Environmental Impact Assessment of Agricultural Production Using LCA: A Review

: Life cycle impact assessment (LCA) provides a better understanding of the energy, water, and material input and evaluates any production system’s output impacts. LCA has been carried out on various crops and products across the world. Some countries, however, have none or only a few studies. Here, we present the results of a literature review, following the PRISMA protocol, of what has been done in LCA to help stakeholders in these regions to understand the environmental impact at different stages of a product. The published literature was examined using the Google Scholar database to synthesize LCA research on agricultural activities, and 74 studies were analyzed. The evaluated papers are extensively studied in order to comprehend the various impact categories involved in LCA. The study reveals that tomatoes and wheat were the major crops considered in LCA. The major environmental impacts, namely, human toxicity potential and terrestrial ecotoxicity potential, were the major focus. Furthermore, the most used impact methods were CML, ISO, and IPCC. It was also found that studies were most often conducted in the European sector since most models and databases are suited for European agri-food products. The literature review did not focus on a speciﬁc region or a crop. Consequently, many studies appeared while searching using the keywords. Notwithstanding such limitations, this review provides a valuable reference point for those practicing LCA.


Introduction
Food supply chains (FSCs) are very complex. There are many components involved in FSCs that process, produce, package, store, transfer, distribute, and market food products to final consumers [1]. Each element in the FSC process is essential, as in any other supply chain; a change in one component affects the others. The relationship between the food system and the economy, environment, and society is mentioned by some organizations and agencies, such as the Food and Agricultural Organization (FAO), Institute of Medicine (IOM), and National Research Council (NRC), when they define the FSC [2]. Therefore, the most crucial question is as follows: Which food production system is more sustainable for the environment and communities?
There are many concerns about food resources and massive population growth, such as meeting the food demand for the world's population, production, and food consumption [1]. The total crop production must double or increase by at least 70% to meet the increasing world population's demand by 2050 [3]. Models have estimated that a 2.4% annual increase in crop yield is necessary to reach the 2050 demand [4]. The rise in food demand results in substantial energy and resource use by the food supply chain, leading to different environmental impacts. Many organizations have mentioned environmental impacts associated with food production, including the use of land, water, and climate and/or ecotoxicity, and (iii) researching agriculture products. A total of 36 research ar were eliminated because they were about FSCs and not GHG/ecotoxicity as an effec egory, did not apply the LCA methodology, or utilized the LCA method for nonag tural products. The LCA studies were analyzed extensively considering four phases o LCA: • Goal and scope definition, • Life cycle inventory, • Life cycle impact assessment, • Life cycle interpretation/recommendation options.

Search Strategy
The literature review was done through the Google Scholar database. The key "LCA crop production" was used in the initial step, which yielded 59,100 studies July 2021. Later, more specific keywords were used, such as "agri-food supply chain LCA" and "agri-food supply chain and GHG" combined with different fruit and ve ble products such as corn, peanuts, wheat, tomato, and apple. Nevertheless, the nu of studies available remained enormous, the largest number of articles we got whe used the above key word with different crops was 7330, while the smallest number 1820. A total of 110 articles were downloaded and analyzed. Twenty-nine studies excluded because they were about FSCs and not about GHG/ecotoxicity as an effect gory, or because they utilized the LCA method.
Furthermore, seven more were excluded because they used the LCA method for agricultural products. Accordingly, we ended up with 74 articles after applying the s tion criteria. Figure 1 shows the steps used throughout the review and the inclusion ria for the literature. The release of the latest version of the ISO 14040 standard in 2006 explains why LCA research is attracting more attention. Moreover, some have recently gone so far as to state that the ISO 14040: 2006 series "has proven a suitable tool for sustainability assessment" [13,14]. Fava et al. (2009) claimed that ISO 14040 should be the basis for future LCA studies [23].
Studies found that the most common tool to study the impact on the environment associated with a product over its life cycle in the agri-food sector was the LCA ISO 14040 standard [14,15]. LCA ISO 14040 has four main phases: (1) goal and scope, which is the essential component of the LCA, (2) qualitative and/or quantitative inventory analysis of the used resources and the emissions released from the life cycle of a product, (3) life cycle impact assessment, which can be divided into classification, characterization, and evaluation, and (4) the interpretation, involving the identification of key issues, evaluation (including checking completeness, sensitivity, and consistency), and development of conclusions together with recommendations, as defined by ISO 14043 (Figure 3). The details of each phase are discussed below.

