The Impact of Training Systems on Productivity and GHG Emissions from Grapevines in the Sughd Region in Northern Tajikistan

Abstract

Northern Tajikistan creates favorable conditions for growing grapes due to its climate. The choice of method of grape production to ensure a high-quality yield, while reducing the negative effects of such production on the environment, poses a serious challenge to implementation regarding the principles of sustainable production. In addition to the essential techniques associated with grapevine production, such as irrigation, fertilization, and plant protection, a training system plays a significant role. The objective of this research was to evaluate the environmental efficiency of vineyard training systems in northern Tajikistan. The indicators accepted for the evaluation process were the vegetative growth of plants, yield, and environmental pressure of production expressed by greenhouse gas (GHG) emissions. The single-factor experiment was conducted in Ghafurov District, and the following training systems were used: (A) multi-arm fan; (B) Umbrella Kniffin; (C) one-side multi-arm, paired planting. Growth parameters, yield, yield quality, and GHG emissions were evaluated. The cultivation of grapes in training system “C” resulted in higher values of parameters such as Practical Bud Fertility coefficient and fruiting shoots pcs and higher yields. The training systems were ranked according to GHG emissions per yield unit in the following order, from lower to higher emissions: C < B < A.

1. Introduction

Grapevine cultivation in Sughd Region in the Republic of Tajikistan is an important part of crop production. It is an integral part of the regional landscape, both historically and culturally. In 2019, the area covered by vineyards in that region was 9722 ha [1,2]. Hot and arid climate (total annual precipitation in lowlands ranges from 200 to 300 mm) in conjunction with access to watercourses and water reservoirs, which constitute an irrigation source in this region (basins of Zeravshan, Syr Darya, Khojabakirgan, and Isfara rivers), create good conditions for grape cultivation [3,4]. Apart from the essential techniques associated with grapevine production, such as irrigation, fertilization and plant protection, a training system (with a focus on row spacing and plant height) also plays a significant role. A training system determines crop-production potential by the spatial distribution of leaves and shoots within the canopy. This is because such distribution substantially affects the level of interception of sunlight, which leads to the intensification of photosynthesis in leaves. Photosynthetic productivity of plants is the factor that limits the grape yield in modern plantations. Photosynthetic productivity is directly associated with interception of sunlight by ground cover [5,6]. Plant spacing affects not only the amount of yield but also its qualitative parameters. Varied plant spacing affects plant vigor indirectly through a change in shading and has an impact on shoot growth. According to Archer and Strauss [7] and Greer and Weedon [8], when the spacing is wider, vegetative development extends over time, which delays maturation and has an adverse effect on the quality of grapes, with particular reference to the content of sugars. According to Morris et al. [9] and Shaulis et al. [10], over-cropped vines are another cause of delays in fruit maturation and of decreased content of sugars in fruits. The possibility of creating a specific amount of produce of desired quality from a space unit is the most important factor shaping grapevine production economics. Rashidov et al. [3] and Niemiec et al. [4] concluded that fertilization strategy is the key element affecting production effectiveness. Modifications in terms of fertilization strategy, amount and technology of irrigation, and plant-cultivation methods are being introduced in production practice to increase productivity and obtain appropriate quality crops. Depending on climate conditions, such as changing temperature during the vegetation period, the level of sunlight in critical stages for production, the amount of available water, as well as agriculture and the level of mechanization of production, those modifications take different courses and levels. An example of such long-term modification is introducing new systems of plantation management. According to Rashidov [1,2], most popular systems for the area of Northern Tajikistan are two-sided training systems with 3 × 2 m of spacing and low (80 cm) trunk. In 2019 in the Sughd region, 5350 out of 9722 ha of grapes were cultivated using this method, and another popular system, with a total area of 2954 ha, is the no-trunk system. The popularity of both trunk and no-trunk systems with 3 × 2 m of spacing can be explained by their introduction and intense popularization in the 1960s and 1970s by the Michurin All-Union Research Institute of Horticulture. They were characterized by small expenditure and sufficient crop. To improve production efficiency, in recent years, systems with high trunks have been introduced. In 2019, the surface of cultivation based on 3 × 2 m on a 120 cm trunk amounted to 1200 hectares, and a 140-cm trunk system amounted to 218 hectares. One of the newest implemented systems is a system trained on a high trunk (140 cm), type “one-side multi-arms, paired planting” with spacing of 4 × 3.7 + 0.6 m. It was introduced by the Institute of Horticulture of Tajik Academy of Agricultural Sciences, however, it is not very popular in the region where the investigation took place. The main factor hindering development is the poor level of knowledge among farmers and their mistrustful attitude towards change. In 2019, the area cultivated with this system was 97 hectares of grapevines in the region of northern Tajikistan. The correct management of the production area in the plantation by the selection of a training system that matches the environmental conditions allows for optimal use of soil and rational management of water resources [11]. Production planning and management are key factors in implementing the principles of sustainable production. Production efficiency is defined by product quality and by the reduction in the human impact on the environment [12,13]. Environmental aspects are an integral part of all modern quality-management systems in primary production [14]. The proper management of soil, particularly in the context of maintaining soil fertility and preventing erosion, is important in terms of rationalization consumption of environmental resources in agricultural production [15]. Determination of the effect of a particular type of production on the environment by using standardized assessment methods makes it possible to evaluate the implementation of principles [16]. One such method is the assessment of the effect of a given production system on greenhouse gases (GHG) emissions. In the case of agri-food products, the grape supply chain is one of the most frequently analyzed, owing to its importance in both economic production and global distribution. The global importance of grape production stems from the fact that they are raw materials for wine production [17,18]. In the case of grape production in Tajikistan, they are intended mainly for direct consumption or the production of raisins, juices, and concentrates. Considering the effect of grape production on GHG emissions, according to many authors, fundamental factors generating the highest emissions include applied pesticides, fertilizers, and energy expenditure for irrigation and cultivation measures [19,20]. In December 2016, the Republic of Tajikistan adopted a policy direction included in the UN 2030 Agenda for Sustainable Development, the objectives of which include increasing the quality of food products and protection of natural resources. In accordance with the idea of sustainable development, agricultural systems should be evaluated in the context of multifaceted impact on the natural environment. One of the most frequently defined anthropopressure factors on the part of agriculture is GHG emissions, and it is this very parameter that should be considered while evaluating agricultural production systems and recommendations within the frames of certified management systems, such as GLOBAL G.A.P. Despite the extensive knowledge of production size achieved with specific technical and technological solutions, there is a lack of knowledge of the effect of training systems on the environmental aspect of production. The objective of this research was to evaluate the environmental efficiency of vineyard training systems in northern Tajikistan. The indicators accepted for the evaluation process were the vegetative growth of plants, yielding, and environmental pressure of production expressed by GHG emissions.

