Energy use and the sustainability of intensifying food production

Energy fuels, and is central to, all physical and biological systems, including the human population and economy. Yet science has missed the significance of civilization’s growing energy consumption. The energetics of the global food system illustrate the counterintuitive aspects of present energy consumption circumstances.

Contemporary humans use enormous amounts of energy (Fig. 1). Compared to the billions of years needed to store billions of tons of living biomass in complex ecosystems and vast reserves of fossil fuels, the speed of our recent 70-year consumptive flash is an explosion. Because energy is the ability to cause change, the laws of physics obligate an exponentially rising energy discharge begets exponentially rising change. The rapidly rising Earth-system transformations documented in the ‘Age of Acceleration’ metrics, for example, are a few of the expected consequences now underway1.

Fig. 1: Global primary energy consumption and food production by human society.

Food production is shown from 1965 to 2013 (dashed line). Food systems from farm- or ocean-to-table require from 15% (dark green) to 30% (light green) of total primary energy consumed.

One reason we fail to comprehend the gravity of these circumstances is that few really understand the fundamental concept of energy. Most of energy science focuses on energy quality, quantity and source. But all energy used as intended, causes change, thereby altering materials and changing the environment. Metaphorically, it is not only the origin, quantity or quality of the gas in your tank, it is what you are doing to the environment with your truck. Thus, for example, it only partially matters if this forcing function is from renewable or non-renewable fuels. We are detonating a large amount of energy in a closed system and causing dramatic environmental change and no mechanical or environmental system can indefinitely withstand an exponentially rising forcing function.

The role of energy in technology and innovation in the modern industrial–technological–informational economy is complex2. For example, in the mid-1800s, the English economist William Stanley Jevons showed how improved efficiency of engineering designs often leads to increased — rather than reduced — energy and resource consumption; the innovations allow the technologies to be more profitable, with lower prices leading to increased demand3. Energy consumption and the other rapidly accelerating socioeconomic markers depicted in the Age of Acceleration metrics aptly represent the pièce de résistance of Jevons’s predictions. The materials and energy ‘saved’ through more efficient innovative technologies are not preserved in the biosphere, but simply diverted to alternative uses, with concomitant even more rapid depletion of the resources.

Global food-system energetics and recent hopeful shifts toward ‘sustainable intensification’ of our food system provide one example of the complex and sometimes counterintuitive energy reduction challenges which lie ahead.

Food systems and energy

All food comes from plants, which transform sunlight into organic energy. To support biological metabolism, every human being requires a minimum ration of about 100 W per capita (~2,000 kcal d–1). Our hunter–gatherer ancestors used an inconsequential amount of Earth’s biomass energy to meet this need, because they lived at very low population densities and used the food to fuel their biological metabolism. The rise of contemporary civilization was accomplished by burning first living biomass and then fossil fuels, and using the resulting energy to perform the work of first agricultural, then industrial and informational economies. Humans currently use more than 2,300 W per capita, to fuel population, innovation and economic growth. Estimates are that 15–30% of humanity’s primary energy consumption is used for the global food system for material things including machines, fertilizers, pesticides and irrigation water; and for dynamical processes including harvesting, transportation, storage, refrigeration, marketing and preparation (Fig. 1). Gradients of energy intensity extend from subsistence small gardens and pastures up to mechanized industrial agriculture and marine fisheries.

We can be more quantitative (Table 1)4. Allowing a margin of calories to achieve enough macro and micronutrients, an adequate diet requires 2,000–2,400 kcal d–1. Thus, using 2013 data, a global population of 7.2 billion needed 22–26 EJ yr–1 of food (Fig. 1) (EJ = exajoule = 1018 joules). Humans cultivated edible harvests (65.1 EJ yr–1) or facilitated the additional growing and grazing of inedible harvests (41.8 EJ yr–1) totalling 106.9 EJ yr–1 of plants to supply downstream 27.7 EJ yr–1 of edible plant and animal calories at the table. The wide 15–30% variation in energy expenditure (from 80 to 160 EJ yr–1, depending mostly on the use of fossil fuels) to grow and get these calories to the table is due to the diversity of farm-to-table systems, the speed of the transportation and marketing, and the low resolution of the data and modelling. Nevertheless, 3 to 6 units of mostly fossil energy (80/27.8 = 3, 160/27.8 = 6) are reasonable estimates of the energy needed to acquire and deliver each unit of food energy. The many complicated detailed issues underlying the entries in Table 1, including the uncertainty bounds for every value, are outside the scope of this Comment, but the present data allow us to make a few systems-level observations for discussion.

Table 1 Global crop and meat energy flows for 2013

Methods intended to improve environmental impacts by reducing energy consumption can have counterintuitive consequences. For example, liquid biofuel energetics are problematic. Liquid biofuel manufacturing starts with 5.8 EJ yr–1 of mostly human-edible plant energy, diverted from the food system, to produce only 2.9 EJ yr–1 of ethanol and biodiesel (Table 1)5. Studies show 2.9 EJ yr–1 of liquid biofuel requires roughly equivalent amounts of fossil fuels to produce. Currently, liquid biofuels represent a lacklustre 2.6% of the 110 EJ yr–1 of petroleum used for global transportation and meaningful growth in the future has limits given growing land-use constraints including conservation needs6. Also, as this growth is pursued, to compensate for their lower energy densities (14% less for biodiesel and 32% less for ethanol), more liquid biofuels are needed to replace equivalent liquid petroleum.