Goal
According to Lee and Inaba (2004), the following questions should be addressed to set up the goal: Why perform LCA, who is the target audience, and what is the product under the LCA study [10]? These were recognized from the reviewed articles while examining the first phase of the LCA, as given in Figure 4. Some of the studies stated the answers to these questions directly, whereas others addressed them indirectly.  show the most common responses to each question. Studies found that the most common tool to study the impact on the environment associated with a product over its life cycle in the agri-food sector was the LCA ISO 14040 standard [14,15]. LCA ISO 14040 has four main phases: (1) goal and scope, which is the essential component of the LCA, (2) qualitative and/or quantitative inventory analysis of the used resources and the emissions released from the life cycle of a product, (3) life cycle impact assessment, which can be divided into classification, characterization, and evaluation, and (4) the interpretation, involving the identification of key issues, evaluation (including checking completeness, sensitivity, and consistency), and development of conclusions together with recommendations, as defined by ISO 14043 (Figure 3). The details of each phase are discussed below. The release of the latest version of the ISO 14040 standard in 2006 explains why LCA research is attracting more attention. Moreover, some have recently gone so far as to state that the ISO 14040: 2006 series "has proven a suitable tool for sustainability assessment" [13,14]. Fava et al. (2009) claimed that ISO 14040 should be the basis for future LCA studies [23].
Studies found that the most common tool to study the impact on the environment associated with a product over its life cycle in the agri-food sector was the LCA ISO 14040 standard [14,15]. LCA ISO 14040 has four main phases: (1) goal and scope, which is the essential component of the LCA, (2) qualitative and/or quantitative inventory analysis of the used resources and the emissions released from the life cycle of a product, (3) life cycle impact assessment, which can be divided into classification, characterization, and evaluation, and (4) the interpretation, involving the identification of key issues, evaluation (including checking completeness, sensitivity, and consistency), and development of conclusions together with recommendations, as defined by ISO 14043 (Figure 3). The details of each phase are discussed below.

Goal
According to Lee and Inaba (2004), the following questions should be addressed to set up the goal: Why perform LCA, who is the target audience, and what is the product under the LCA study [10]? These were recognized from the reviewed articles while examining the first phase of the LCA, as given in Figure 4. Some of the studies stated the answers to these questions directly, whereas others addressed them indirectly.  show the most common responses to each question.

Goal
According to Lee and Inaba (2004), the following questions should be addressed to set up the goal: Why perform LCA, who is the target audience, and what is the product under the LCA study [10]? These were recognized from the reviewed articles while examining the first phase of the LCA, as given in Figure 4. Some of the studies stated the answers to these questions directly, whereas others addressed them indirectly. Figures 5-7 show the most common responses to each question. common aim was to assess agricultural production, cultivation, processing, packaging, transport, and emission at all production stages to recognize the vast issues and to propose reasonable alternatives that decrease the environmental effects ( Figure 5). The purpose of this review was to better understand how to use LCA to evaluate the environmental impact of agricultural production. The least common goal was to compare LCA to other methods, which may be due to the difficulty of making a fair comparison in terms of method performance.

Target Audience
The target audience defines who undertakes or commissions an LCA and for whom. It is critical to understand who will use the LCA results to provide them with helpful information. The majority of articles have multiple target audiences (TAs). Politicians working on climate change, decision-makers, and policymakers on global warming potential (GWP) footprints related to food and common agricultural policy (CAP) were the most common TAs, with 10 studies. Additionally, several studies targeted government sectors such as food sector policymakers, the country's agriculture sector, and the fruit and vegetable sector. Following that, the producers, namely, the farmers and the producing industry, were targeted in eight studies, six of which provided information to the consumer on a local and international scale (see Figure 6). People working on social and economic development, such as government policymakers for sustainable consumption and production, future ecolabeling programs, and those working to improve the environmental and financial sustainability of existing agricultural systems, were also targeted. Another target audience was represented by the Florida food, agri-food, and citrus  industries. As shown in Figure 6, only 35 of the 74 research articles analyzed clearly stated their target audience. The frequency of target audiences is also displayed as a word cloud for a rapid overview.

