences of energy inputs and energy forms on output levels and evaluation of inputs sensitivity for alfalfa production in Hamedan province, Iran. The sensitivity of energy inputs was estimated using the marginal physical productivity (MPP) method and partial regression coefficients on alfalfa yield. Data were collected from 80 alfalfa farms in August and September 2009. The sample volume was determined by random sampling method. The total energy of 810.57 GJ ha− 1 was calculated for establishment and 7 years production life. The electricity used in pumping system was the highest energy inputs for alfalfa production (75.79%). Econometric model evaluation showed that the machinery energy was the most significant input affecting the output level. Sensitivity analysis results indicate that with an additional use of 1 MJ of each machinery and seeds energy, would lead to an additional increase in yield by 5.094 and 4.986 kg, respectively. The MPP of human labor, farmyard manure and biocides was negative. It can be because of applying the inputs more than required or improperly applying. © 2011 International Energy Initiative. Published by Elsevier Inc. All rights reserved.
Introduction Alfalfa (Medicago sativa L.), a perennial legume is used primarily as hay, silage and pasture for animal fodder and also as a source of protein. It is grown in monoculture, in rotation with grain crops and in mixtures with various species of forage grasses and is entered into the rotation after winter cereals or before spring crops such as corn or cotton. Alfalfa is sown, managed and harvested for hay production using a variety of cultural practices that greatly influence the energetics of its production. It is established preferably in the autumn or in the spring. During the four or five years after sowing, the crop is harvested two to three times per year in dryland areas and five to seven times in irrigated lands. Various systems are used for irrigation. After harvest, the hay is preserved mainly as small to medium rectangular bales (25–45 kg per bale) (Tsatsarelis and Koundouras, 1994). The production of alfalfa in 2007 was about 5,132,500 t/yr in Iran and the cultivation land area was about 50,000 ha. Hamedan province with 11% share is ranked second in alfalfa production in Iran (Anonymous, 2007). In Hamedan province alfalfa is established preferably in autumn (September and October) and it is harvested three or four times per year (in June, July, August and also in September if it is harvested four times a year) and up to 6 or 7 years after sowing. Efficient use of energy is one of the principal requirements of sustainable agriculture. Energy use in agriculture has been increasing
⁎ Corresponding author. Tel.: + 98 2612801011; fax: + 98 2612808138. E-mail address: [email protected] (H.G. Mobtaker).
in response to increasing population, limited supply of arable land, and a desire for higher standards of living. Continuous demand in increasing food production resulted in intensive use of chemical fertilizers, pesticides, agricultural machinery, and other natural resources. However, intensive use of energy causes problems threatening public health and environment. Efficient use of energy in agriculture will minimize environmental problems, prevent destruction of natural resources, and promote sustainable agriculture as an economical production system (Erdal et al., 2007). Modern farming has become very energy intensive; therefore there is a great need to balance the use and availability of energy (Singh et al., 2000). Production functions are central to the determination of the efficient allocation of resources. Many researchers have studied energy analysis and relationship between inputs and yield to determine the energy efficiency of plant production. Singh et al. (2000) applied different mathematical functions to establish relationship between the yield and total energy input for cotton. Their result showed that the average yield of cotton can be increased by 6 ± 8% with an additional energy input of 1 ± 3%, mainly through tillage, irrigation and spraying. Singh et al. (2004) studied energy use for wheat crop in India. They applied Cobb–Douglas function to establish relationship between energy inputs and yield and accounted regression coefficients and marginal physical productivity for each of energy inputs. Mohammadi et al. (2010) estimated the production function of kiwifruit. They investigated different functions in order to analyze the relationship between energy inputs and energy output and at last the Cobb–Douglas function was selected as the suitable function. Results revealed that energy inputs of human labor, water for irrigation, total fertilizer and machinery contributed
