What We Publish

Food without agriculture

Steven J. Davis, Kathleen Alexander, Juan Moreno-Cruz, Chaopeng Hong, Matthew Shaner, Ken Caldeira, and Ian McKay · Nature Sustainability 7, 90-95 · 2023

Key finding. Dietary fats can be synthesized chemically at under 0.8 grams of CO2-equivalent per kilocalorie, well below the more than 1.5 grams per kilocalorie now emitted producing palm oil in Brazil or Indonesia — and with large reductions in land and water use besides.

Two contour plots of greenhouse gas emissions per edible kilocalorie. Panel a, agricultural fats, plots land-use emissions against energy emissions in grams of CO2-equivalent per kilocalorie, with red dots for Sub-Saharan African oilcrops near 3.5, Brazilian soy above 2.5, and Indonesian, Brazilian and Colombian palm oil between 1.0 and 2.0. Panel b, synthetic fats, plots feedstock emissions against the emissions intensity of the energy used, with dots for coal and natural gas feedstocks and, at the origin, carbon dioxide feedstock with non-fossil energy at close to zero.
The two panels are on the same units, and the synthetic route (b) reaches values the agricultural route (a) cannot. Real palm oil and soy sit between 1 and 3.5 grams of CO2-equivalent per kilocalorie; fats synthesized from captured carbon dioxide with non-fossil energy approach zero. Figure 2 from Davis et al. (2023), Nature Sustainability 7, 90-95. Reproduced under CC BY 4.0. Extracted from the published PDF and resized for web display.

What question did this research address?

Making food systems more sustainable is almost always framed as making agriculture less damaging: less land, less water, fewer emissions per tonne. The alternative — producing food by chemical or biological synthesis, with no farm at all — has attracted comparatively little attention or money.

This paper asked whether that alternative is quantitatively serious for at least one class of food, and what the emissions, land and water consequences would be.

What did we find?

The analysis targets fats specifically, because they are chemically simple enough to synthesize at scale and because they are a large share of dietary calories.

Synthesis comes in below 0.8 grams of CO2-equivalent per kilocalorie against palm oil's more than 1.5 grams — less than half, before counting the land and water that agriculture also consumes.

The land and water reductions are described as enormous rather than marginal, since synthesis needs neither cropland nor irrigation.

The authors are explicit that a broad array of attractive chemosynthetic foods remains a possibility rather than a near-term prospect. Fats are the tractable case, not a template for everything on a plate.

Two obstacles are named that are not technical. Scaling up would disrupt agricultural economies, and it depends on consumers being willing to eat the result.

Why does it matter?

It reframes what "sustainable food" can mean. The entire debate is normally about which crops and livestock to favour and how to farm them better; this proposes that some food need not be farmed at all, and puts a number on the advantage.

Fats are a good place to start precisely because they are unglamorous. Much of their dietary role is as an ingredient rather than as a recognisable food, which lowers the consumer acceptance barrier that would block synthetic staples.

The disruption caveat deserves its place in the finding rather than a footnote. Palm oil supports large agricultural economies, so displacing it is a distributional question as much as an environmental one.

Citation

Steven J. Davis, Kathleen Alexander, Juan Moreno-Cruz, Chaopeng Hong, Matthew Shaner, Ken Caldeira, and Ian McKay (2023). Food without agriculture. Nature Sustainability 7, 90-95.

Related