2. Materials and Methods

2.1. Field Trial Design and Treatments

The experiment was conducted between 2016 and 2018 in northern Tajikistan, in Sughd Region, Ghafurov district on the “Abukarimov” plantation (40°19′18″ N 69°46′35″ E). The climate of the region is dry and hot, characterized by an annual rainfall of about 200 to 300 mm, with evaporation ranging from 1.188 to 1.573 mm and 2.000 to 5.000 of total annual heating degree days (°C) [1,2,3]. The average temperature and precipitation for 2016–2018 are presented in

Table 1. Temperature and precipitations in Ghafurov region in 2016–2018.

	2016		2017		2018
	P1 [mm]	T2 [°C]	P [mm]	T [°C]	P [mm]	T [°C]
january	13.3	4.3	29.7	1.8	2.3	1.0
february	2.2	6.5	29.6	2.5	20.2	3.8
march	28.8	13.3	31.8	8.1	57.9	12.4
april	6.2	16.8	40.9	14.9	6.6	15.9
may	64.4	22.5	18.4	23.9	17.6	21.3
june	4.8	27.6	0.8	28.2	19.2	26.5
july	8.7	29.0	-	29.5	-	30.5
august	-	28.0	0.3	27.0	-	27.0
september	-	24.3	7.2	22.5	0.7	21.6
october	23.4	12.9	13.8	14.7	21.4	10.5
november	29.2	4.7	4.5	9.9	22.6	7.7
december	28.4	3.6	10.3	1.6	14.9	0.3

¹P—precipitation, ²T—temperature.