0973-0826/$ – see front matter © 2011 International Energy Initiative. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.esd.2011.10.009
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
Nomenclature n N s d t Yi X1 X2 X3 X4 X5 X6 X7 X8 ei αi βi γi DE IDE RE NRE MPPxj αj GM(Y) GM(Xj)
required sample size number of holdings in target population standard deviation acceptable error (permissible error was chosen as 5%) confidence limit (1.96 in the case of 95% reliability) yield level of the ith farmer labor energy machinery energy diesel fuel energy chemical fertilizer energy farmyard manure energy biocides energy electricity energy seed energy error term coefficients of the exogenous variables coefficients of the exogenous variables coefficients of the exogenous variables direct energy indirect energy renewable energy non-renewable energy marginal physical productivity of jth input regression coefficient of jth input geometric mean of yield geometric mean of jth input energy
significantly to the yield. The impact of human labor energy was found to be the highest among the other inputs in kiwifruit production. Mohammadi and Omid (2010) studied energy inputs–yield relationship for greenhouse cucumber production in Iran. The regression results revealed that the contribution of energy inputs on yield (except for fertilizers and seeds energies) was significant. Mobtaker et al. (2010) investigated the energy consumption and inputs sensitivity for barley production in Iran. Econometric model evaluation showed that machinery energy was the most significant input which affects the output level. As well, sensitivity analysis indicates that MPP of biocides energy was negative. Rafiee et al. (2010) determined the energy balance between the energy inputs and yield for apple production in Tehran, Iran. Results showed the energy input of diesel fuel had the biggest share within the total energy inputs. The impact of human labor, farmyard manure, water for irrigation, electricity and chemical fertilizer energy inputs was significantly positive on yield. Although many studies have conducted on energy use in agricultural crops (Tsatsarelis, 1993; Mandal et al., 2002; Yilmaz et al., 2005; Jianbo, 2006; Strapatsa et al., 2006; Cetin and Vardar, 2008; Mohammadi et al., 2008), there is no any study on the energy consumption and sensitivity analysis for alfalfa production in Iran. The aims of this study were to determine the amount of energy use and the relationship between energy inputs and yield of alfalfa production in Iran. Also sensitivity analysis of energy inputs for alfalfa crop production was investigated and the Marginal Physical Product technique was utilized to analyze the sensitivity of energy inputs on alfalfa yield.
85
1,008,038 ha. The central region of Hamedan province is an important alfalfa production center. According to Ministry of Jihad-e-Agriculture (MAJ) experts and local officers, there were 1500 alfalfa farms in 2007 which were distributed homogeneously in this region (Anonymous, 2007). Therefore, a simple random sampling method was used to determine survey volume and the farms were chosen randomly from study region. This method is expressed as below (Kizilaslan, 2009; Mobtaker et al., 2010): n¼
Nðs t Þ2 ðN−1Þd2 þ ðs t Þ2
ð1Þ
where n is the required sample size; s, the standard deviation; t, the t value at 95% confidence limit (1.96); N, the number of holding in target population and d, the acceptable error (permissible error 5%). Consequently calculated sample size in this study was 68, but it was considered to be 80 to ensure the accuracy. The data were collected using a face to face questionnaire in August and September 2009. The questionnaires included total energy inputs from different sources and yield weight. The inputs used in the production of alfalfa were specified in order to calculate the energy equivalences in the study. Inputs in alfalfa production were: human labor, machinery, diesel fuel, chemical fertilizers, farmyard manure, biocides, electricity and seeds. It should be mentioned that the free sources of energy (solar energy input for photosynthesis) are not accounted for. The output was dry hay (15% w.b.). To calculate the energy equivalent of inputs and output, units in Table 1 were used. In order to obtain a relationship between inputs and yield, a mathematical function needs to be specified. For this purpose different