Agricultural
We divided the products into 11 categories: tomato, fruits, citrus, vegetable, fresh salad, grains, seeds, oil, sugar, flower, and trees, as shown in Figure 7. The most common product was tomato; 13 studies analyzed tomato production, including fresh tomato, canned tomato (whole peeled, paste, and diced), and ketchup. The second most common product was wheat with nine studies. Because some studies involved more than one crop,

Aims of LCA
As indicated in the literature, LCA studies can be partitioned into two major categories: descriptive and comparative. Descriptions aim to recognize the natural load of a chosen framework, while comparisons aim to differentiate between two frameworks. Among the discussed papers, 48 were descriptive, while 30 were comparative. As noted, the most common aim was to assess agricultural production, cultivation, processing, packaging, transport, and emission at all production stages to recognize the vast issues and to propose reasonable alternatives that decrease the environmental effects ( Figure 5). The purpose of this review was to better understand how to use LCA to evaluate the environmental impact of agricultural production. The least common goal was to compare LCA to other methods, which may be due to the difficulty of making a fair comparison in terms of method performance.

Target Audience
The target audience defines who undertakes or commissions an LCA and for whom. It is critical to understand who will use the LCA results to provide them with helpful information. The majority of articles have multiple target audiences (TAs). Politicians working on climate change, decision-makers, and policymakers on global warming potential (GWP) footprints related to food and common agricultural policy (CAP) were the most common TAs, with 10 studies. Additionally, several studies targeted government sectors such as food sector policymakers, the country's agriculture sector, and the fruit and vegetable sector. Following that, the producers, namely, the farmers and the producing industry, were targeted in eight studies, six of which provided information to the consumer on a local and international scale (see Figure 6). People working on social and economic development, such as government policymakers for sustainable consumption and production, future ecolabeling programs, and those working to improve the environmental and financial sustainability of existing agricultural systems, were also targeted. Another target audience was represented by the Florida food, agri-food, and citrus industries. As shown in Figure 6, only 35 of the 74 research articles analyzed clearly stated their target audience. The frequency of target audiences is also displayed as a word cloud for a rapid overview. that explains why the same reference was used for multiple crop groups and why the number of studies on the chart exceeds the number of studies covered. Tomato production was separated into three categories since three types of tomato products (fresh tomato, canned tomato, and tomato ketchup) were considered, as indicated in the diagram.

Agricultural
We divided the products into 11 categories: tomato, fruits, citrus, vegetable, fresh salad, grains, seeds, oil, sugar, flower, and trees, as shown in Figure 7. The most common product was tomato; 13 studies analyzed tomato production, including fresh tomato, canned tomato (whole peeled, paste, and diced), and ketchup. The second most common product was wheat with nine studies. Because some studies involved more than one crop, that explains why the same reference was used for multiple crop groups and why the number of studies on the chart exceeds the number of studies covered. Tomato production was separated into three categories since three types of tomato products (fresh tomato, canned tomato, and tomato ketchup) were considered, as indicated in the diagram.

Scope
The scope defines the product system boundaries that determine which unit processes should be included in the LCA analysis and which should be excluded. Table A2 (Appendix A) includes more information on all 74 studies, including their inputs and outputs inside and outside of the scope. Most studies (14) contained three to four phases in their boundaries, as shown in Figure 8A. There are two explanations for not including the eliminated phases in the majority of articles. The first is a lack of data and knowledge about individual inputs, making it difficult to get a decent overall view. Secondly, some authors excluded the minor influence stages because it was impossible to include all phases.
cesses should be included in the LCA analysis and which should be excluded. Table A2 (Appendix A) includes more information on all 74 studies, including their inputs and outputs inside and outside of the scope. Most studies (14) contained three to four phases in their boundaries, as shown in Figure 8A. There are two explanations for not including the eliminated phases in the majority of articles. The first is a lack of data and knowledge about individual inputs, making it difficult to get a decent overall view. Secondly, some authors excluded the minor influence stages because it was impossible to include all phases.
Since we are looking at the agri-food supply chain, most of the articles noticeably had similar steps when designing their boundaries. Depending on the selected crop and the target audience, there were slight differences in the scope's starting point and finishing point ( Figure 8B). According to the review, 47 studies started their scope from the nursery stage (cradle), which involves preparing the raw materials, buildings, and field or land. Furthermore, 25 studies began their scope from the farming stage (farm gate). Considering our focus on agricultural production, only one study started their scope after the farming stage.
Similarly, the final stage differed from one study to another, ranging from the farming stage to the grave, including the product's processing, packaging, storing, and transferring stages. Thirty-one studies in the literature review included steps until the crop harvesting stage, whereas 16 authors included some or all of the processing, packaging, and storing stages in the study's scope. A number of reviewed studies reached the point of distribution and consumption in their analysis. Disposal and waste management were the final stages in some studies, with 10 articles including the end-of-life phase in their analysis ( Figure 8B). One study did not specific boundaries; thus, the number in Figure  8B is less than the number of studies reviewed [24].  Since we are looking at the agri-food supply chain, most of the articles noticeably had similar steps when designing their boundaries. Depending on the selected crop and the target audience, there were slight differences in the scope's starting point and finishing point ( Figure 8B). According to the review, 47 studies started their scope from the nursery stage (cradle), which involves preparing the raw materials, buildings, and field or land. Furthermore, 25 studies began their scope from the farming stage (farm gate). Considering our focus on agricultural production, only one study started their scope after the farming stage.
Similarly, the final stage differed from one study to another, ranging from the farming stage to the grave, including the product's processing, packaging, storing, and transferring stages. Thirty-one studies in the literature review included steps until the crop harvesting stage, whereas 16 authors included some or all of the processing, packaging, and storing stages in the study's scope. A number of reviewed studies reached the point of distribution and consumption in their analysis. Disposal and waste management were the final stages in some studies, with 10 articles including the end-of-life phase in their analysis ( Figure 8B). One study did not specific boundaries; thus, the number in Figure 8B is less than the number of studies reviewed [24].

