A new paper Food-Miles and the Relative Climate Impacts of Food Choices in the United States by Christopher Weber and H. Scott Matthews (Environ. Sci. Technol.,42(10), 3508–3513, DOI: 10.1021/es702969f ) conducted a life cycle analysis of the greenhouse gas outputs of foods in the USA (but they didn’t manage to include the impacts of land cover and soil change).

I expect weightings would probably not change a lot if other environmental impacts, such as declines in other ecosystem services, were considered as well. However, I expect they would change a lot in countries with less intensive agriculture. It would be interesting to see someone do the math for these cases.

Environmental Science and Technology news, reports on a paper in Do food miles matter?:

It’s how food is produced, not how far it is transported, that matters most for global warming, according to new research published in ES&T (DOI: 10.1021/es702969f). In fact, eating less red meat and dairy can be a more effective way to lower an average U.S. household’s food-related climate footprint than buying local food, says lead author Christopher Weber of Carnegie Mellon University.

Weber and colleague Scott Matthews, also of Carnegie Mellon, conducted a life-cycle assessment of greenhouse gases emitted during all stages of growing and transporting food consumed in the U.S. They found that transportation creates only 11% of the 8.1 metric tons (t) of greenhouse gases (in CO2 equivalents) that an average U.S. household generates annually as a result of food consumption. The agricultural and industrial practices that go into growing and harvesting food are responsible for most (83%) of its greenhouse gas emissions.

For perspective, food accounts for 13% of every U.S. household’s 60 t share of total U.S. emissions; this includes industrial and other emissions outside the home. By comparison, driving a car that gets 25 miles per gallon of gasoline for 12,000 miles per year (the U.S. average) produces about 4.4 t of CO2. Switching to a totally local diet is equivalent to driving about 1000 miles less per year, Weber says.

A relatively small dietary shift can accomplish about the same greenhouse gas reduction as eating locally, Weber adds. Replacing red meat and dairy with chicken, fish, or eggs for one day per week reduces emissions equal to 760 miles per year of driving. And switching to vegetables one day per week cuts the equivalent of driving 1160 miles per year.

Several other recent studies have analyzed particular foods and poked holes in the food mile concept. For example, it can be more energy efficient for a British household to buy tomatoes or lettuce from Spain than from heated greenhouses in the U.K.

The new work expands on those studies by providing a comprehensive look at the U.S. food supply. Weber used an input–output life-cycle assessment, which counts not only the CO2 produced when food is shipped but also all greenhouse gases, including methane (CH4) and nitrous oxide (N2O), emitted from farm production. This means counting all the way back to the fossil fuels used to manufacture fertilizer and tractors.

“There is more [total] greenhouse gas impact from methane and nitrous oxide than from all the CO2 in the supply chain,” Weber says. In large part, he adds, this is because N2O and CH4 emission in the production of red meat “blows away CO2”. Cows burp CH4, and growing their feed uses large amounts of fertilizers that are converted to N2O by soil bacteria.

Update: Simon Donner on Maribo also discusses this paper.