functions were investigated and finally Cobb–Douglas production function was selected; because it produced better results (yielded better estimates in terms of statistical significance and expected signs of parameters). The Cobb–Douglas production function is frequently used in both energy and economics studies to show the relationship between input factors and the level of production (Singh et al., 2000; Singh et al., 2004; Mohammadi and Omid, 2010; Mobtaker et al., 2010). This function is nothing but the logarithmic function. Logarithmic functions are used where changes of variables in the model show as many folds compared to one another. The coefficient of variables in this function which is in log form also represents elasticities (Mohammadi and Omid, 2010). Also, it is easy to analyze, and it seems to be a good approximation for actual productions (Singh et al., 2000). The Cobb–Douglas production function is expressed as: Y ¼ f ðxÞ expðuÞ:
ð2Þ
Eq. (2) can be further re-written as: n lnY i ¼ a þ ∑j¼1 α j ln X ij þ ei i ¼ 1; 2; 3;…; n
ð3Þ
where Yi denotes the yield of the ith farmer; Xij, the vector of inputs used in the production process; a, the constant term; αj, represent coefficients of inputs which are estimated from the model and ei, the error term. Assuming that when the energy input is zero, the crop production is zero too, Eq. (3) is reformed to (Singh et al., 2003; Hatirli et al., 2006; Mohammadi and Omid, 2010): n lnY i ¼ ∑j¼1 α j ln X ij þ ei :
ð4Þ
Materials and methods The study was performed in central region of Hamedan province which is located in the west of Iran; within 33° 59′ and 35° 48′ north latitude and 47° 34′ and 49° 36′ east longitude. Hamedan region has an area of 1,949,400 ha; and the farming area, with a share of 51.7%, is
In the present case, n = 8; therefore Eq. (4) can be expressed in the following form: lnY i ¼ α 1 lnX 1 þ α 2 lnX 2 þ α 3 lnX 3 þ α 4 lnX 4 þ α 5 lnX 5 þ α 6 lnX 6 þ α 7 lnX 7 þ α 8 lnX 8 þ ei ð5Þ
86
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
Table 1 Energy equivalent of inputs and output in agricultural production. Inputs (unit)
Unit
A. Inputs 1. Human labor
h
1.96
2. Machinery
h
64.80
3. Diesel fuel
l
56.31
4. Chemical fertilizers (a) Nitrogen
kg
66.14
(b) Phosphate (P2O5) (c) Potassium (K2O) 5. Farmyard manure 6. Biocides 7. Electricity
kg
12.44
kg
11.115
kg
0.30
8. Seed
kg
28.1
B. Output 1. Dry hay (15% w.b.)
kg
15.8
kg kWh
Energy equivalent (MJ unit− 1)
120 11.93a
Reference
Yilmaz et al. (2005); Unakitan et al. (2010) Singh (2002); Hatirli et al. (2006) Rafiee et al. (2010); Unakitan et al. (2010)
Esengun et al. (2007); Mobtaker et al. (2010) Esengun et al. (2007); Mobtaker et al. (2010) Shrestha (1998); Mobtaker et al. (2010) Singh (2002); Bojaca and Schrevens (2010) Mobtaker et al. (2010) Ozkan et al. (2004); Hatirli et al. (2005) Tsatsarelis and Koundouras (1994)
Tsatsarelis and Koundouras (1994)
where MPPxj is marginal physical productivity of jth input; αj, regression coefficient of jth input; GM(Y), geometric mean of yield; and GM(Xj), geometric mean of jth input energy on per hectare basis. A positive value of MPP of any factor indicates that with an increase in input, production is increased and a negative value of MPP of any factor input indicates that additional units of inputs contribute negatively to production. Hence, it is better to keep the variable resource in surplus rather than utilizing it as a fixed resource (Singh et al., 2004). In last part of the study the returns to scale of production was calculated. In production, returns to scale refer to changes in output, subsequent to a proportional change in all inputs (where all inputs increase by a constant factor). In the Cobb–Douglas production function, n P it is indicated by the sum of the elasticities ( α i ) derived in the form of i¼1 regression coefficients. If the sum of the coefficients is greater than unity n P ( α i > 1), then it could be concluded as the increasing returns to scale, n i¼1 P on the other hand if the latter parameter is less than unity ( α i b1), i¼1
then it is indicated as the decreasing returns to scale; and, if the result n P is unity ( α i ¼ 1), it shows as the constant returns to scale (Singh et al., 2004).i¼1 Basic information on energy inputs and alfalfa yields were entered into Excel's spreadsheet and SPSS 17.0 software. Modeling carried out using linear regression technique. Results and discussion
a
This coefficient used according to the efficiency of power plants and power loss of distribution networks reported in references for Iran.
