Difference in Energy Input and Output in Agricultural Production under Surface Irrigation and Water-Saving Irrigation: A Case Study of Kiwi Fruit in Shaanxi

Abstract

China’s kiwi industry has seen rising production costs and shrinking planting areas in recent years; at the same time, the lack of professional production standards leads to the input redundancy and waste of production factors in the production process of kiwifruit, which intensifies the dilemma of unsustainable agricultural production. This has brought more and more serious challenges to the sustainable development of the industry. In order to solve this problem and clarify the composition and utilization efficiency of energy in the production process of kiwifruit, this study took Chinese kiwifruit production as the research object and analyzed the energy input and output under surface irrigation and water-saving irrigation from the perspective of energy. The results show that the energy input of kiwifruit production under traditional surface irrigation was 85.4 GJ/ha, and the energy output was 59.7 GJ/ha. Among all energy input elements, mineral fertilizers accounted for the highest proportion of energy input, accounting for 48.31%. Under water-saving irrigation, the energy input and output of kiwifruit production are 72.3 GJ/ha and 62.3 GJ/ha; the highest energy input is also mineral fertilizer. The data envelopment analysis results also confirmed that there is a large redundancy in the amount of mineral fertilizer. Compared with surface irrigation, water-saving irrigation technology has effectively improved the energy ratio (from 0.70 to 0.86), energy productivity (from 0.37 kg/MJ to 0.45 kg/MJ) as well as net energy (from −25.8 GJ/ha to −9.93 GJ/ha). Thus, promoting the application of water-saving irrigation technology and increasing the proportion of fertigation during the kiwi production process are necessary measures to promote the sustainable development of China’s kiwi industry.

1. Introduction

Agricultural production is one of the most ecologically impactful human activities. The Food and Agriculture Organization (FAO) points out that a high external input and resource-intensive agricultural systems cause large-scale deforestation, water scarcity, biodiversity loss, soil degradation and greenhouse gas emissions [1]. Agricultural production consumes 70% of the water consumption of the whole society and produces 13% of greenhouse gases. Effectively improving the efficiency of resource utilization and promoting resource recycling is a key and urgent issue [2]. The northern foothills of Qinling Mountains in Shaanxi Province, China, is the eugenic area and the main planting area for kiwifruit. As of the end of 2017, the kiwifruit planting scale in Shaanxi Province reached 69,000 hm², and the output reached 1.31 million t, accounting for 55.3% of the national output and one-third of the world’s total output. It has become the world’s largest kiwifruit production base [3].

Since 2016, the development of the kiwifruit industry in Shaanxi Province has encountered a bottleneck. Due to the influence of farmers’ cognitive limitations on the input of production factors, ineffective decisions regarding the input of production factors have been made in production activities. This has caused the production of kiwifruit to be in a state of wasteful cost, and it is difficult to improve the production efficiency, which has seriously affected the sustainability of the development of the kiwifruit industry [4]. According to the data released by Shaanxi Provincial Bureau of Statistics [3,5,6], the kiwifruit production and planting area of kiwifruit trees in Shaanxi province continued to rise from 2011 to 2017 but fell sharply in 2018 (Figure 1a). The increasing input year by year and the decrease in operating income have affected the confidence of growers and have caused the destruction of orchards and the abandonment of planting. As can be seen from Figure 1b, the production cost has increased year by year since 2011. At the same time, the net income reached the peak in 2016 and continued to decrease in the following years, falling to CNY 74,011/ha in 2019, basically the same as that in 2011. Therefore, it is an urgent issue to optimize the number of input factors of production as well as the related management strategies.

On the other hand, in order to solve the shortage of water resources and ensure the security of food production, water-saving irrigation technologies (drip irrigation, sprinkler irrigation, etc.) have been rapidly developed and applied in agricultural production in China, with the policy guidance and financial support of the Chinese government [7]. As a result, the utilization coefficient of irrigation water has increased significantly, from 0.5 in 2011 to 0.554 in 2018, and it also shows obvious advantages in terms of saving labor and increasing production [8,9]. However, the application of water-saving irrigation systems has caused an increase in initial investment [10,11] as well as an increase in energy consumption for water lifting due to the higher operating pressure of the irrigation system [12,13]. It can be concluded that the large-scale promotion and application of water-saving irrigation technology have a significant impact on the input of various agricultural production factors and the yields. Since the 12th Five-Year-Plan, kiwifruit irrigation in Shaanxi Province has been largely transformed from traditional ground irrigation to water-saving irrigation, such as drip irrigation or micro-spray irrigation [14]. However, due to the low education level of users and the lack of effective guidance in the operation stage, the management and operation level of the water-saving irrigation system is low, and the application effect is poor [15,16]. However, at present, there is still a lack of quantitative assessment on the impact of the water-saving irrigation technology application and management level on the input of kiwifruit production factors and the sustainable development of kiwifruit production.

The input and yield analysis of production factors based on an energy perspective is a method that is often used to evaluate agricultural production efficiency [17,18,19]. Different production input elements and crop yields are quantified into evaluation indicators based on the energy perspective, and production efficiency evaluation is carried out so that producers can form an objective understanding of the production level under different production modes and crop types, which is an important method for promoting sustainable agricultural production [20,21,22]. Compared with conventional economic analysis, evaluation based on an energy perspective is less susceptible to price fluctuations of input and output factors [23], which helps to provide useful, comprehensive information for decision makers [24], so this method has been widely used in evaluating the energy use efficiency and production sustainability of field crops, vegetables, economic forest fruits, etc. [25,26,27,28,29]. Such studies generally focus on the composition of energy inputs in the production process and the energy ratio of each production factor [24,30]. A sensitivity analysis has been conducted on the impact of inputs of various production factors on the yield [31,32,33]. The crop yield and greenhouse gas emissions have been forecast based on the relationship between the energy input and output [34,35,36] with the commonly used methods of Multiple Linear Regression [31,37], Artificial Neural Network [38,39]), etc. The energy use efficiency of different production factor input schemes [24,40] has been evaluated with the commonly used method of Data Envelopment Analysis [41,42,43].