Functional Unit
Another step of the goal and scope phase is to choose a functional unit of the scope. A functional unit is the reference unit in which elementary flows from the inventory until the impact assessment stage are represented. Selecting the ideal functional unit is necessary during the boundary designation step. The functional unit is dependent on the type of input materials (raw material) and the final products. Accordingly, the input unit might be separate from the outputs. For example, the output such as GHG emissions could be in kg·ha −1 while the final product could in tons or the input material could be in kWh for energy consumption and kg for fertilizers. Figure 9 shows the most common functional units used in previous studies.

Functional Unit
Another step of the goal and scope phase is to choose a functional unit of the scope. A functional unit is the reference unit in which elementary flows from the inventory until the impact assessment stage are represented. Selecting the ideal functional unit is necessary during the boundary designation step. The functional unit is dependent on the type of input materials (raw material) and the final products. Accordingly, the input unit might be separate from the outputs. For example, the output such as GHG emissions could be in kg·ha −1 while the final product could in tons or the input material could be in kWh for energy consumption and kg for fertilizers. Figure 9 shows the most common functional units used in previous studies.

Data Quality Requirement
The reliability of the results from LCA studies strongly depends on how data quality requirements are met. The following parameters should be considered: time-related coverage (selected year), geographical coverage (study area), and technology coverage (technology used in the processes stages). This paper examined the temporal and spatial data

Data Quality Requirement
The reliability of the results from LCA studies strongly depends on how data quality requirements are met. The following parameters should be considered: time-related coverage (selected year), geographical coverage (study area), and technology coverage (technology used in the processes stages). This paper examined the temporal and spatial data in detail and the used machinery in general.
It is understood from the literature review that most studies collected their data for a single year of cultivation ( Figure 10B). The spatial scale of the analysis (global or regional) depends on the impact category. For example, global warming is a worldwide issue, whereas acidification is a regional issue. Furthermore, two countries were commonly represented in the evaluated research, Italy and the United States, with 17 and 14 studies, respectively ( Figure 10A). When it comes to the technology used in each activity, the majority of the tools mentioned were agricultural equipment, which is to be expected given that we are investigating crop production.

Phase 2: Life Cycle Inventory
The second step of the LCA is the life cycle inventory analysis (LCI). The product's life cycle inventory results in an LCA study are obtained by summing up all fractional contributions of the input and output from each unit process in the product's production system. Thus, LCI generates quantitative environmental information of a product throughout its entire life cycle.