. The single-factor experiment was set up using a randomized block design. Treatments were repeated three times, and the area of each plot was 0.10 ha. A vineyard training system (scheme of planting, height of trunks) was a factor of the experiment. Within the scope of realization of the objective of the research, an environmental analysis of three vineyard training systems was conducted (Figure 1): (A) spacing 3 × 2 m, trained on a low trunk (80 cm), “multi-arm fan”; (B) spacing 3 × 2 m, trained on a high trunk (120 cm), “Umbrella Kniffin”, (C) spacing 4 × 3.7 + 0.6 m, trained on a high trunk (140 cm), “One-side, multi-arm, paired planting” [1,2]. This research has allowed a thorough evaluation of the environmental and production aspects of the process and creating recommendations for newly established plantations. The economic growth of the Republic of Tajikistan in terms of agricultural produce exports observed in recent years indicates that, in the near future, fruit production will increase, both for the fresh-produce market and processing. Optimizing production at the stage of starting a plantation therefore has strategic value in the context of future economic effects and ensuring compatibility with certified systems of quality control, which are mandatory in the target countries. All the examined objects were equally fertilized and irrigated. Soil and climate conditions in all variants were the same. The experiment was conducted on Gypsic Calcisols [21]. Physico-chemical properties of the soil are presented in

Table 2. Physical and chemical properties of investigated soil.

Soil Layer (cm)	pH	C org %	N	P	K	Bulk Density, g cm−3	Particle Density, g cm−3
			mg kg−1
0–25	7.1	0.51	370	910	1693	1.60	2.54
25–50	7.0	0.27	240	690	1751	1.90	2.63
50–75	6.9	0.18	180	600	1834	1.95	2.63
75–100	6.7	0.09	180	540	1759	2.04	2.37

. The vineyard was set up in 2009. The husain white variety was cultivated at the plantation. In each year of the experiment, standardized fertilization was applied in the amount of 200 kg N ha⁻¹, 100 kg P ha⁻¹, and 100 kg K ha⁻¹. Nitrogen was applied in the form of ammonium nitrate (34% N), phosphorus in the form of triple super phosphate (21% P), and potassium in the form of potassium chloride (50% K). Nitrogen was applied in two doses (50/50): the first dose was applied in early spring (first decade of March), and the second dose was applied during the blooming period (third decade of May). Phosphorus was applied in one dose in late autumn (the second decade of November). Potassium was applied in two doses (50/50): the first dose was in late autumn and the second dose was applied with nitrogen in the blooming period (third decade of May). The plots were irrigated every year with amounts ranging from 500 to 550 mm ha⁻¹, with irrigation on 6–7 planned dates. Formation pruning was performed in November.

2.2. Measurements

The soil bulk density was determined by use of the core method and the particle density by the use of the pycnometer method [21]. The chemical properties of air-dry soil samples were determined with the following method: pH measured by the potentiometric method, soil organic carbon content measured using Tiurin’s method [14], a Kjedahl method was applied to determine the content of nitrogen in soil [14], the contents of phosphorus and potassium were determined by an ICP-EAS method (JY 238 ULTRACE, Jobin Yvon Emission, Longjumeau, France). Every year during the growing season, selected yield-forming and vegetation indices were measured. The results presented in this study come from measurements done on 10 plants from each plot: the number of buds, number of shoots, number of fruiting shoots as well as the number of non-fruiting shoots, and inflorescences, and the amount and mass of bunches. Each year, after the grape harvest, yield was calculated per 1 ha. Grapes were taken from each plot for analysis. One collective sample consisted of 10 primary samples from different parts of the plot. Total sugar content and total acidity were determined in fresh samples of fruit. Total sugar assay was performed using the anthrone method [3]. Total acidity was measured by titrating sodium hydroxide into a sample of grape juice to neutralize the acid in the juice. The coefficient by practical bud fertility (K) was calculated as the average number of bunches on one shoot which grew during the growing season from the winter bud [22,23].

K = B/S,

where B is the number of bunches on one shoot, and S is the number of shoots which grew during the growing season from the winter bud.

Yield index by practical bud fertility (YIPF) was calculated multiplying bunch weight and practical bud fertility data [23]

YIPF = B_w × K,

where B_w is the average weight of bunches.