The analysis of the input and output of production factors based on an energy perspective provides a research idea for solving the sustainable problem of kiwifruit production in Shaanxi under the background of the rapid development of water-saving irrigation technology. Several studies have been carried out on the energy input and output of kiwifruit [35,38]. The results show that, even for the same crop type, when it is located in different producing areas, there may be huge differences in the energy structure composition of the input production factors, which is closely related to the local production management level, meteorological conditions, water resources and so on. As the largest kiwifruit production base in the world, Shaanxi Province still lacks quantitative research on the energy input and output and energy utilization efficiency of kiwifruit under different irrigation modes. On the other hand, previous studies on energy utilization efficiency lacked the evaluation and sequencing of technical schemes and did not propose the optimal technical scheme suitable for local production conditions and production modes. This study is mainly aimed at clarifying the energy composition and energy utilization efficiency of input production factors in the planting of kiwifruit in Shaanxi, analyzing the impact of water-saving irrigation technology applications on energy input and output efficiency and comprehensively evaluating the energy utilization efficiency of kiwifruit production under different irrigation technologies. Through energy evaluation, the optimal input scheme of production factors under different irrigation technology modes is screened, which provides a reference for the efficient use of energy and the sustainable production of Shaanxi kiwi.

2. Materials and Methods

In this study, we first obtained the data on the input and output factors of kiwifruit production under traditional ground irrigation and water-saving irrigation technology through the household questionnaire method, and then a series of energy characteristic parameters were extracted and calculated through the survey data. On this basis, a comprehensive evaluation of energy efficiency based on data envelopment analysis was carried out to analyze the impact of the application of water-saving irrigation technology on energy utilization efficiency in the production process of kiwifruit and to determine the optimal factor input scheme under different irrigation methods.

2.1. Data Acquisition

This study was conducted in Zhouzhi County and Mei County, Shaanxi Province, located at the northern foot of the Qinling Mountains (108°4′28″ E, 34°16′56″ N). According to the statistics on the website of the Shaanxi Provincial Fruit Industry Bureau, the kiwifruit planting area in these two counties is 33,306 ha, with an annual output of 725,700 tons, accounting for 62.65% of the province’s total planting area and 76.56% of the total output [6]. Therefore, the irrigation technology and management mode adopted in kiwifruit orchards in this area are representative. The region has a temperate continental climate, with an average altitude of 521 m, a frost-free period of 210 days and an annual evaporation and annual rainfall of 1500 mm and 632 mm, respectively’ the rainfall is mainly concentrated from July to September each year.

Regarding the questionnaire adopted in this study to reflect the input and output factors in kiwifruit production, three aspects were specifically investigated and considered. First, the kiwifruit planting technical manual was consulted, and the kiwifruit planting input and production factors listed in the manual were summarized [44]). Second, pre-research was carried out, and the input elements in production were summarized through face-to-face communication with kiwifruit growers. Third, the questionnaire used by [34,35] in their research on kiwi fruit in Iran was also referred to. In the questionnaire, the technology used for irrigation was clearly identified as surface irrigation and water-saving irrigation (drip irrigation, micro-sprinkler irrigation). In this survey, a household face-to-face survey was adopted to conduct a questionnaire survey on 60 households of kiwifruit orchard growers in the above-mentioned two regions, among which 30 households used surface irrigation and the others used water-saving irrigation. In the preliminary analysis of the questionnaire results, it was found that the varieties of kiwifruit grown in different orchards were diverse, mainly including Xuxiang, Hayward, Qin Mei, Hong Yang and so on. Considering that different varieties of kiwifruit often have large differences in yield (this was not reflected or explained in the studies of [34,35], in order to avoid the yield differences caused by different kiwifruit varieties, the questionnaire was screened, and the kiwifruit orchards that only planted a single variety of Xuxiang were used as valid survey objects (the planting area of this variety accounted for more than 60% of the total planting area, and it was the main kiwifruit variety in Shaanxi Province). A total of 31 valid questionnaires were obtained, of which 15 were ground irrigation and 16 were water-saving irrigation.

2.2. Feature Parameter Extraction

2.2.1. Quantification of Production and Investment Factor Indicators

The input and output elements of the whole process of kiwifruit production are shown in the first column of

Table 1. Energy equivalents of inputs and outputs in agricultural production.

Production Process	Unit	Energy Equivalents (MJ Unit−1)	References
A. Input
Human Labor	h	1.96	[34,47]
Machinery	h	62.7	[30]
Diesel fuel	L	56.31	[45]
Electricity	kW·h	11.93	[25,45,48]
Mineral fertilizer
(a) Nitrogen	kg	66.14	[25,34,35]
(b) Phosphate (P2O5)	kg	12.44	[31,34,35]
(c) Potassium (K2O)	kg	11.15	[34,35,49]
Farmyard Manure	kg	0.3	[34,35,50]
Biocides
(a) Fungicide	kg	199	[51,52]
(b) Insecticide	kg	92	[51,52]
WSIS
(a) Head system	kg	148	[51,53]
(b) Pipeline	kg	112	[51,53]
B. Output
Kiwifruit	kg	1.9	[20,34]

Note: WSIS-Water-saving irrigation system.