Phase 2: Life Cycle Inventory
The second step of the LCA is the life cycle inventory analysis (LCI). The product's life cycle inventory results in an LCA study are obtained by summing up all fractional contributions of the input and output from each unit process in the product's production system. Thus, LCI generates quantitative environmental information of a product throughout its entire life cycle.
Most studies at this stage specified the input material (water, fertilizer, pesticide, diesel, etc.) in each process of the production included in the scope, as well as the output (harvested crop, waste, emission to the air, soil, and water, etc.). Furthermore, they mentioned the sources of the inventory data ( Figure 11), typically being from primary and/or secondary data sources. Primary data are obtained from specific processes throughout the life cycle of the researched product. Process activity data (physical measures of a process that results in GHG emissions or removal), direct emissions data (determined through direct monitoring, stoichiometry, mass balance, or similar methods) from a specific site, or data averaged across all sites containing the specific process are all examples of primary data [25]. Secondary data are collected from government departments, organizational records, and studies that previously gathered information from primary sources and made it available to other researchers.
Climate 2021, 9, x FOR PEER REVIEW 12 o direct monitoring, stoichiometry, mass balance, or similar methods) from a specific s or data averaged across all sites containing the specific process are all examples of prim data [25]. Secondary data are collected from government departments, organizational ords, and studies that previously gathered information from primary sources and m it available to other researchers. About 48% of the studies used secondary data, 13% used primary data, and 35% u both. One study collected data from a real farm experience. Three authors conducted terviews with owners to collect the data. Two studies used surveys with specific questi to collect the required information. One study mentioned that the source was prima but the article did not specify their method. Seven studies utilized primary data, while About 48% of the studies used secondary data, 13% used primary data, and 35% used both. One study collected data from a real farm experience. Three authors conducted interviews with owners to collect the data. Two studies used surveys with specific questions to collect the required information. One study mentioned that the source was primary, but the article did not specify their method. Seven studies utilized primary data, while the other nine used secondary data. The authors of the examined research utilized two types of secondary data methods: databases and previous studies. Eleven of the studies used databases, while five of them used previous studies. Five writers, on the other hand, gathered inventory data from databases and prior studies. Twenty-six studies utilized both primary and secondary approaches to reduce the uncertainty of their findings ( Figure 12A,B).

Phase 3: Life Cycle Impact Assessment
In life cycle impact assessment (LCIA), the significance of a product system's potential environmental impacts, based on life cycle inventory results, is evaluated using LCIA. The LCIA consists of several elements: classification, characterization, normalization, and weighting. Of these four elements, normalization and weighting are considered optional,

Phase 3: Life Cycle Impact Assessment
In life cycle impact assessment (LCIA), the significance of a product system's potential environmental impacts, based on life cycle inventory results, is evaluated using LCIA. The LCIA consists of several elements: classification, characterization, normalization, and weighting. Of these four elements, normalization and weighting are considered optional, while the first two are mandatory elements in LCIA [10] (Figure 13). As shown in Figure 14, all 74 reviewed studies completed the classification and characterization phases, whereas 14 studies completed normalization and 10 completed weighting. Few studies included the waiting stage since it is optional and challenging.
The first step is classification, which involves identifying the impact assessment method. The most common standard method was the CML with various versions, such as CML 2 baseline 2000 V2/world, developed by the Center for Environmental Studies, and CML 2000 produced by the Center of Environmental Science of Leiden University. The second most common methods were ISO 14044 (2006), ISO (2000), and ISO 14040, followed by many other methods, such as IPCC 2001 GWP 100, proposed by the Intergovernmental Panel on Climate Change. For more information about the methods used in the studies, see Figure 15. The model used to calculate the impact is determined by the impact category the author intends to examine. As a result, LCA, ISO, and IPCC were the most commonly used impact methods since they provide categorization factors for ecotoxicity and climate change, which were among the criteria used to select articles for this review.   category the author intends to examine. As a result, LCA, ISO, and IPCC were the most commonly used impact methods since they provide categorization factors for ecotoxicity and climate change, which were among the criteria used to select articles for this review. Choosing the correct method for the LCA's impact assessment stage depends on the impact category under investigation. Each method has categories; for example, CML 2000 has 10 environmental impact categories: abiotic depletion, global warming, ozone layer depletion, human toxicity, freshwater aquatic ecotoxicity, marine aquatic ecotoxicity, terrestrial ecotoxicity, photochemical oxidation, acidification, and eutrophication.
In the process to quantify the impact of a procedure or material used, impact categories are first chosen, followed by quantifying environmental impact in each impact category using the equivalency approach. This process is termed "characterization" [10]. Characterization includes the emissions to air, soil, and water, as represented in Figure  16. The most prevalent impact categories in the 74 papers were human toxicity and ecotoxicity, with 48 and 41 studies, respectively. Moreover, 34 studies included global warming potential as an effect category, whereas marine pollution (26 articles), freshwater aquatic ecotoxicity (23 articles), and acidification potential (22 articles) were topics of the remaining studies ( Figure 16). Choosing the correct method for the LCA's impact assessment stage depends on the impact category under investigation. Each method has categories; for example, CML 2000 has 10 environmental impact categories: abiotic depletion, global warming, ozone layer depletion, human toxicity, freshwater aquatic ecotoxicity, marine aquatic ecotoxicity, terrestrial ecotoxicity, photochemical oxidation, acidification, and eutrophication.
In the process to quantify the impact of a procedure or material used, impact categories are first chosen, followed by quantifying environmental impact in each impact category using the equivalency approach. This process is termed "characterization" [10]. Characterization includes the emissions to air, soil, and water, as represented in Figure 16. The most prevalent impact categories in the 74 papers were human toxicity and ecotoxicity, with 48 and 41 studies, respectively. Moreover, 34 studies included global warming potential as an effect category, whereas marine pollution (26 articles), freshwater aquatic ecotoxicity (23 articles), and acidification potential (22 articles) were topics of the remaining studies ( Figure 16).

