Multi-layer systems approach for assessing the socio-economic metabolism of food
MetadataShow full item record
Population growth combined with increased per capita food demands and changing diets will have far-reaching resource and environmental consequences. This is primarily due to food production systems being both i) dependent on the availability of energy and finite resources, such as fresh water, land and phosphorus, and ii) inefficient, resulting in large losses, high resource requirements and the generation of substantial amounts of pollution. In addition, material cycles in food systems are coupled, where changes in the management of one material will lead to intended and unintended consequences for others. This makes solving food system challenges complex, as, first, the resource and environmental consequences of production must be understood and, secondly, cycle linkages must be considered in order to evaluate tradeoffs. Currently, a wide variety of approaches are used to address food systems including technological methods (e.g. genetic engineering and adjusting feeding techniques), frameworks (e.g. the waste hierarchy), and analytical tools (e.g. life cycle assessment and material flow analysis). However, we argue that these methods do not effectively address resource challenges because they either i) disregard the physical system for which they are trying to change and/or ii) are based on a fragmented or aggregated understanding of the system and do not allow for an evaluation of impacts on larger scales. As a result, alternative approaches are needed that consider the full systems context both along supply chains and between materials. In this thesis, we test a socioeconomic metabolic approach, multi-layer material flow analysis (ML-MFA), to address food production systems under the postulation that this approach considers the systems context, links cycles and creates synergies. We employ this method to identify methodological and policy benefits and limitations through four case studies, represented as appended articles 1-4. We find a major strength of ML-MFA is that the obtained results are based on mass balance consistency both along the supply chain and between layers and, thus, robust conclusions can be made. With such an approach, cycle linkages are made explicit, the results reflect reality and are reproducible, and uncertainties are transparent. This makes for effective policymaking, as it i) allows for a detailed and comprehensive assessment of different materials and their coupling effects and ii) enables the identification of goal conflicts and ensures balanced priorities. In addition, once initial modeling efforts are made, updating and refining the system as well as further layer development becomes trivial. Thus, after model development, continual monitoring of the ever-changing socio-economic landscape is feasible, requiring only minimal time and resource inputs. Barriers of ML-MFA primarily relate to the time, resource and data intensive nature of initial modeling efforts. These limitations, however, are primarily due to i) substantial data gaps for food systems and ii) the lack of a systems approach during the collection, formatting and reporting of statistics. Thus, considerable efforts are required to collect, understand and interpret how data fit into the systems context. However, we find that, in a broader sense, the above barriers are largely caused by the tendency of societies to monitor individual activities instead of entire systems. This is not only the case for food systems but also all parts of society (e.g. manufactured goods, services, etc.). In this thesis, we argue that optimizing resource/environmental systems requires a transition towards monitoring systems through a concerted effort from policy, industry and researchers. In order for this to occur, an essential first step is to improve i) data availability and ii) data reporting, through the use of a systems understanding as the basis for data collection, by all parties. This would increase the robustness of results and, also, substantially reduce modeling efforts, costs and time demands. Secondly, monitoring systems requires the involvement of researchers, government institutions and industry and, therefore, the role of each party needs to be defined. We argue that governmental institutes should have the task of maintaining databases, while research and industry both report and supply data. All parties would have the task of building models, informing policy and onitoring the changing landscape.