, and the corresponding values are obtained through questionnaires. In order to carry out a unified evaluation from the perspective of energy, it is necessary to quantify each element, that is, on the basis of the data obtained from the questionnaire, multiply each element by an energy conversion factor. Based on the previous research results, this paper determines the energy conversion coefficient of each element, as shown in the third column of Table 1. It should be pointed out that the input factors in Table 1 do not list the energy consumption of irrigation water but instead use the energy consumption of electricity. According to the research of [34,45], the energy conversion factor of irrigation water was taken as 1.02 MJ/m³, which can not indicate the impact of irrigation technology differences on the energy consumption of irrigation water. Water-saving irrigation technologies such as drip irrigation and sprinkler irrigation operate under higher pressure, and the energy consumption per unit volume of water is significantly higher than that of surface irrigation [13]. This study focuses on the production energy input under different irrigation methods, so a unified 1.02 MJ/m³ cannot be used as the energy conversion standard for irrigation water. During the investigation, it was found that all the electricity consumption of the orchard was consumed by groundwater extraction for irrigation, and the electricity consumption and electricity bills were recorded in detail. Therefore, the energy consumption of irrigation water was replaced by electricity consumption to express the difference in energy consumption under different irrigation technologies. In addition, the conversion of mechanical equipment energy needs to consider the operating time, the weight, the average economic life of the mechanical equipment and the equivalent energy consumption per unit weight of the equipment. The calculation formula is as follows [46]:

$${\mathrm{Machinery} \mathrm{energy} (\mathrm{MJha}}^{-1}) ={ (\mathrm{T}}_{\mathrm{M}}{\mathrm{E}}_{\mathrm{M}}{\mathrm{W}}_{\mathrm{M}}{) / \mathrm{L}}_{\mathrm{M}}$$

(1)

where T_M is the running time of machinery (h/ha); E_M is the equivalent energy consumption of machinery (93.61 MJ/kg for tractors, 87.63 MJ/kg for combine harvesters and 62.7 MJ/kg for other types of agricultural machinery) [30]; W_E (kg) is the weight of agricultural machinery, which can be obtained by querying the machine nameplate or product manual; L_M is the service life of agricultural machinery (12,000 h for tractors, 3000 h for combine harvesters, 3000 h for rotary tillers, 1500 h for drilling machines and 15,000 h for ridge machines) [46].

2.2.2. Energy Characteristic Parameters

Based on the quantified input and output production factor indicators, the relationship between the energy input and output during kiwifruit production can be quantitatively described by energy characteristic parameters, including the Energy productivity (E_p), Specific energy (E_s), Energy ratio (E_r) and Net energy gain (NEG) [46]. The calculation formulas for each parameter are as follows.

$$\mathrm{Energy}\hspace{0.33em}{\mathrm{productivity} (\mathrm{E}}_{\mathrm{P}}{,\mathrm{kg} \mathrm{MJ}}^{-1}) =(\frac{\mathrm{Kiwifruit} \mathrm{yield},{ \mathrm{kg} \mathrm{ha}}^{-1}}{\mathrm{Total} \mathrm{energy}\hspace{0.33em}{\mathrm{input}, \mathrm{MJ} \mathrm{ha}}^{-1}})$$

(2)

$${\mathrm{Specific} \mathrm{energy} (\mathrm{E}}_{\mathrm{S}}{, \mathrm{MJ} \mathrm{kg}}^{-1}) =(\frac{\mathrm{Total} \mathrm{energy}\hspace{0.33em}{\mathrm{input}, \mathrm{MJ} \mathrm{ha}}^{-1}}{\mathrm{Kiwifruit} \mathrm{yield},{ \mathrm{kg} \mathrm{ha}}^{-1}})$$

(3)

$${\mathrm{Energy} \mathrm{ratio} (\mathrm{E}}_{\mathrm{R}}) =(\frac{\mathrm{Total} \mathrm{energy}\hspace{0.33em}{\mathrm{output}, \mathrm{MJ} \mathrm{ha}}^{-1}}{\mathrm{Total} \mathrm{energy}\hspace{0.33em}{\mathrm{input}, \mathrm{MJ} \mathrm{ha}}^{-1}})$$

(4)

$$\begin{array}{ll}{\mathrm{Net} \mathrm{energy} \mathrm{gain} (\mathrm{NEG}, \mathrm{MJ} \mathrm{ha}}^{-1}) & =(\mathrm{Total} \mathrm{energy}\hspace{0.33em}{\mathrm{output}, \mathrm{MJ} \mathrm{ha}}^{-1})\\ & -\hspace{0.33em}(\mathrm{Total} \mathrm{energy}\hspace{0.33em}{\mathrm{input}, \mathrm{MJ} \mathrm{ha}}^{-1})\end{array}$$

(5)

In addition, in order to clarify the structure of input energy, the factors of production input energy are listed in Table 1, which can be divided into direct energy and indirect energy, as well as renewable energy and non-renewable energy [20]. Among them, direct energy includes labor, diesel and electricity, indirect energy includes pesticides, inorganic fertilizers, organic fertilizers and machinery, renewable energy includes labor and organic fertilizers and non-renewable energy includes diesel, inorganic fertilizers, pesticides and machinery [54].