Phase 4: Life Cycle Interpretation/Recommendation Options
The primary purpose of interpretation, which is the last phase of the LCA, is to use the inventory results and impact assessment analysis to evaluate the starting point for product improvement. The starting point is to understand the process tree and then identify the key issues, i.e., the key processes, materials, activities, components, or even life cycle stages in developing a product. The primary purpose is followed up with improvement recommendations to find more environmentally friendly designs and/or process modification. Studies applied dominance analysis and marginal analysis to identify the key issues. The dominant aspects of the inventory table may be revealed by studying the

Phase 4: Life Cycle Interpretation/Recommendation Options
The primary purpose of interpretation, which is the last phase of the LCA, is to use the inventory results and impact assessment analysis to evaluate the starting point for product improvement. The starting point is to understand the process tree and then identify the key issues, i.e., the key processes, materials, activities, components, or even life cycle stages in developing a product. The primary purpose is followed up with improvement recommendations to find more environmentally friendly designs and/or process modification. Studies applied dominance analysis and marginal analysis to identify the key issues. The dominant aspects of the inventory table may be revealed by studying the environmental elements of a process matrix. An arbitrarily chosen criterion, such as "contribution greater than 1% of the total impact", can be applied in identifying key issues from the matrix. Marginal analysis illustrates the changes in the process to which the intervention, effect, or index is most sensitive. In theory, marginal analysis is a powerful tool in determining product improvement options [8,26].
Many studies stated that, for a complete understanding of the significant driver of the impacts, it is necessary to include all stages and material used through a product's life cycle, which is very challenging due to a lack of information and databases. However, depending on the aim of the LCA research, the literature review revealed a number of critical concerns, such as emissions from chemical and energy usage, the cultivation method used, land-use problems, and consumption waste.
Furthermore, studies in the literature proposed several recommendations for improving the agri-food system and reducing environmental consequences. One of them was adhering to the EPA and USDA pesticide and fertilizer guidelines. A frequent proposal was to use agricultural waste as animal feed. The most common request, however, was to enhance production without increasing inputs ( Figure 17).
Climate 2021, 9, x FOR PEER REVIEW 17 of 67 environmental elements of a process matrix. An arbitrarily chosen criterion, such as "contribution greater than 1% of the total impact", can be applied in identifying key issues from the matrix. Marginal analysis illustrates the changes in the process to which the intervention, effect, or index is most sensitive. In theory, marginal analysis is a powerful tool in determining product improvement options [8,26]. Many studies stated that, for a complete understanding of the significant driver of the impacts, it is necessary to include all stages and material used through a product's life cycle, which is very challenging due to a lack of information and databases. However, depending on the aim of the LCA research, the literature review revealed a number of critical concerns, such as emissions from chemical and energy usage, the cultivation method used, land-use problems, and consumption waste.
Furthermore, studies in the literature proposed several recommendations for improving the agri-food system and reducing environmental consequences. One of them was adhering to the EPA and USDA pesticide and fertilizer guidelines. A frequent proposal was to use agricultural waste as animal feed. The most common request, however, was to enhance production without increasing inputs ( Figure 17).

