|WHITE, ROBIN - Virginia Polytechnic Institution & State University|
|Hall, Mary Beth|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 3/15/2016
Publication Date: 9/12/2016
Citation: White, R.R., Hall, M. 2016. Should we feed animals to feed people? An optimization-based evaluation of environmental, economic, and health aspects of human diets in the United States [abstract]. 5th EAAP International Symposium on Energy and Protein Metabolism and Nutrition, September 12-15, 2016, Krakow, Poland. p. 10-26.
Technical Abstract: Given global pressure to improve food security, there is increasing interest in the question of whether we should feed animals to feed humans. Another question is whether there is sufficient domestically produced food to meet a population’s nutrient requirements. A linear programming approach to identify least-cost nutritionally adequate diets can be used to objectively formulate human diets for comparing across-alternative food availability scenarios. Such a formulation approach is regularly used by livestock nutritionists to identify least-cost, nutritionally adequate rations for animals. The objective of this study was to evaluate the ability of food production in the U.S. to adequately meet the nutrient requirements of the current U.S. population with or without domestically produced food derived from animal products, and to determine the daily cost of diets, environmental impact, and nutrients available for export to other countries in these systems. Current U.S. food consumption patterns and optimized diets derived from systems with or without livestock products were evaluated. Domestic production of 93 non-animal and 24 animal-derived food products was sourced from USDA/ERS, and chemical composition (41 nutrients) of each food product was used to identify the total amount of energy, amino acids, vitamins, minerals, and fatty acids produced domestically. From those 41 nutrients, requirements for 36 were identified for humans based on age and gender. Distribution of the U.S. population by analogous age and gender groups was used to identify a weighted average nutrient requirement of the U.S. population. Carbon footprints for all food products were sourced from International Standards Organization-compliant life cycle assessments specific to U.S. production systems. The potential effects of removing livestock from U.S. agriculture were simulated following these assumptions: 1) grain previously consumed by animals will be available for human consumption; 2) land previously used for hay and silage production and cropable pasture and grazing land will be used for human food production; 3) the nutrients from animal products previously provided to humans will no longer be available; 4) emissions from livestock production will no longer occur; and 5) human-edible grain and oilseed processing byproducts will be available for consumption by people, but human-inedible byproducts will need to be disposed of by alternative avenues (composting, burning, etc.). Current food consumption patterns suggest that the U.S. diet is comprised of 40% animal products, 25% vegetables, 24% fruits, 10% grains, and 1% nuts and legumes. This diet is consumed at a rate of 1,336 g/d and costs an average of $3.91/d. With the entire U.S. population consuming this diet, 85.9 x 106 yearly human energy requirement equivalents of food are available for export. Additionally, 59.9 x 106 yearly human protein and 17.8 x 106 yearly human lysine requirement equivalents can be exported. The carbon footprint of this diet was calculated at 3.3 kg CO2-equivalents/d. For comparison, a least-cost diet was balanced using the constraints that all nutrients must be greater than the minimum requirement, and that the total amount of any foodstuff consumed by the U.S. population has to be less than the total domestic production of that food. This least-cost diet retailed for $1.42/d, had very similar consumption patterns to the current diet but required less food per day (1,200 g/d), and had a lower carbon footprint (0.92 kg CO2-equivalents/d). By consuming less food, substantially greater quantities of energy (418 x 106 yearly requirement equivalents) and protein (546 x 106) were available for export. The same optimization exercise was conducted on the simulated system without animals. Despite greater total availability of energy, a feasible solution could not be identified, me