2.3. Evaluation of Energy Utilization Efficiency

In the absence of unified guidance standards, for the same planting area, although the water sources, climate, soil types and other conditions are almost the same, growers often have their own planting and management preferences, which are consistent with the growers’ own cognition, management level, personal habits, etc. Therefore, different energy inputs and outputs will also be caused. The rational allocation of energy during the production process helps to improve production efficiency [55] and economic benefits [56], as well as improve the sustainability of crop production systems by saving resources [57]. The data envelopment analysis method is often used for energy utilization efficiency evaluation and energy allocation optimization [58]. This method can evaluate multiple input and output indicators without assuming a specific production function and has been applied to the optimal allocation of energy in the production of rice, soybean, cucumber, strawberry and other crops [25,40,59]. This paper adopts the CCR-DEA model [60,61] and the ACE-DEA (adversarial cross-evaluation-DEA) model to evaluate the energy utilization efficiency and configuration optimization of kiwifruit. The CCR-DEA model can be expressed as follows [24]:

$$Max\hspace{0.33em}{h}_j={\displaystyle \sum _{r=1}^s{\mu }_r{y}_{rj}/{\displaystyle \sum _{i=1}^m{v}_i{x}_{ij}}}\hspace{1em}s.t.\hspace{0.33em}{\displaystyle \sum _{r=1}^s{\mu }_r{y}_{rj}-{\displaystyle \sum _{i=1}^m{v}_i{x}_{ij}}}\le 0$$

(6)

$${\mu }_r\ge 0,\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}{v}_i\ge 0$$

where h_j is the comprehensive efficiency index (CEI) of the jth DMU; x_ij and y_rj are, respectively, the ith input indicator and the rth output indicator of the jth DMU; v_i and μ_r are the weight coefficients of the ith input indicator and the rth output indicator, respectively, which are determined by the relationship between the input and output indicators; and s and m indicate the number of output and input indicators, respectively.

The ACE-DEA model can avoid the defect that the CCR-DEA model is dominated by self-evaluation, which is prone to producing false DEA effective units. The best weight of each DMU_j is used to calculate the efficiency value of other DMU_ks, and the cross-evaluation value E_jk is obtained, as shown in the following formula.

$$E_{jk}=\frac{{\displaystyle \sum _{r=1}^s}{\mu }_{rj}^{*}{y}_{rk}}{{\displaystyle \sum _{i=1}^m}{v}_{ij}^{*}{x}_{ik}},\hspace{1em}k,j=1,2,\dots ,n$$

(7)

Taking the max ${\displaystyle \sum _{r=1}^s{u}_r{y}_{rj}}$ as the first objective and the min ${\displaystyle \sum _{r=1}^s{\mu }_{rj}{y}_{rk}/{\displaystyle \sum _{i=1}^m{v}_{ij}{x}_{ik}}}$ as the second objective, an adversarial cross-evaluation is established, and the cross-evaluation value constitutes the cross-efficiency matrix E:

$$E=\left[\begin{array}{l}{E}_{11}\hspace{0.33em}\hspace{0.33em}{E}_{12}\hspace{0.33em}\hspace{0.33em}\dots \hspace{0.33em}\hspace{0.33em}{E}_{1n}\\ E_{21}\hspace{0.33em}\hspace{0.33em}{E}_{22}\hspace{0.33em}\hspace{0.33em}\dots \hspace{0.33em}\hspace{0.33em}{E}_{2n}\\ \hspace{0.17em}\dots \hspace{0.33em}\hspace{0.33em}\dots \hspace{0.33em}\hspace{0.33em}\dots \hspace{0.33em}\hspace{0.33em}\dots \\ E_{n1}\hspace{0.33em}\hspace{0.33em}{E}_{n2}\hspace{0.33em}\hspace{0.33em}\dots \hspace{0.33em}\hspace{0.33em}{E}_{nn}\end{array}\right]$$

(8)

Take the average of the self-evaluation efficiency value on the main diagonal and the remaining n − 1 cross-evaluation efficiency values as a measure of the efficiency level of unit j. The larger e_j is, the better the decision-making unit j is, realizing the full ordering of the evaluated decision-making units.

2.4. Data Range Standardization

In order to maintain data consistency and facilitate evaluation and analysis, interval standardization can be carried out for each indicator so that all indicators can be converted into values of the [0, 1] interval. The range standardization formula is as follows. In the evaluation, the expectation of the input index is that the smaller the better, it belongs to the inverse index and the reciprocal processing is performed before the data standardization.

$$A_{ij}=\frac{X_{ij}-\min (X_{ij})}{\max (X_{ij})-\min (X_{ij})}$$

(9)

DAEP Version 2.1 and MATLAB (The MathWorks Inc., Natick, MA, USA) were used for the CCR-DEA model and the ACE-DEA model.

3. Results Analysis

3.1. Quantification of Production and Investment Factor Indicators

The energy input composition and proportion of kiwifruit orchards using surface irrigation (Figure 2a) and water-saving irrigation (Figure 2b) are shown in Figure 2. It can be seen in Figure 2c that the total energy input under the surface irrigation method is 85.4 GJha⁻¹, of which the largest energy input is inorganic fertilizer, accounting for 48.31%, followed by electricity and labor, accounting for 13.74% and 12.19%, respectively. The smallest proportion of energy input is organic fertilizer, accounting for 2.58%. The total energy input under the water-saving irrigation method is 72.3 GJha⁻¹, showing a decrease of 15.42% compared with the total energy input of the surface irrigation method. The largest proportion of the energy input is still inorganic fertilizer, accounting for 35.7%, followed by the materials of the water-saving irrigation system and electricity. The proportion of the energy input is 16.06% and 13.73% respectively. The smallest proportion of the energy input is organic fertilizer, which accounts for 1.25%.

By comparing the energy inputs of the two irrigation methods, it can be seen that the energy inputs of labor, irrigation electricity, pesticides, inorganic fertilizers, organic fertilizers and fossil fuels are reduced to varying degrees after the water-saving irrigation technology is adopted. Among them, the energy input of organic fertilizer decreased the most from 0.221 GJha⁻¹ to 0.091 GJha⁻¹, a decrease of 59%, and the energy input of inorganic fertilizer decreased from 41.3 GJha⁻¹ to 25.8 GJha⁻¹, a decrease rate of 37.5%.