where Xi stands for corresponding energies as X1, human labor; X2, machinery; X3, diesel fuel; X4, chemical fertilizers; X5, farmyard manure; X6, biocides; X7, electricity; and X8, seed. Energy demand in agriculture can be divided into direct energy (DE), indirect energy (IDE), renewable energy (RE) and non-renewable energy (NRE). Direct energy is directly used at farms and on fields. Indirect energy, on the other hand, consists of the energy used in the manufacture, packaging and transport of fertilizers, pesticides and farm machinery (Ozkan et al., 2004). The DE includes human labor, diesel fuel and electricity while the IDE covers chemical fertilizers, farmyard manure, biocides, machinery and seeds used in the alfalfa production. Renewable energy includes human labor, seeds and farmyard manure and non-renewable energy consists of machinery, diesel fuel, chemical fertilizers, biocides and electricity. The effect of direct, indirect, renewable and non-renewable energies on production was also studied. For this purpose, Cobb–Douglas function was determined as Eqs. (6) and (7): ln Y i ¼ β1 ln DE þ β2 ln IDE þ ei
ð6Þ
ln Y i ¼ γ1 ln RE þ γ2 ln NRE þ ei
ð7Þ
where Yi is the ith farm's yield; DE, IDE, RE and NRE are direct, indirect, renewable and non-renewable energies used for alfalfa production, respectively. βi and γi are coefficient of exogenous variables. Eqs. (5)–(7) were estimated using ordinary least square technique. The Marginal Physical Product (MPP) technique, based on the response coefficients of the inputs, was utilized to analyze the sensitivity of energy inputs on alfalfa yield. The MPP which is a factor indicates the change in the output with a unit change in the factor input in question, keeping all other factors constant at their geometric mean level. The MPP of the various inputs was computed using the αj of the various energy inputs as (Singh et al., 2004): MPP xj ¼
GM ðY Þ αj GM X j
ð8Þ
Table 2 shows the quantities of inputs used in whole production life (7 years) and establishment of baled alfalfa hay production and their energy equivalences. Also Fig. 1 shows the distribution percent of the energy associated with the inputs. The results revealed that around 576 h of human labor and 255 h of machinery power per hectare were required to produce alfalfa in the research area. The total energy input for various processes in the alfalfa production was calculated to be 810.57 GJ ha − 1. Tsatsarelis and Koundouras (1994) calculated the energy inputs for alfalfa whole production life and establishment in Greece as 116 GJ ha − 1. The highest average energy consumption of inputs was for electricity (614.37 GJ ha − 1) which was accounted for about 76% of the total energy input (Fig. 1), followed by chemical fertilizers (105.45 GJ ha − 1, 13%). The electricity consumption in alfalfa production is for irrigation purposes. Having deep wells in the region and not using modern efficient irrigation methods are among the reasons of high consumption of electrical energy in the studied region. In order to reduce the electricity consumption, using of modern methods of irrigation with high efficiency (which leads in saving water consumption) can be suggested. Table 2 Amounts of inputs, outputs and energy inputs and output in alfalfa production. Inputs (unit)
Quantity per unit area (ha)
Inputs 1. Human labor (h) 2. Machinery (h) 3. Diesel fuel (L) 4. Chemical fertilizers (kg) (a) Nitrogen (kg) (b) Phosphate (P2O5) (kg) (c) Potassium (K2O) 5. Farmyard manure (ton) 6. Biocide (kg) 7. Electricity (MWh) 8. Seeds (kg) The total energy input (GJ)
575.65 254.53 1144.46 2878.61 1296.90 1575.62 6.09 12.78 18.04 51.50 95.87
Output 1. Dry hay (15% w.b.) (ton) Total energy output (GJ)
96.53
Total energy equivalent (GJ ha− 1) 1.13 16.49 64.44 105.45 85.78 19.60 0.067 3.83 2.16 614.37 2.69 810.57
1525.21 1525.21
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
80.0
614.37
Percentage (%)
70.0 60.0 50.0 40.0 30.0
105.44
20.0 10.0
1.12
16.49
64.44
3.83
2.69
2.16
0.0
Fig. 1. The share and amount (GJ ha − 1 ) of energy inputs for alfalfa production in Hamedan, Iran.