To calculate the amount of GHG emissions in grape production, the ISO 14040:2006 “Environmental management—Life cycle assessment—Principles and framework” [24] standard was used. The timeframe of the system was one year. The potential to generate a greenhouse effect was estimated basing on the emission of greenhouse gases in carbon dioxide (CO₂) equivalent. One Mg of commercial product was adopted as a functional unit. The amount of GHG emissions (E kg Mg⁻¹) in carbon dioxide equivalent was calculated according to the IPCC methodology [25,26] and according to the formula

E = e_ec/yield,

e_ec = e_chem + e_field + e_mm+ e_irr,

where e_chem is the emissions associated with the manufacture of fertilizers and agrochemicals used for crop cultivation [27,28]; e_field is the emissions from the field, direct and indirect emission, associated with fertilizer use, mineralization of organic matter in the soil, and management of post-harvest residues and mineralization of organic matter in the soil [25,29]. The value of nitric oxide emissions from harvesting residue and mineral fertilizers was adopted at 1.00% [24,26]. The value “N–N₂O emissions” was multiplied by 44/28 to convert it into N₂O. N₂O emissions were used as a CO₂ equivalent by multiplying them by the global warming potential of 298. The adopted content of carbon fraction in dry matter of harvesting residue was 50%, the adopted mineralization rate for carbon from residues was 25%, and the adopted soil mineralization rate of organic matter was 2% [25,26], e_mm is emissions associated with field work at the farm, e_irr is emissions associated with irrigation (for irrigation, electric water pumps were used, emissions were calculated by using the GHG emissions coefficient) [30].

2.3. Statistical Analysis

ANOVA was applied to analyze the results. The significance of mean differences among the treatments was tested with the multiple comparison procedure, and Tukey’s range test was applied at a significance level of α = 0.05. The analysis was performed using the statistical software package Statistica v. 13.0 (StatSoft Inc. Tulsa, OK, USA).

3. Results and Discussion

3.1. Development and Growth

Proper selection of the method of vineyard training has a direct effect on the quantity and quality of grape yields in a given region [31,32]. The selection of a training method for the vineyard as well as its location diversified the studied yield-forming parameters. Based on our study, it was established that, at 100 buds per plant, on average, 80.0% of buds were developed at the plantation. Considering the vineyard training scheme, in the period 2016–2018, a considerably lower number of buds developed in treatments “A” and “B”. In the case of fruiting shoots, the lowest number was observed for variants “A” and “B”, and the highest for variant “C” (

Table 3. Vegetation and growth indicators with dependence on the training systems (2016–2018).

Training System	2016	2017	2018	Mean
Buds pcs. plant−1
A	80.4 ± 1.72a1	80.1 ± 1.56a	78.3 ± 2.97a	79.62 ± 1.10a
B	80.5 ± 1.81a	80.3 ± 1.28a	78.9 ± 1.43a	79.94 ± 0.86a
C	82.4 ± 1.76b	81.7 ± 1.34b	81.6 ± 1.52b	81.91 ± 0.51b
Shoots pcs. plant−1
A	63.4 ± 0.16a	63.2 ± 0.56a	63.0 ± 0.69a	63.2 ± 0.18a
B	64.3 ± 0.95a	64.0 ± 1.12a	63.8 ± 1.52a	64.0 ± 0.46a
C	66.4 ± 0.63b	65.7 ± 0.55b	65.3 ± 0.74b	65.8 ± 0.61b
Fruiting shoots pcs. plant−1
A	33.4 ± 1.68a	33.0 ± 1.23a	32.4 ± 1.48a	32.9 ± 0.63a
B	33.9 ± 1.33a	33.4 ± 2.10a	32.9 ± 1.34a	33.4 ± 0.73a
C	40.4 ± 1.52b	40.3 ± 1.32b	40.3 ± 1.79b	40.3 ± 0.53b
Practical bud fertility
A	0.61 ± 0.09ab	0.59 ± 0.07a	0.55 ± 0.06a	0.58 ± 0.07a
B	0.54 ± 0.07a	0.55 ± 0.08a	0.52 ± 0.09a	0.54 ± 0.05a
C	0.67 ± 0.10b	0.68 ± 0.05b	0.64 ± 0.05b	0.68 ± 0.08b

¹Means with different letters within column are significantly different, according to Tukey test (p ≤ 0.05), ± standard deviation.