It is worth noting that the energy consumption of electricity consumed by irrigation was reduced from 11.7 GJha⁻¹ to 9.92 GJha⁻¹, a decrease of 15.5%. According to the data recorded by the flow meter, the average irrigation water consumption per unit area of surface irrigation is 7891 m³/ha, while the water consumption per unit area of water-saving irrigation is only 3103 m³/ha, a decline of 60.7%. Correspondingly, the electricity consumption per unit volume of water under surface and water-saving irrigation is 0.25 kWh/m³ and 0.54 kWh/m³ respectively, which also confirms that pressurized water-saving irrigation significantly increases the energy consumption for water extraction. Therefore, the conversion factor of water energy per unit volume is different under different irrigation methods. It can also be seen in Figure 2c that, after the water-saving irrigation technology is adopted, the mechanical energy consumption increases by 12%, and the energy consumption of the materials of the water-saving irrigation system itself is about 11.6 GJha⁻¹, which is unique to the water-saving irrigation method.

In order to explore the differences between the energy input and output factors of different growers under the two different irrigation methods, a box diagram is drawn (Figure 3). Figure 3 shows that the median value of each input factor has the same trend as that in Figure 2c. In addition to the mechanical energy and the energy consumption of the water-saving irrigation system itself, the energy input values of other production factors after water-saving irrigation showed a decrease, and the energy value corresponding to the kiwifruit yield increased significantly. From the scatter plot and normal curve of each energy element shown in the figure, it can be seen that, regardless of whether traditional ground irrigation or water-saving irrigation is used, the input of the same type of production factor shows a great difference among different growers, which reflects that obvious personal preferences exist during the production process.

3.2. Energy Characteristic Parameters

The energy characteristic parameters of surface irrigation and water-saving irrigation in Shaanxi Province, China, are listed in

Table 2. Comparison of energy characteristic parameters under different irrigation methods.

Energy Indices	Unit	Without WSIS	With WSIS	Guilan #	Mazandaran ##
Total energy input	MJha−1	85,444.07	72,268.24	104,156.03	30,285.62
Total energy output	Mjha−1	59,660	62,341.2	49,512.14	46,639.85
Energy ratio	—	0.7	0.86	0.48	1.54
Energy productivity	kgMJ−1	0.37	0.45	0.25	0.81
Specific energy	MJkg−1	2.72	2.2	4.01	1.23
Net energy	Mjha−1	−25,784.07	−9927.05	−54,643.89	16,354.23
Direct energy	Mjha−1	30,319.98	22,909.84	73,088.3	9110.45
Indirect energy	Mjha−1	55,124.09	49,358.41	31,067.74	21,275.17
Renewable energy	Mjha−1	12,624.18	6860.83	21,826.2	7713
Non-renewable energy	Mjha−1	72,819.89	65,407.42	82,329.84	22,572.62

Notes: ^# Soltanali et al., 2017 [38]; ^## Mohammadi et al., 2010 [64]; WSIS—Water-saving irrigation system.

. As a comparison, the table also lists the research results of energy investment in kiwifruit production in Guilan Province ([38] and Mazandaran Province, Iran [34]. The energy inputs of surface irrigation and water-saving irrigation in Shaanxi Province were 85.4 and 72.3 GJ/ha, while the energy inputs of kiwifruit production in the Guilan and Mazandaran provinces of Iran were 104.2 and 30.3 GJ/ha, respectively. This indicates that the characteristics of the production area and the production management mode and level have a significant impact on the total energy input of agricultural production. As reported by [34], a high electricity consumption in agricultural production is the most important reason for the higher energy input for kiwifruit production in Guilan province.

The output energy of kiwifruit production under surface irrigation and water-saving irrigation in Shaanxi Province were 59.6 and 62.3 GJ/ha, respectively, indicating that the application of water-saving irrigation technology can reduce the input energy while increasing yield. Therefore, water-saving irrigation technology turns out to be effective in improving agricultural production efficiency, which is consistent with the conclusions of [62]. From the perspective of other energy characteristic parameters, when the surface irrigation was changed to water-saving irrigation, the energy ratio and energy productivity increased from 0.70 and 0.37 kgMJ⁻¹ to 0.86 and 0.45 kgMJ⁻¹, respectively, and the specific energy decreased from 2.72 MJkg⁻¹ to 2.20 MJkg⁻¹. The changes in these characteristic parameters all reflect that the application of water-saving irrigation technology can optimize the energy structure of kiwifruit production and improve the energy utilization efficiency. In the provinces of Guilan and Mazandaran in Iran, the energy output of kiwifruit is 46.6–49.5 GJ/ha, which is slightly lower than that in Shaanxi, China. On the one hand, this may relate to the management level of the orchard; on the other hand, it may be affected by the differences in kiwifruit varieties.

In addition, the net energy under both irrigation techniques was negative, reaching −25.8 and −9.93 GJ/ha, respectively. As a comparison, the net energy of kiwifruit production in Guilan Province was −54.6 GJ/ha, while that of kiwifruit production in Mazandaran Province was 16.3 GJ/ha. Some related research shows that, as opposed to cereal crops, vegetables, melons and fruits belong to the types of a high input and low output, and the net energy is generally negative [28]. For example, the net energy of cucumber and strawberry was −58.5 GJ/ha [28] and −26.7 GJ/ha [40], respectively, while the net energy of maize and wheat could reach 95.6 GJ/ha [63] and 41.7 GJ/ha [28].

The research results also show that the direct energy consumption of surface irrigation and water-saving irrigation is 30.3 GJ/ha and 22.9 GJ/ha, respectively, accounting for 35.5% and 31.7% of the total energy input. The application of water-saving irrigation technology can reduce direct energy consumption, which is consistent with the research results of [62]. In wheat production, the proportion of direct energy under surface irrigation, sprinkler irrigation and drip irrigation is 55%, 49.79% and 34.38%, respectively. In addition, the proportion of the renewable energy input in Shaanxi kiwifruit production was 14.8% (surface irrigation) and 9.5% (water-saving irrigation).