The energy share of nitrogen fertilizer was about 81% of total energy of used chemical fertilizer. As alfalfa is a leguminous and nitrogen fixing plant, the nitrogen fertilizer can be substituted by other fertilizers such as potassium, phosphorus and even farmyard manures. This can lead to less consumption of energy. The energy consumption was minimum for potassium (67.90 MJ ha− 1) which accounted for about 0.01% of the total energy consumption. The consumption of human labor was only 0.14% (1.13 GJ ha − 1). Similar results have been reported in the literature implying that the energy input of human labor has a little share of total energy input in agricultural production (Sartori et al., 2005; Strapatsa et al., 2006; Kizilaslan, 2009). The average yield of alfalfa (for 7 years) was obtained to be 96.53 t ha − 1, accordingly, the total energy output per hectare was calculated as 1525.21 GJ ha − 1 (Table 2). The share of energy input as direct, indirect, renewable and nonrenewable forms is illustrated in Fig. 2. The total consumed energy input could be classified as direct energy (83.88%), and indirect energy (16.12%) or renewable energy (00.95%) and non-renewable energy (99.05%). This indicates that alfalfa production depends mainly on non-renewable energy (electricity, diesel fuel and chemical fertilizers) in the studied area. Therefore, it is clear that non-renewable energy consumption was higher than that of renewable in alfalfa production, which is in agreement with the literatures for different crops (Yilmaz et al., 2005; Erdal et al., 2007; Kizilaslan, 2009; Mobtaker et al., 2010). Since the main non-renewable inputs were electricity and chemical fertilizers, efficient use of water and management of plant nutrients
1.2
802.90
Percentage (%)
1
using renewable resources like farmyard manure would increase the rate of renewable energy. Regression results for Eq. (5) are shown in Table 3. For data used in this study, autocorrelation was tested using Durbin–Watson method (Hatirli et al., 2005; Mohammadi and Omid, 2010). The Durbin–Watson value was found to be 2.671 for Eq. (5) which indicates that there was no autocorrelation at the 5% significance level in the estimated model. The R2 value was determined as 0.99 for this equation, implying that around 0.99 of the variability in the energy inputs was explained by this model. The results of assessment of Cobb–Douglass function on each inputs in alfalfa production indicates that the impact of each inputs differ in constitution of yield. The results revealed that the impact of energy inputs could have positive effect on yield (except for human labor, farmyard manure and biocides). As it can be seen from Table 3, Machinery had the highest impact (0.875) among the other inputs and significantly contributed on the yield at 1% level. It indicates that a 1% increase in the energy machinery input led to 0.875% increase in yield in these circumstances. The second important input was found to be the diesel fuel with 0.193 elasticity, followed by seeds and electricity, with elasticity of 0.135 and 0.032 respectively. Mobtaker et al. (2010) concluded that machinery energy was the most significant input affecting the barley output level in Iran. Hatirli et al. (2006) developed an econometric model for greenhouse tomato production in Antalya province of Turkey and reported that the human labor, chemical fertilizers, biocides, machinery and water energy were important inputs, which significantly contributed to yield. The human labor, farmyard manure and biocides had a negative impact on alfalfa yield. The value of return to scale for the model (1) was calculated by gathering the regression coefficients as 1.136. The higher value of return to scale than unity implies increasing return to scale. The sensitivity of energy inputs on output level was analyzed using the marginal physical productivity method and partial regression coefficients and the results are provided in Table 3. As it can be seen, the major MPP was drown for machinery energy (5.094), followed by seeds energy (4.986). It indicates that using an additional of 1 MJ either for machinery or seeds energy would result in increasing the yield by 5.094 and 4.986 kg, respectively. As a consequence, parameters with a large sensitivity coefficient have a strong influence on the state variable. This identifies which factors should be identified and measured most carefully to assess the state of the environmental system, and which environmental factors should be managed preferentially (Drechsler, 1998). The MPP of human labor, farmyard manure and biocides energy were found to be −9.462, − 1.391 and −2.635; a negative value of MPP implies that additional units of inputs are contributing negatively
