Optimization of residents' dietary structure with consideration of greenhouse gas mitigation and nutritional requirements

Food-related GHG emissions are ultimately driven by residents' dietary structures. Influenced by socio-economic development and consumer performances, modification for residents' dietary structures should reflect the variations in food demand and nutritional requirements. An integrated approach was developed by incorporating autoregressive integrated moving average (ARIMA), multiple linear regression, and an interval linear programming model into a general life cycle analysis (LCA) framework. In detail, (a) the GHG emissions from agricultural systems were assessed in the LCA framework; (b) the variations in food demand influenced by socio-economic development were identified; and (c) a specific model for dietary structure optimization that focused on GHG emission reduction and dietary nutritional balance was established. The variations of food supply in different seasons and food demands in different age and gender groups were considered in the model. A case study was proposed to illustrate the application of the approach in Guangdong Province, China. Compared with the predicted food demands based on multiple linear regression models, per capita cereal, fruit, and meat intake under the optimized dietary structure would be slightly smaller than the predicted demands, as well as vegetable, bean and nut intake would be more than the predicted demands. Among different age and gender groups, adult and adolescent males would contribute the biggest GHG emissions of diets, due to their high daily energy requirement. Conversely, adult and older women would contribute the smallest GHG emissions in food consumption.