3.3. Evaluation of Energy Utilization Efficiency

3.3.1. Evaluation of the CCR-DEA Model

The energy input and output indicators of 31 valid questionnaires were brought into the CCR-DEA model, and the comprehensive evaluation efficiency of the energy utilization of kiwifruit planting under the two irrigation methods was obtained, as shown in

Table 3. Comprehensive evaluation efficiency of kiwifruit planting energy utilization under different irrigation methods.

DMU No.	CRSTE	DMU No.	CRSTE	DMU No.	CRSTE	DMU No.	CRSTE
N1	1.000	N9	1.000	S16	1.000	S24	0.772
N2	1.000	N10	0.950	S17	0.848	S25	1.000
N3	1.000	N11	1.000	S18	1.000	S26	0.983
N4	0.970	N12	1.000	S19	0.916	S27	1.000
N5	0.868	N13	1.000	S20	1.000	S28	0.961
N6	1.000	N14	1.000	S21	1.000	S29	1.000
N7	0.771	N15	1.000	S22	1.000	S30	1.000
N8	1.000			S23	0.737	S31	0.733

. In this table, N1–N15 are decision-making units (DMUs) using surface irrigation, of which 11 are efficient DMU, accounting for 73.3% of the total number of decision-making units. The number of inefficient DMUs is four, of which the most inefficient DMU is N7, and its comprehensive evaluation efficiency value is 0.771. In this table, S1–S16 are DMUs that use water-saving irrigation, of which the number of effective DMUs is nine, accounting for 56.3% of the total number of decision-making units. The number of inefficient DMUs is six, of which the most inefficient DMU is S31, and its comprehensive evaluation efficiency value is 0.733.

For the inefficient DMUs, when the value of each input indicator increases by the same percentage, the increase in output energy will be less than this percentage, which reflects the redundancy of the input index, and the comprehensive energy utilization can be improved by reducing the value of each input indicator [65]. The proportion of inefficient DMUs using water-saving irrigation technology reached 43.7%, while the proportion of ineffective DMUs using surface irrigation was 26.7%, which shows that there is an obvious excess input of production materials in the production process of kiwifruit. This is also an important reason for the increase in the production cost of kiwifruit year by year since 2011 (Figure 1b). Although the application of water-saving irrigation technology can optimize the energy structure of kiwifruit production and improve energy utilization efficiency on the whole, the effectiveness of its energy utilization efficiency is lower than that of surface irrigation (56.3% vs. 73.3%), which may be related to the significant differences in the management level of users who adopt water-saving irrigation technology.

Project the inefficient DMUs on the effective frontier to obtain the optimal adjustment amount of the energy input indicator of each inefficient DMU, as shown in

Table 4. Energy saving in different inputs for an inefficient DMU, if recommendations are followed.

Type	Particular	Average Energy Use	Average Energy Saving	% of Total Saving	% of Total Energy Used
		MJ ha−1	MJ ha−1
Without WSIS	Human labor	11,507.1	1060.7	15.0	1.2
	Biocides	6309.3	201.7	2.9	0.2
	Mineral fertilizer	39,253.3	425.0	6.0	0.5
	Farmyard manure	2921.0	0.0	0.0	0.0
	Electricity	12,827.7	3007.1	42.7	3.4
	Machinery	6168.0	1540.7	21.9	1.8
	Diesel	8498.0	815.4	11.6	0.9
	WSIS	0.0	0.0	0.0	0.0
	Total	87,484.4	7050.6	100.0	8.1
With WSIS	Human labor	6538.6	529.9	3.74	0.7
	Biocides	4292.2	300.6	2.12	0.4
	Mineral fertilizer	26,540.3	7913.3	55.83	10.7
	Farmyard manure	685.4	98.6	0.70	0.1
	Electricity	10,370.5	2326.0	16.41	3.2
	Machinery	6555.1	1577.7	11.13	2.1
	Diesel	7307.6	832.9	5.88	1.1
	WSIS	11,543.7	594.7	4.20	0.8
	Total	73,833.4	14,173.7	100.0	19.2

Note: WSIS—Water-saving irrigation system.

. For the inefficient DMUs using surface irrigation, the average energy input is 87.5 GJ/ha. After optimization, the average energy input can be reduced by 7.1 GJ/ha, a decrease of 8.1%, of which the electric energy input has the largest decrease, accounting for 42.7% of the total energy reduction. For the inefficient DMUs using water-saving irrigation, the average energy input is 73.8 GJ/ha. After optimization, the average energy input can be reduced by 14.2 GJ/ha, a decrease of 19.2%. Among them, the energy input of mineral fertilizers has the largest decrease, accounting for 55.8% of the total energy reduction. The results show that the energy-saving potential of the water-saving irrigation mode is higher than that of surface irrigation. On the other hand, the power consumption redundancy of irrigation is high under surface irrigation, which means there is a large potential to reduce the amount of irrigation water (the energy of electricity was adopted to estimate the irrigation energy consumption). In contrast, the application of mineral fertilizers in the water-saving irrigation mode has a high redundancy, This is mainly because when farmers use the water-saving irrigation system, they add small fertilizing devices in the field, realizing the precise control of fertilization in the production process and reducing the energy input of the inorganic fertilizer, thus reducing the energy input of water-saving irrigation. Thus, reducing the use of inorganic fertilizers may be an important way to improve the energy efficiency of kiwifruit production.