). Fruiting shoots constituted, on average for 2016-2018, 54.6% of all shoots. Variant “C” had a lower density (1350 plants per hectare) than variants “A” and “B” (density 1650 plants per hectare) and, by extension, a larger nutrition area. This difference affected both the appearance of the plants and the internal state of developmental stages, shoot growth energy, and the character of fructification. When the plant nutrition area is larger, relatively strong skeleton branches develop, and shoots have more moderate growth but higher fruitfulness. Access to daylight was an important factor influencing the intensity of flowering [33]. Plant spacing in variant “C” was higher and because of that there was less mutual shading among plants, which had a positive effect on their development. It is widely assumed that the direct capture of sunlight by buds improves grapevine fertility [33,34,35]. Coefficient of practical bud fertility is an index that enables assessment of the yield-forming potential of a given method of cultivation and of running a plantation [22,36]. The highest value of this index was observed for system “C”, and the lowest for system “B” (Table 3). In the case of this study, the coefficient of practical bud fertility did not translate accurately into obtained yield.

3.2. Yield and Productivity

Table 4. The productivity of the grape depending on the training systems (2016–2018).

Training System	2016	2017	2018	Mean
Bunches pcs. plant−1
A	38.4 ± 2.99b1	37.0 ± 1.32b	34.7 ± 1.71b	36.7 ± 1.72b
B	34.7 ± 3.63a	35.3 ± 1.68a	32.9 ± 1.54a	32.9 ± 1.44a
C	45,8 ± 1.42c	45.0 ± 2.14c	41.6 ± 1.41c	43.8 ± 1.23c
Average mass (g) of bunch
A	304 ± 6.7a	300 ± 7.7a	295 ± 8.3a	300 ± 1.9a
B	354 ± 1.5b	353 ± 1.0b	351 ± 3.0b	353 ± 1.7b
C	383 ± 2.2c	381 ± 1.3c	380 ± 1.7c	381 ± 1.4c
YIPF
A	184 ± 10.6a	176 ± 14.2a	162 ± 8.7a	173 ± 12.0a
B	191 ± 8.4a	195 ± 15.1b	181 ± 5.4b	189 ± 10.5b
C	258 ± 11.3b	261 ± 7.2c	242 ± 13.7c	258 ± 10.9c
Yield kg plant−1
A	11.5 ± 0.77a	11.1 ± 0.42a	10.4 ± 0.40a	11.0 ± 0.52a
B	12.3 ± 0.41b	12.4 ± 0.34b	11.6 ± 0.36b	12.1 ± 0.33b
C	17.2 ± 0.69c	17.1 ± 0.47c	15.8 ± 0.49c	16.7 ± 0.61c
Yield Mg ha−1
A	18.4 ± 0.34a	18.6 ± 0.21a	18.2 ± 0.21a	18.4 ± 0.21a
B	19.9 ± 0.22b	20.3 ± 0.24b	20.1 ± 0.13b	20.1 ± 0.23b
C	21.0 ± 0.27c	20.9 ± 0.22c	20.8 ± 0.24c	20.9 ± 0.09c

¹Means with different letters within column are significantly different, according to Tukey test (p ≤ 0.05), ± standard deviation.

shows the yield of the investigated variety for 2016-2018. The lowest number of bunches was recorded for variant “B”, and the highest for variant “C” for all investigated years. The number of bunches did not correlate with the amount of yield. The bunch weight (which can be presented in the descending order C > B > A) was the factor that ultimately influenced yield results. Based on research results, the productivity of grape yield in a given system was determined and expressed through YIPF (Table 4). Significantly higher values were recorded for system “C”, and the lowest for system “A” in all years of experiment, 2016-2018. In the case of diversification associated with the training system, the highest yield was obtained in variant “C”, and the lowest in variant “A”. Apart from a lower planting density, variant “C” was characterized by high training, variant “B” was also included in high-training systems, whereas “A” was classified as low. Application of variant “C” allowed the obtainment of, on average, a 2.5-ton higher yield than in the case of variant “A”. The difference between variants “B” and “A” was 1.7 ton per hectare. Various authors’ studies [37,38,39,40] indicate that in low-training systems with downward positioning of shoots, there is a reduction in plant vigor and in yielding in relation to ‘high’ systems. According to Van Leeuwen et al. [39], vineyards with a wider planting adapt better to dry conditions. The presented statement is associated with the assumption that at a lower density each plant has easier access to water resources in the soil. This statement is valid when grapevine roots spread all over the soil. According to Champagnol [40], grapevines can spread over 10 m² of soil surface. In this study, the developmental area of a single plant was 6 m² in variants “A” and “B”, and 8 m² in variant “C”.