Table 3 Econometric estimation results of inputs. Endogenous variable: yield
679.86
87
Coefficient
t-ratio
MPP
0.8
Exogenous variables
0.6
Model 1: ln Yi = α1 ln X1 + α2 ln X2 + α3 ln X3 + α4 ln X4 + α5 ln X5 + α6 ln X6 + α7 ln X7 + α8 ln X8 + ei Human labor − 0.110 − 1.475 − 9.462 Machinery 0.875 8.268⁎ 5.094 Diesel fuel 0.193 2.633⁎⁎ 0.289 Chemical fertilizers 0.025 1.677⁎⁎⁎ 0.028 Farmyard manure − 0.002 − 0.699 − 1.391 Biocides − 0.012 − 3.258⁎ − 2.635 Electricity 0.032 1.276 0.006 Seeds 0.135 3.175⁎ 4.986 Durbin–Watson 1.761 R2 0.998 n P Return to scale ( α i ) 1.136
0.4 0.2
130.61 7.64
0
Fig. 2. The share and amount (GJ ha− 1) of total energy input in the form of direct, indirect, renewable and non-renewable for alfalfa production in Hamedan, Iran.
i¼1
⁎, ⁎⁎, ⁎⁎⁎ indicates significance at 1%, 5% and 10% levels, respectively.
88
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89
to production, i.e. less production with more input. The other important variables that affect alfalfa yield are diesel fuel and chemical fertilizer energy with MPP of 0.289 and 0.028, respectively. Singh et al. (2004) estimated the sensitivity of energy inputs on wheat productivity for five agro-climate zones in India. They reported that MPP of chemicals in zones 1–5, were calculated to be 0.385, 2.816, − 0.211, 0.610 and 0.624, respectively. Rafiee et al. (2010) estimated the sensitivity of energy inputs on apple production in Iran. The results showed that the MPP value of water for irrigation was the highest, followed by human labor and chemicals energy inputs, respectively. The energy obtained from existing inputs was divided into two direct and indirect forms. The assessed trends of both forms of energy were positive, indicating the positive impacts on the output level. Impact of direct energy (0.566) was more than that of indirect energy (0.340). This impact was significant at 1% level. The regression coefficient for renewable energy (0.322) and non-renewable energy (0.644) was significant at 1% level. It is concluded that impact of non-renewable energy was higher than that of renewable energy in Alfalfa production. Similar results have been reported in the literatures (Mohammadi and Omid, 2010; Unakitan et al., 2010). Durbin–Watson values were calculated as 2.290 and 2.371 for Eqs. (6) and (7), respectively; indicating that there is no autocorrelation at the 5% significance level in the estimated models. The R 2 values were found to be 0.99 and 0.96, respectively. As it can be seen from Table 4 the MPP of direct, indirect, renewable and non-renewable energy was found to be 0.087, 0.293, 0.188 and 0.012, respectively. This indicates that an additional use of 1 MJ of each of direct, indirect, renewable and non-renewable energy would lead to an additional increase in yield by 0.087, 0.293, 0.188 and 0.012 kg, respectively. It is concluded that impact of non-renewable energy was higher than that of renewable energy in alfalfa production. Mohammadi and Omid (2010) and Hatirli et al. (2005) have also reported similar results for greenhouse cucumber production in Iran and energy use in Turkish agriculture, respectively. The return to scale values for the models (2) and (3) were 0.906 and 0.966, respectively, implying the decreasing return to scale. Energy management is an important issue in terms of efficient, sustainable and economic use of energy. Optimization is an important way to maximize the amount of productivity, which can significantly impact the energy consumption and production costs. Linear programming is among the ways that can be used to optimize the energy inputs. So, the present study can be extended to determine wasteful uses of energy inputs by inefficient farms and suggest necessary quantities of various inputs should be used by them. More studies in this direction are currently underway.