3.3.2. Evaluation of the ACE-DEA Model

For the DMUs evaluated as effective by the CCR-DEA model, range standardization was performed on the input and output indicators, and the scores of each indicator were obtained, as shown in Figure 4. The percentage system is used for the scoring of each sub-item. For the input indicator, the lower the energy input, the higher the score, while for the output indicator, the higher the energy output, the higher the score. In Figure 4, N1–N15 represent the surface irrigation mode and S16–S30 represent the water-saving irrigation mode; there is no energy input for the water-saving irrigation system, and this item is recorded as 100 points. It can be seen in the figure that, although these schemes are all efficient DMUs (CRSTE = 1), there are still obvious differences between the energy input and energy output of each DMU; the input quantities of various production factors have obvious personal preferences and lack a unified standard.

The efficient DMUs were sorted by the adversarial cross-evaluation method, and the cross-evaluation efficiency values are shown in Figure 5a. Under the surface irrigation method, the orchards with the highest and lowest evaluation scores are N1 (0.69) and N12 (0.381). The orchards with the highest and lowest evaluation scores are S21 (0.699) and S20 (0.286). According to the results of the questionnaire survey and cross-evaluation, for the kiwifruit growers who use ground irrigation and water-saving irrigation, the recommended production factor input schemes are shown in

Table 5. Recommended inputs of various production factors under different irrigation methods.

Particular	Human Labor	Biocides	Mineral Fertilizer	Farmyard Manure	Electricity	Machinery	Diesel	WSIS
	h/ha	kg/ha	kg/ha	kg/ha	kWh/ha	kg/ha	l/ha	kg
Without WSIS	5683.8	24.3	864.0	2495.99	620.00	113.13	115.37	0
With WSIS	1942.8	24.7	488.6	956.25	600.00	74.58	79.56	102.24

Note: WSIS—Water-saving irrigation system.

. It can be seen in the table that, compared with surface irrigation, the use of water-saving irrigation in kiwi fruit orchards can greatly reduce the amount of human labor and farmyard manure by 65.8% and 61.7%, respectively. At the same time, it can effectively reduce the amount of mineral fertilizers, machinery and diesel fuel, with decline rates of 43.4%, 34.1% and 31%, respectively. The consumption of pesticides and electricity consumption are basically the same under different irrigation methods. In addition, the weight of water-saving irrigation equipment (pumps, controllers, pipes, filters, etc.) needs to be increased by about 102.24 kg per hectare.

The radar charts under the four DMUs of N1, N12, S20 and S21 are shown in Figure 5. It can be seen that N1 has a better performance than N12 in terms of the indicators of fuel, machinery, electricity and organic fertilizer (Figure 5b). Although both N1 and N12 are efficient DMUs, as evaluated by the CCR-DEA model, the result of adversarial cross-evaluation shows that the production factor input plan of N12 seems to be more reasonable. Similarly, among the growers who adopted the water-saving irrigation technology, S21 had the highest adversarial cross-evaluation efficiency value, and the energy consumption of human labor, mineral fertilizer, farmyard manure, electricity, machinery and diesel fuel performs better than S20. The performance of diesel is better than that of S20, but S21 consumed more pesticides and has a higher input in water-saving irrigation systems during kiwifruit production (Figure 5c). The comparison between the optimal input schemes of water-saving irrigation and surface irrigation is shown in Figure 5d. Although the application of water-saving irrigation technology increases the system’s own energy consumption, the energy inputs of labor, mineral fertilizer, farmyard manure, machinery and diesel fuel have been significantly reduced, which further proves that the application of water-saving irrigation technology can improve energy utilization efficiency and play a positive role in promoting the sustainable development of kiwifruit production.

4. Discussion

4.1. Huge Energy Saving Potential of Mineral Fertilizers

According to the results of this study, the energy input of mineral fertilizers has the highest proportion in both surface irrigation and water-saving irrigation modes, among which nitrogen fertilizer is the most important type of energy input. In this study, the average amount of mineral fertilizer consumed under surface irrigation is 1038.9 kg/ha, of which the amount of nitrogen fertilizer is 535.6 kg/ha, and the equivalent energy input of nitrogen fertilizer is 35.4 GJ/ha, accounting for 85.8% of the total energy input of mineral fertilizers. The average amount of mineral fertilizer consumed under water-saving irrigation is 679.6 kg/ha, of which the amount of nitrogen fertilizer is 327.4 kg/ha, and the equivalent energy input is 21.7 GJ/ha, accounting for 83.9% of the total energy input of mineral fertilizers. This result is consistent with the studies of [66,67] on Chinese kiwi fruit orchards (the amount of nitrogen fertilizer is around 636–891 kg/ha). However, according to the studies of [34,64], the application rate of nitrogen fertilizer per unit area of kiwifruit planting in Iran is only 169.4–316.3 kg/ha.

China is striving to reach the United Nation’s Sustainable Development Goals by 2030 [68]. However, the excessive use and low utilization efficiency of mineral fertilizers, mainly nitrogen fertilizers, have long brought severe challenges to the sustainable development of China’s agriculture: China’s average annual nitrogen application per unit area of farmland was 305 kg/ha, while the world’s average nitrogen application rate was 74 kg/ha [69]. Among them, the excessive application of nitrogen fertilizer to fruit trees is the most serious, with an average excess rate of 429 kg/ha [68]. In China, the utilization efficiency of nitrogen fertilizer is 0.25, which is lower than the world average of 0.42, and this value can reach 0.65 in North American [70]. The excessive application of nitrogen fertilizer can significantly increase both the energy input and capital input in production, and at the same time, it will easily lead to soil acidification [71], water pollution [72] and massive greenhouse gas emission [73]. The main reason for the serious excessive use of mineral fertilizers is the one-sided pursuit of yield maximization and the absence of effective fertilization guidance. In fact, there is a law of diminishing returns for plant nutrient demand. After the amount of fertilization reaches a certain level, with more fertilizer input, the yield begins to decline. Therefore, in view of the problems of excessive fertilization and low fertilizer utilization efficiency during the production process of kiwifruit orchards in China, scientific and reasonable fertilization techniques should be promoted to the growers. For instance, the precise application of various elements of fertilizer could be formulated by testing the soil nutrients in advance [74], which could be an important way to drastically reduce the energy input in the production of kiwifruit.

