As some of these foods are produced primarily outside of the regions of interest in this paper , they were cross referenced with the top ten most produced agricultural commodities in each of those regions to create a list of commodities to be considered in this paper.Table 3.1 shows which of the globally most produced food and agricultural commodities are also in the top ten most produced commodities of the United States, Europe and Australia. Note that some commodities, such as rice, vegetables, and cassava, are not widely produced in any of these geographic regions. Therefore, commodities of interest were those that existed in the both the global top ten most produced and the top ten list of at least two out of three regions. Figure 3.3 shows that enteric fermentation and three manure related activities , contribute 66.1% of agriculture’s emissions, resulting in significant interest in better understanding the environmental impacts of livestock farming. Beef is one of the top ten most produced commodities in both the United States and Australia. Similarly pork is one of the most produced commodities in Europe. As they are therefore responsible for a large proportion of global agricultural emissions, these two livestock farming commodities, vertical grow beef and pork, were added to the list of commodities of interest in this paper.Most of the LCA studies found, 18 of the 25, were comparative studies. Many looked at the differences between conventional and organic farming systems.
As there is a current push toward both growing and buying more organic food, it makes sense that the community is interested in finding empirical evidence to support the environmental benefits of organic food. Comparisons of different cropping systems that result in a similar processed product is also common: For example, three cropping systems are compared, as all three result in the production of vegetable oil. In all of the studies in this category, the agricultural systems are manually compared. Surprisingly, there were only two connective studies and one update study found. It may be the case that updates to LCI data are not commonly published in academic literature, and are instead directly updated in LCI databases . The lack of papers in the connective category may be evidence that there truly is a lack of connectivity across the plethora of LCA models created. Four methodology studies were found. There is substantial literature on improving LCA methods. The search criteria that were used aimed to find papers that specifically involved one or more LCA studies and their details. The papers in this category did conduct an LCA study, but the purpose was to test proposed methods.LCA Methods allow for the modeling of different types of agricultural systems . Some methods, such as EIOLCA, have been developed to try to reduce the overhead inherent in LCA, by allowing analysts to calculate estimates. However, the process is very involved, requires expertise in the method, and it is difficult to reuse models, thereby taking advantage of effort already expended. In general, accurate, thorough, and rigorous LCAs are effort-intensive. Each of the LCA studies presented in this chapter provide further insight into how the LCA technique is customized for and commonly used in the assessment of agricultural systems,such as the creation of hybrid LCA methods and streamline LCAs.
In this section, I articulate these findings as seven observations regarding LCA for agriculture.The most common type of LCA used in the analysis of agricultural systems is a cradle-to gate analysis, since once the product is ready for shipping, the storage, variety of packaging, distribution methods, preparation, and consumption, among others, vary widely. Table 3 lists the LCA types used in the representative sample of LCA studies, highlighting those that involved a cradle-to-gate LCA. The scope and boundary of the cradle-to-gate agricultural system LCAs are very similar to each other. The final product of the agricultural system is usually some form of raw product, such as a meat, grain, fruit or vegetable. The amount of processing this food product undergoes within the system of interest also varies widely in some cases, the final product is frozen, ground, deboned, packaged, or transformed into some derivative transportable product such as sugars from sugarcane.Table 3 overviews the functional units that are used in each LCA study. For instance, “1 ha of land used” is a popular metric, which allows for the calculation of energy intensity with respect to land use, demonstrating how much strain the system puts on the land. For single product systems, a functional unit is often in terms of produce weight, as it allows for the calculation of energy, emissions or impacts per unit weight of the product at the gate. The decision of where the gate lies depends the system boundary, i.e., which of these processing techniques occurs on the farm. For example, in beef production, the farm-gate may be pre-slaughter or post-slaughter. This also determines the functional unit: for pre-slaughter the functional unit would be live-weight, while post-slaughter, a common unit is Hot Standard Carcass Weight.Sometimes, midpoint functional units are used to analyze system subcomponents or to allow for the discretization of processes. For example, HSCW , is used.