Table 4 Econometric estimation results of direct, indirect, renewable and non-renewable energies. Endogenous variable: energy output
Coefficient
t-ratio
MPP
Exogenous variables Model 2: ln Yi = β1 ln DE + β2 ln IDE + ei Direct energy Indirect energy Durbin–Watson R2 n P Return to scale ( β i )
0.566 0.340 2.290 0.99 0.906
8.198⁎ 4.324⁎
0.087 0.293
0.322 0.644 2.371 0.96 0.966
6.930⁎ 21.278⁎
0.188 0.012
i¼1
Model 3: ln Yi = γ1 ln RE + γ2 ln NRE + ei Renewable energy Non-renewable energy Durbin–Watson R2 n P Return to scale ( γ i ) i¼1
⁎ Indicates significance at 1%.
Conclusion The purpose of this study was to investigate influences of energy inputs and energy forms on output levels and evaluation of inputs sensitivity for alfalfa production in Hamedan province, Iran. Data were collected from 80 alfalfa farms and the sample volume was determined by random sampling method. Total energy input was found to be 810.57 GJ ha − 1 and total energy output was calculated as 1525.21 GJ ha − 1. It was founded that the electricity was the most energy consuming input followed by chemical fertilizers. The share of non-renewable energy was higher than that of renewable energy consumption. Econometric estimation results revealed that machinery had the highest impact (0.875) among other inputs and significantly contributed on yield at 1% level. The estimated MPP for machinery energy was the biggest among inputs of energy and MPP of Human labor, farmyard manure and biocides energy were negative. The MPP of direct, indirect, renewable and nonrenewable energy were found to be 0.087, 0.293, 0.188 and 0.012, respectively. Optimal consumptions of electricity, chemical fertilizers and other major inputs would be useful not only in reducing negative effects to environment, but also in maintaining sustainability. Lack of soil analysis in the area leads to unconscious usage of chemical fertilizer. In order to reduce the electricity consumption, using of modern methods of irrigation with high efficiency (which leads in saving water consumption) can be suggested. Also it is suggested that new policies are to be taken to reduce the negative effects of energy inputs such as plant, soil and climate pollution. Therefore, analysis of energy consumption is an important task.