4.2. The Operation Quality of Water-Saving Irrigation Systems Needs to Be Improved Urgently

The application of water-saving irrigation systems can save water, labor and fertilizer [8,9,75], which have also been confirmed in this study. However, it can also be seen in Figure 3, Figure 4 and Figure 5 that there are significant differences in the management and operation levels of water-saving irrigation technology among different users, and there are very large differences in their input to various production factors in the kiwifruit production process. Still taking the amount of mineral fertilizer input as an example, our survey results show that the average input amount of mineral fertilizer under water-saving irrigation was 679.6 kg/ha, but the maximum and minimum values are 934 kg/ha and 305 kg/ha, respectively. The maximum value of mineral fertilizer is even higher than that under surface irrigation. Through the survey process, it was found that the main reason for this phenomenon is that growers that consumed lower mineral fertilizer tended to have the ability to use fertigation properly in production. On the contrary, although there are some orchards where water-saving irrigation systems have been installed, they were only used for irrigation, and the traditional manual throw of solid fertilizer was still used for fertilization. This means that the technical advantage of water-saving irrigation systems—that the fertilizer can be saved through fertilization—has not been brought into play, which leads to increased production costs of kiwifruit planting and a low energy utilization efficiency [68].

The reasons for the low management and operation level of water-saving irrigation systems are mainly as follows: on the one hand, the construction of water-saving irrigation systems for a small household in China is mostly invested in by the government, which aims to ensure the rapid promotion and application of water-saving irrigation technology in China so as to save water in agricultural production [76]. However, taking government investment as the leading factor in promoting the application of water-saving irrigation technology also has obvious drawbacks. For example, it often attaches importance to the construction of water-saving irrigation systems but pays insufficient attention to the management and operation of the system; for instance, there is always a lack of necessary training for users of water-saving irrigation systems, and it is difficult to obtain effective services when there is a problem with the operation of the system and it needs to be repaired. Affected by the acceleration of China’s urbanization and industrialization, the aging phenomenon of China’s rural labor force has become serious [77]. The ability of the population to accept new technologies is weak [78]. When there are problems in the operation of the water-saving irrigation system and it is difficult to solve them in time, the phenomenon of abandonment and use will occur.

In addition, the distribution of cultivated land in China is scattered, the per capita cultivated land area is small and the orchard area of a small household does not exceed 0.67 ha. The design of a water-saving irrigation system does not always take this fully into account, and it is generally more suitable for the management of large-scale orchards. In actual operation, multiple growers often use one pump and fertilization system together. Due to the lack of a unified irrigation and fertilization plan, the coordination of the irrigation and fertilization time for multiple users is inconvenient. During the investigation, it was found that some growers added small fertilization devices in the field according to the small and scattered characteristics of the field in China, as shown in Figure 6b, so as to realize the independent and flexible control of the fertilization time and ensure the integration of water and fertilizer. This method has been proved to be feasible and has achieved the obvious effect of saving fertilizer, saving energy and increasing the yield. The original smallholder growers can continue to engage in kiwifruit production through franchise means and abide by the unified production management norms. For those small-sized orchard operators who have a difficult time achieving large-scale operation, they should also develop appropriate water and fertilizer integrated irrigation technology according to local conditions so as to improve the management and operation level of water-saving irrigation systems.

5. Conclusions

In recent years, there have been unsustainable phenomena such as the continuous increase in production costs and the redundant input of production materials in the production process of kiwifruit in Shaanxi, China. Based on the methods of a questionnaire survey and efficiency evaluation, this study clarified the structure and composition of the energy input in the production process of kiwifruit from the perspective of energy. The key reasons that currently limit the sustainable production of kiwifruit in Shaanxi were expounded upon, and the optimal input scheme of production factors under different irrigation technology modes was screened out, which provided the basis for the efficient use of energy in kiwifruit production and sustainable production.

When surface irrigation was used, the energy input of kiwifruit production was 85.4 GJ/ha, and the energy output was 59.7 GJ/ha. The highest energy input was mineral fertilizer (41.3 GJ/ha, accounting for 48.31% of the total input energy). When water-saving irrigation was used, the energy input of kiwifruit production was 72.3 GJ/ha, and the energy output was 62.3 GJ/ha. The highest energy input was also mineral fertilizer (25.8 GJ/ha, accounting for 35.7% of the total input energy). The application of water-saving irrigation technology has effectively improved the energy utilization efficiency of kiwifruit production. Compared with surface irrigation, the energy ratio has increased from 0.70 to 0.86, the energy productivity has increased from 0.37 kg/MJ to 0.45 kg/MJ and the net energy has increased from −25.8 GJ/ha to −9.93 GJ/ha.

Under the traditional surface irrigation mode, the redundancy of irrigation power consumption is high, and the water saving potential is obvious; under the water-saving irrigation mode, the redundancy of the mineral fertilizer application amount is high, and the fertilizer saving potential is high. The large input of inorganic fertilizer and the low management level of water-saving irrigation systems are the main limiting factors for the sustainable production of Chinese kiwifruit at present; it is urgent to popularize the fertilization technology of scientific systems, formulate the appropriate fertilizer amount and proportion and promote the scale, unification and standardization of kiwifruit production.