Table 5. Sugar content, total acidity and dry mass depending on the training systems (2016–2018).

Training System	2016	2017	2018	Mean
Sugar content %
A	18.4 ± 0.45b1	18.5 ± 0.67b	18.6 ± 0.81b	18.5 ± 0.51b
B	18.0 ± 0.29b	18.1 ± 0.41ab	18.2 ± 0.37b	18.1 ± 0.44ba
C	17.5 ± 0.38a	17.7 ± 0.22a	17.7 ± 0.44a	17.6 ± 0.19a
Total acidity %
A	0.50 ± 0.158a	0.60 ± 0.135a	0.55 ± 0.145a	0.55 ± 0.124a
B	0.52 ± 0.143a	0.59 ± 0.141a	0.63 ± 0.108a	0.58 ± 0.091a
C	0.64 ± 0.113a	0.66 ± 0.127a	0.68 ± 0.119a	0.66 ± 0.083a
Dry mass %
A	19.3 ± 1.51a	21.9 ± 0.88a	21.0 ± 0.80a	21.6 ± 1.2a
B	19.9 ± 0.82a	20.4 ± 0.57a	21.8 ± 0.68a	20.6 ± 0.6a
C	20.3 ± 1.13a	21.6 ± 0.60a	22.9 ± 0.71a	20.7 ± 0.3a

¹Means with different letters within column are significantly different, according to Tukey test (p ≤ 0.05), ± standard deviation.

shows selected qualitative parameters of the yield for years 2016-2018. It is widely believed that the intensity of solar radiation, the amount of solar energy reaching a plant, is one of the factors influencing the content of sugars [41,42,43,44,45]. This study showed a significant difference between the investigated treatments in all years. The highest sugar content was recorded in fruits from the vineyard which used system “A” and “B”, and the lowest was from the vineyard which used system “C”. In the case of fruit acidity, there were no significant differences between treatments. According to numerous authors [7,33,34,39], higher plant density on a plantation increases the leaf-to-fruit ratio, which translates directly into an increase in sugar content in fruits [34]. Parker et al. [44] determined that a reduction in leaf-to-fruit ratio may delay veraison, which translates into a decrease in grape sugar without a significant effect on grape acidity. In their study, Sabbatini et al. [45] established that the greater the number of buds, the greater the productivity of a given plant and the lower the content of soluble substances. Our study also showed a strong positive correlation between the number of buds and the weight of bunches (Figure 2). Figure 3 shows a significant negative correlation between the number of buds on a plant and sugar content in fruits. An increased number of developing buds and shoots is indicative of substantial plant vigor and considerable plant vegetative development, and it increases yields. Menezes Feitosa et. al. [46] determined that, in terms of product quality, excessive plant vigor has an adverse effect on many physiological processes in plants. Vigorously developing shoots produce more sugars. However, at the same time, a plant requires higher respiratory activity; the produced sugars are used for vegetative development, which inhibits accumulation of these soluble substances in berries. Estimation of the obtained biological yield (in the case of grapes, this will be the amount of sugar obtained from a hectare) is an important factor which enables characterization of the efficiency of a given production method. Based on this study, it was established that the lowest biological yield was obtained with system “A”, as seen in Figure 4. The most efficient in terms of biological yield were systems “B” and “C”, where the obtained yield was much higher than in system “A”.

3.3. GHG Emissions

Estimation of carbon dioxide emissions from grapevine cultivation enabled evaluation of the environmental efficiency of a given training system. The effect of the training system on GHG emissions per 1 ton of yield is presented in Figure 5, and input data are shown in

Table 6. Energy consumption and greenhouse gas (GHG) emissions associated with agricultural treatments, fertilizers and pesticides.