References Anonymous. Annual agricultural statistics. Ministry of Jihad-e-Agriculture of Iran; 2007http://www.maj.ir. Bojaca CR, Schrevens E. Energy assessment of peri-urban horticulture and its uncertainty: case study for Bogota, Colombia. Energy 2010;35(5):2109–18. Cetin B, Vardar A. An economic analysis of energy requirements and input costs for tomato production in Turkey. Renew Energy 2008;33:428–33. Drechsler M. Sensitivity analysis of complex models. Biol Conserv 1998;86(3): 401–12. Erdal G, Esengun K, Erdal H, Gunduz O. Energy use and economical analysis of sugar beet production in Tokat province of Turkey. Energy 2007;32(1):35–41. Esengun K, Gunduz O, Erdal G. Input–output energy analysis in dry apricot production of Turkey. Energy Convers Manage 2007;48:592–8. Hatirli SA, Ozkan B, Fert C. An econometric analysis of energy input–output in Turkish agriculture. Renewable Sustainable Energy Rev. 2005;9(6):608–23. Hatirli SA, Ozkan B, Fert C. Energy inputs and crop yield relationship in greenhouse tomato production. Renew Energy 2006;31(4):427–38. Jianbo L. Energy balance and economic benefits of two agroforestry systems in northern and southern China. Agric Ecosyst Environ 2006;116:255–62. Kizilaslan H. Input–output energy analysis of cherries production in Tokat Province of Turkey. Appl Energy 2009;86(7–8):1354–8. Mandal KG, Saha KP, Ghosh PK, Hati KM, Bandyopadhyay KK. Bioenergy and economic analysis of soybean-based crop production systems in central India. Biomass Bioenergy 2002;23(5):337–45. Mobtaker HG, Keyhani A, Mohammadi A, Rafiee S, Akram A. Sensitivity analysis of energy inputs for barley production in Hamedan Province of Iran. Agric Ecosyst Environ 2010;137(3–4):367–72. Mohammadi A, Omid M. Economical analysis and relation between energy inputs and yield of greenhouse cucumber production in Iran. Appl Energy 2010;87(1):191–6. Mohammadi A, Tabatabaeefar A, Shahan S, Rafiee S, Keyhani A. Energy use and economical analysis of potato production in Iran a case study: Ardabil province. Energy Convers Manage 2008;49(12):3566–70. Mohammadi A, Rafiee Sh, Mohtasebi SS, Rafiee H. Energy inputs–yield relationship and cost analysis of kiwifruit production in Iran. Renew Energy 2010;35(5): 1071–5. Ozkan B, Akcaoz H, Fert C. Energy input–output analysis in Turkish agriculture. Renew Energy 2004;29(1):39–51. Rafiee Sh, Mousavi Avval SH, Mohammadi A. Modeling and sensitivity analysis of energy inputs for apple production in Iran. Energy 2010;35:3301–6. Sartori L, Basso B, Bertocco M, Oliviero G. Energy use and economic evaluation of a three year crop rotation for conservation and organic farming in NE Italy. Biosyst Eng 2005;91(2):245–56. Shrestha DS. Energy use efficiency indicator for agriculture; 1998. Available from: http://www.usaskca/agriculture/caedac/PDF/mcrae.
H.G. Mobtaker et al. / Energy for Sustainable Development 16 (2012) 84–89 Singh JM. On farm energy use pattern in different cropping systems in Haryana, India. Master of science. Germany: International Institute of Management, University of Flensburg; 2002. Singh S, Singh S, Pannu CJS, Singh J. Optimization of energy input for raising cotton crop in Punjab. Energy Convers Manage 2000;41(17):1851–61. Singh H, Mishra D, Nahar NM, Ranjan M. Energy use pattern in production agriculture of a typical village in arid zone India: part II. Energy Convers Manage 2003;44(7): 1053–67. Singh G, Singh S, Singh J. Optimization of energy inputs for wheat crop in Punjab. Energy Convers Manage 2004;45(3):453–65.
89
Strapatsa AV, Nanos GD, Tsatsarelis CA. Energy flow for integrated apple production in Greece. Agric Ecosyst Environ 2006;116(3–4):176–80. Tsatsarelis CA. Energy inputs and outputs for soft winter wheat production in Greece. Agric Ecosyst Environ 1993;43:109–18. Tsatsarelis CA, Koundouras DS. Energetics of baled alfalfa hay production in northern Greece. Agric Ecosyst Environ 1994;49(2):123–30. Unakitan G, Hurma H, Yilmaz F. An analysis of energy use efficiency of canola production in Turkey. Energy 2010;35:3623–7. Yilmaz I, Akcaoz H, Ozkan B. An analysis of energy use and input costs for cotton production in Turkey. Renew Energy 2005;30(2):145–55.