Discussion
The present study reviewed articles related to the environmental impacts of agricultural production in LCA assessment. The main steps in conducting an LCA are defining the purpose of the study and boundary stages involved in the analysis, collecting the data of the inventory phase, estimating the impact of the involved process and used material, and then identifying the key issues, followed up with improvement recommendations. Most studies followed these steps, and some of them had common impact categories. However, implementing LCA is challenging and necessitates meticulous data collection.

Choice of Time, Spatial Domain, and Elementary Flows in LCA
Nearly 17% of studies did not mention the temporal scale of their analyses, depicting the inherent limitation of ISO 14040/ISO 14044 in considering the time period of evolution and process variations pertaining to diverse impact categories. The highest temporal resolution obtained from the literature was seasonal (4% of studies). The choice of time in LCA depends on the spatial and temporal scale of the impact categories considered. For example, the temporal scale of ecotoxicity varies from hours to years. On the other hand, ecotoxicity impacts have multiple transport pathways such as air, water, and soil

Discussion
The present study reviewed articles related to the environmental impacts of agricultural production in LCA assessment. The main steps in conducting an LCA are defining the purpose of the study and boundary stages involved in the analysis, collecting the data of the inventory phase, estimating the impact of the involved process and used material, and then identifying the key issues, followed up with improvement recommendations. Most studies followed these steps, and some of them had common impact categories. However, implementing LCA is challenging and necessitates meticulous data collection. in LCA depends on the spatial and temporal scale of the impact categories considered. For example, the temporal scale of ecotoxicity varies from hours to years. On the other hand, ecotoxicity impacts have multiple transport pathways such as air, water, and soil emissions with diverse temporal scales. Establishing a time frame for the evaluation in LCA is challenging, as both very lengthy and very short periods of assessment are not practicable depending on the topic of the LCA. Extremely short timescales violate the concept of intergenerational equality, whereas extremely long ones marginalize short-term actions, lowering the incentive to act [27]. Consequently, care should be taken when defining the temporal scale of inventory flows.
About half of the studies (49%) used secondary data collection for the LCA, acquiring data from websites and previous studies. The studies that constituted primary datasets were fewer due to the trouble of obtaining data at the desired spatial/temporal resolution for the inventory flows. The selection of impact categories and spatial domains (Figure 16) clearly reflects a preference for secondary datasets. The major categories studied were human toxicity potential and terrestrial ecotoxicity (the primary contributor being agricultural pesticide emissions). Studies used the approximated characterization factor from models for a particular spatial and temporal horizon to assess the potential impacts. Multimedia chemical exposure models such as CalTOX [28], USES-LCA [28,29], IMPACT 2002 [30], and USEtox [31] can provide the time-dependent concentrations of a chemical in the environmental compartments of air, soil, water, plants, and sediments. The potential impacts are characterized on the basis of the chemical's fate in an environmental partition and its effect.

Impact Assessment
The quantity of the input material at each stage of the crop production chain can reduce GHG, as well as emissions, including energy use (diesel, fuel, electricity) both on farm (crop production, machinery use) and off farm (transportation, refrigeration). Additional emissions include fertilizer production and use (N, P 2 O 5 , K 2 O), pesticide use (fungicide, herbicide, insecticide), raw material production and transportation, packaging production, and disposal (Table A2). These sources of emissions contribute to environmental impacts in various ways, including human toxicity, terrestrial toxicity, freshwater toxicity, aquatic toxicity, global warming, and acidification ( Figure 16). It has been demonstrated that low-input crops have minimal impacts, but high-input crops have high impacts [32]. Furthermore, the type of input can affect the rate of the impacts. For example, replacing Thomas slag with triple superphosphate reduced the toxicity associated with the presence of heavy metals [33]. Simultaneously, replacing urea with ammonium nitrate reduced the influence of fertilization on eutrophication and acidity induced by ammonia volatilization [34].

LCA as a Tool in Environmental Policy Decisions
In order to achieve the population demand in the future, increasing food production is not the only pathway to increase food availability. Increased food production necessitates either more land or increased fertilizer and pesticide use on current arable land, with negative environmental consequences such as elevated GHG emissions, biodiversity loss, water contamination, and soil erosion [35]. That explains why, among the LCA papers, the most common target audiences were policymakers and producers, whereby policymakers regulate new policies for upcoming issues and producers follow these rules. The LCA methodology can be used to identify parameters and their variability in order to assist producers, wholesale and retail consumers, and policymakers in aligning their practices and purchasing decisions with low-carbon goals. LCA can also be used to analyze different production systems in order to quantify differences in input consumption and environmental consequences. The key parameters and their variability are then addressed to offer stakeholder metrics for evaluating and aligning their agricultural processes, purchasing decisions, and policies to optimize production supply chains.