Agrochemicals
Type of agrochemicals	kg CO2-eq kg−1	kg CO2-eq ha−1
Ammonium nitrate	9.280	1856.0
Triple superphosphate	0.440	100.3
Potassium salt	0.680	81.9
Pesticides	10.97	93.2
Agricultural treatments
Type of agricultural treatment	Diesel use [dm3 ha−1]	kg CO2-eq ha−1
Cultivation with an aggregate 2×	40.0	127.7
Mineral fertilization 3×	17.8	55.5
Application of plant protection products. 4 ×	24.0	76.6
Field transport	7.0	22.3
Other
Irrigation [KWh]	6562.5	108.7
N2O–N emissions produced from managed soils, biomass, fertilizers		1002.7–“A” 1009.5–“B” 1012.7–“C”
Emission related with organic matter mineralization (soil, crop residue)		1002.7–“A” 1009.5–“B” 1012.7–“C”

. The lowest emissions per unit of yield were recorded for system “C” (3.8% lower in relation to system “B” and 12.9% lower than in system “A”). Yield obtained in a given system was the fundamental factor determining such division. The share of individual constituents of emissions is shown in Figure 6. Applied fertilizers were of the greatest importance in emissions in all the studied cases, as the energy inputs for their production constituted, on average, 51.6% of the total emissions, with other constituents being field direct and indirect N₂O (25.5%), organic matter mineralization (10.7%), fuel (7.1%), irrigation (2.8%), and pesticides (2.4%). In other studies [47,48], electric energy used for irrigation was one of the factors generating the highest GHG emissions. GHG emissions related to electric energy are strongly associated with the region where specific energy is generated. For the European Union, the average emissions are 128.2 gCO_2-eq MJ⁻¹ (medium voltage) [30]. For Tajikistan, they are 4.6 g CO_2-eq MJ⁻¹. Very low emissions are associated with the energy production system: as much as 87.4% of electric energy comes from hydroelectric power plants [49]. Using high doses of artificial manure is problematic in plantations in Sughd Region, including the analyzed farm. This particularly applies to phosphorus, as the uptake of this nutrient along with 18 Mg of yield from 1 ha is, on average, 13 kg or less [3,4,50]. In connection with the training system where grapevine leaves and branches are shredded and mixed with soil, phosphorus is not removed from the field in excessive amounts. Annual fertilization with phosphorus at a level of 100 kg ha⁻¹ is not rational [51]. Rashidov et al. [3], when comparing fertilizer doses of 44, 88, and 122 kg of P per ha in grape cultivation on Gypsic Calcisols in northern Tajikistan, found that a dose of 44 kg has the highest agricultural efficiency. In the context of sustainable production, the application of an appropriate training method enables obtainment of the optimum yield with a simultaneous reduction in production pressure on the environment, and proper rationalization of fertilization and reduced doses will provide greater production efficiency and environmental efficiency.

4. Conclusions

The cultivation of grapes in training system “C” with spacing 4 × 3.7 × 0.6 m at a height of 1.4 m: “one-side multi-arm, paired planting” resulted in the obtainment of the best (in terms of plant vigor) parameters, such as practical bud fertility coefficient and fruiting shoots pcs. plant⁻¹. Training systems have a substantial effect on the amount of fruit biomass obtained. The highest yield was obtained in treatment “C”. A significantly lower yield was obtained in system “A” (multi-arm fan with spacing 3 m × 2 m), 9.2% lower than the yield from system “B” and 13.6% lower than that obtained under system “C”. In terms of yield quality, the average weight of a bunch under conditions of system “C” was 381 g, whereas under system “B” it was 353 g and under “A” it was 299 g. The sugar content in fruits was significantly higher in systems “A” and “B”, but because of higher yielding under system “C”, total sugar yield (biological yield) from 1 ha was significantly higher in system “C”. The training system affected the amount of GHG emissions from the plantation. The training systems were ranked according to GHG emissions per yield unit in the following order, from lower to higher emissions: C < B < A. The change from a system with a low trunk, “A”, to a system with a high trunk, “B”, without changing the planting density, resulted in a significant increase in yield, whereas maintaining the sugar content in fruits at the same level contributed to the reduction in GHG emissions per unit of yield. The most beneficial (in terms of the environment and production) under the studied climatic conditions is the variant of training on a highest trunk (140 cm): “one-side multi-arm, paired planting” with lower density of planting.