It represents the end of the production process in the pork supply chain, i.e., the weight of the product at the slaughterhouse gate. This makes the unit incomparable to other pork or meat product LCA studies that may define the endpoint at the consumer side To address this issue, the analysts in the Australian Pork study also use two midpoint functional units: 1 live piglet and 1 live slaughter pig at the farm gate. These units allow the findings to be used in comparative studies. For reference, another pork production study, by Basset-Mens & Van der Werf, has a functional unit of 1kg of live slaughter pig as well, in addition to a land use unit . However, not all LCA studies have a midpoint functional unit or a functional unit that can be used to compare the models produced in the study with other studies, even if they are ISO compliant. Functional units can also be highly specific to the system of interest, product, or location. For example, the functional unit is “1kg of soybean meal produced in Argentina and delivered to Rotterdam Harbor”. The level of granularity is non-negotiable. While there must have been intermediate steps in the LCA that separated the different processes , these numbers are not always released or easily accessible. Various levels of detail are lost to the reader, and more importantly the system cannot easily be broken down into reusable components. Unfortunately, while the functional unit is meant to make LCA studies more comparable and reusable, it is not always the case.Renouf et al. perform a comparative assessment of the production of sugars for fermentation in Australia, corn production in the US, and sugar beets in the UK. The product of interest was a sugar suitable for fermentation, as its bio-products have wide use, including as an alternative energy source. Here, in addition to systems based in different locations, the initial crop is different too. The functional unit in this study is 1 kg of monosaccharide , as this enables comparability across sugarcane, corn and sugar beets. As opposed to conducting a separate LCA study based on the specific sugarcane farms, the researchers used data from a variety of Australian inventory databases, local survey data, and other academic publications that have looked at different processes within the sugarcane production system . Similarly, for the U.S. corn, best way to cure cannabis and the British sugar beet impact numbers, the researchers looked at two sets of studies for each case, and converted their functional units into 1 kg of monosaccharide, based on the yield numbers reported. As all the U.S. corn and British sugar beet studies had a high level of detail available in the report, the resulting analysis is precise comparison between the three sugar production systems.Heller et al. perform a very broad review of the United States food system by using a life cycle perspective to connect systems within different sectors of the industry. They use a product life cycle approach to analyze sustainability indicators across different life cycle stages: resource origin, growing and production, food processing, packaging, and distribution, preparation and consumption, and end of life management. This study is unique in that it attempts to address the entire US food system, connecting different agricultural systems without resorting to a sector based approach like EIO-LCA. Heller et al. did not conduct a new LCA, instead opting to review LCAs in published literature and connect information about the impacts that occurred at each stage to provide a holistic view of the food system.
It is still one of a small number of papers that attempts to connect impacts across products and agricultural sectors over a large region, thereby encompassing a sizable portion of the industry.Another massively-scaled LCA study is available in a report by the Center for Environmental Strategy at the University of Surrey by Mila i Canals et al.. The paper details a series of comparative LCAs, which combined aim to compare the environmental impacts of domestic versus imported vegetables. They compared broccoli production in the United Kingdom and Spain, salad in the UK, Spain, and Uganda, and finally, legume production in the UK, Uganda, and Kenya. The life cycle of each product is geographically disparate, therefore they break it down into three projects/reports chunked as follows: “cradle-to-central-depot”, “retail-to-plate”, and “consumption-to-waste” , The report highlights the importance of connecting LCAs across products, production systems, regions, up to connecting the entire industry. It is because of this highly detailed, connected set of LCA models that they can come to the surprising conclusion that local is not always more environmentally friendly.Once a flow diagram has been created, and the analysts have a feel for how resources move within the system, they gather all the data required to calculate different environmental impacts. This process involves decomposing the high level steps in the flow diagram into individual sub-flows or processes. The basic unit of LCI data that is collected is the “unit process”, defined by ISO 14044 as: “the smallest element considered in the life cycle inventory analysis for which input and output data are quantified” . For each unit process, inputs and outputs , and the associated environmental impact with it are listed. The question to be answered is: how does actually performing this step affect the environment? The data may be collected in several ways: primary data collection , data obtained from published literature , data obtained from the results of simulations of approximately similar systems, or through lookup in a Life Cycle Inventory database. Due to the scope of the LCA, the number and size of the agricultural systems under study, the type and level of detail of the LCA to be conducted, and the availability of existing data, the LCI phase can consume the most time, money, and effort. They contain structured collections of objects representing unit process data. An overview of some LCI databases used in Agricultural LCAs is available in Table 3.4 .Within the LCA studies surveyed in this chapter, the national or regional LCI databases that analysts interact with are: the United States Life Cycle Inventory database, the Australian National Life Cycle Inventory database, and the European Life Cycle Database. Additional databases are listed in Table 3.4. Although this list is not exhaustive, others exist, many of which contain smaller, specialized datasets. Many proprietary databases are populated via primary data collection performed by consultants in partner organizations. These proprietary databases often aggregate existing free databases, and/or resell other proprietary databases as part of a package deal. ecoinvent is such an example, and is popularly used to supplement data regarding machinery, infrastructure, or capital goods in an agricultural LCA. These data are often international in scope. Most databases , contain data gathered during process-based LCAs. Some new database initiatives focus specifically on agricultural and food systems, some of which are also included in Table 3.4. Some databases, such as USLCI and ecoinvent, only release data in the ecoSpold format, ELCD and GaBI only use ILCD. Whereas others, such as AusLCI,, have versions of their data in both formats.