Challenges in Collecting the Information and Limitations
Obtaining each LCA component from the reviewed studies is not simple for the reader due to the authors' descriptive and nonexhaustive approach. Section 3 shows that diverse communities can benefit from this study on a local, international, and global scale. Hence, the author could have used a table or a flow chart to present the flow of components and stages to summarize the four phases and their components to enable the reader to focus on helpful information.
Another challenge is to identify what information needs to be included in the phases of the LCA. One of the essential characteristics of phase one of the LCA is using a functional unit; some authors mentioned it in the goal section while others mentioned it in the scope section. Noticeably, studies with an economic purpose often did not clearly report the functional unit.
The necessity of incorporating all production processes and their input materials, analyzing all phases to understand the environmental effect, and obtaining an optimal outcome from the LCA analysis of food production systems was emphasized by researchers. However, that is neither possible nor practical because of data limitations and cost restrictions [10]. Accordingly, the minor influential stages were excluded. Hence, most studies focused on a single phase of the food production chain. For example, some studies focused on the cultivation phase because they considered that the food production system's environmental impact mainly comes from farming activities.
The literature review did not focus on a specific region or a crop. Consequently, many studies appeared while searching using the keywords. Therefore, we included 74 articles related to LCA in agricultural production in general, as well as GHG emissions and ecotoxicity as an LCA impact category.

Assumptions Used, Benefits, and Recommendations
The LCA of crops along a food supply chain can provide helpful information from an economic, social, and environmental perspective. Using the LCA, stakeholders can better understand the energy, water, and material input and evaluate the outputs' environmental impacts. Thus, they can regulate new policies and use modern practices to improve the production supply chains.
A substantial understanding of each phase of the LCA is required to present an accurate food product's environmental impact. This paper clearly explains the LCA's major components that can serve as a primer for the scientific community. Specifically, because LCA is a systematic tool that allows for analyzing a product throughout its life cycle, LCA is used to study the economic value and importance from the local and global perspectives.
If the final product's functional unit is introduced at either the goal or the scope stage, the study results would be unaffected from our perspective. However, we recommend illustrating the input's measurement unit and the outputs while illustrating the production scope, followed by a table of units to be more readable for the audience to understand at which stage the inputs are being used and to represent the elementary flows. Defining the system boundary determines the impact pathway for an impact category that links the elementary flows from inventory to the endpoint of analysis. It is clear that the system boundary processes need to be defined according to the study's goal and the impact category. Furthermore, the functional unit must be clearly defined to explain the elementary flows from inventory to the endpoint. It is essential to know the impact category that the LCA aims to estimate, which processes are related to it, and their cause-effect relationships. The impact assessment studies were mostly conducted in the European sector since most models and databases are suited for European agri-food products.

Research Gaps
The information obtained from the literature sheds light on some of the future research needs: (a) the impact of land use on GHG emissions [36], (b) LCA applications based on irrigation techniques using solar energy dealing with waste streams [37], (c) LCA of processed and homegrown vegetables [38], (d) packaging of foods with eco-design solutions [8], and (e) applications of LCA in organic agricultural practices, fertilization practices, mulching and milling techniques, and achievable production yields [39]. Some studies have called for more LCA applications in non-European and non-OECD countries to make their agri-food sector more environmentally friendly [40]. Therefore, it is understood that LCA can be used to make the agri-food supply chain more sustainable.
The inventory flows obtained from the present review point to the inter-dependency of three sectors in LCA: energy, food, and water. Consequently, policymakers can use LCA as a tool to spot the crucial areas that need improvisation within the framework of the food-energy-water nexus. Moreover, it is imperative to understand the drivers of environmental policy for selecting an environmentally friendly agri-food supply system. The regional variation of this nexus calls for more regional LCA assessments based on the allocation of resources. More research is needed to explore future scenarios [41] that drive resource consumption and policy design for long-term sustainability utilizing the LCA framework.

Conflicts of Interest:
The authors declare no conflict of interest.
Appendix A Table A1. Common aims in the selected studies.