Excess meat consumption is the more commonly known high-energy luxury. Edible plant calories (19.1 EJ yr–1) together with harvested or grazed inedible calories of grass, pasture and stover (41.8 EJ yr–1) combine for a total feedstock of 60.9 EJ yr–1 to later produce only 6.5 EJ yr–1 of meat, dairy and fish for consumption (Table 1). Most calories lost (54.3 EJ yr–1) when converting plants to animals are due to unavoidable metabolic losses. The global rate of feedstock return represented here for 2013 data is typical (6.5/60.9 = 0.11 = 11%)4. While feedstock return efficiencies improve with economic development, for example as diets shift from beef to poultry and pork, or through improved genetics and feeding technology and controlled environment production facilities, conversion efficiencies remain constrained from large stepwise advances by the biological limits of the metabolism required to morph plants into meat.

From this perspective, reducing meat and liquid biofuel consumption offers a remarkably clear improvement to energy consumption and upstream resources, particularly land and water use. Obviously, we recognize the enormity of the socioeconomic, behavioural, and nutritional complexities and challenges associated with these paths forward. But we also contend that a sustainable future without substantive reductions in energy consumption is not possible.

Alternatively, we also pursue additional innovation and technology to simultaneously improve yields and restore natural balances throughout the food system7. For example, better management of material resources at smaller scales with big data, informatics, artificial intelligence and other technologies normally target various aspects of the 15.9 EJ yr–1 in food-system processing losses (Table 1). These data-intensive redesigns can provide a greater degree of balanced control of organic material resources like carbon, nitrogen, soils, water and biodiversity, or services like harvest date, transportation and refrigeration temperature. Technology has also recently produced genetically modified crops, for example, which reduced photorespiration losses, thereby resulting in increased yields.

However, in accordance with Jevons’s paradox, the pros and cons associated with using technology to increase yields and reduce processing losses remains a convoluted cost–benefit multi-scale systems science, particularly as the outcomes are judged on the concept of global sustainability and energy consumption. The indirect energy expended for education, housing, food, mining, manufacturing, transportation, infrastructure, and ongoing energy and information flows required by existing and new technologies are largely unknown, difficult to quantify and require a multi-scaled approach to be successful8. For example, rapid growth in data generation, storage and analysis are energy intensive and generate indirect damages removed from the daily direct energy needs of food-system operations9. Energy reduction or efficiency improvement success stories, in an otherwise rapidly growing economy (Fig. 1) risk being largely antidotal, or worse, stimulating additional economic growth through increases in useful energy availability. Given these convoluted trends over the last 70 years, reducing energy consumption through innovation is a far more complex science than we have previously acknowledged.


With regard to the future, we should first concede that although scientists understand the role of energy in agriculture10, our reduction commitments, ability to measure, and ability to improve these energy returns on investment ratios in the food system continue to remain unknown or at least insufficiently focused11. The current state of energy modelling at all scales of food systems is time consuming, expensive, involves extensive or proprietary data, and altogether faces a mindboggling quantity of farm-to-table complexities. (How much energy is used to manufacture drip tape internationally, transport, and install it in a strawberry field in California’s rural Central Valley?) Hopeful reductions in CO2 or a need for increased yields per energy input are mentioned12,13 and resource-conserving redesigns like integrated crop or pest management or organic agriculture methods can demonstrate energy savings on a case-by-case basis14,15. But energy efficiency assessments throughout all scales of the food system are not universal components of detailed redesigns16,17 or are only partially addressed, for example, through greenhouse gas impact studies18,19.

Altogether, the enormous energy inputs required for the global food system remain an Achilles heel and serve as a metaphor for energy consumption in a growing society. The energy outlays dedicated to transportation, medicine and construction have similar confounding quandaries. Nevertheless, we contend that stepwise reductions in global energy consumption are unavoidable in any credible sustainability plan. But solutions to reduce society’s power curve will require humans choosing to do less with less, which is not currently being pursued20. For example, limiting human behaviour such as consuming less meat clearly provides more predictable and sizable energy reductions. Otherwise, we need innovation, but not with an uncertain but otherwise rapid increase in global energy consumption. For example, the current expansion of additional liquid biofuel production to meet the Intergovernmental Panel on Climate Change (IPCC) decarbonization goals is questionable. The energetics of liquid biofuels suggest a bigger problem with the promise of bioenergy in total, whose production consumes fossil fuels, and also competes for land, water, and food production, yet is still rapidly falling behind its decarbonization promises necessary to meet near-term climate goals established by the IPCC. Science has failed to properly convey the extent of runaway energy discharge, its front-seat role in innovation, and its fundamental and dangerous role in the rearrangement of nature. Malthus, Jevons, Carson, Ayres and Kneese, H. Odum, Ehrlich, Meadows, and others articulate the outcomes attributed to runaway resource consumption in a closed system. We point out the underappreciated simplicity, importance and consequence of our energy consumption rate.


  1. 1.

    Steffen, W., Broadgate, W., Deutsch, L., Gaffney, O. & Ludwig, C. Anthropocene Rev. 2, 81–98 (2015).

    Article  Google Scholar 

  2. 2.

    Nekola, J. C. et al. Trends Ecol. Evol. 28, 127–130 (2013).

    Article  Google Scholar 

  3. 3.

    Jevons, W. S. The Coal Question 2nd edn (Macmillan and Company, 1866).

  4. 4.

    Berners-Lee, M., Kennelly, C., Watson, R. & Hewitt, C. N. Elem. Sci. Anth. 6, 52 (2018).

    Article  Google Scholar 

  5. 5.

    Biofuels for Transport (IEA, 2019);

  6. 6.

    Hall, C. A. S., Dale, B. E. & Pimentel, D. Sustainability 3, 2413–2432 (2011).

    Article  Google Scholar 

  7. 7.

    Basso, B. & Antle, J. Nat. Sustain. (2020).

  8. 8.

    United Nations, European Union, Food and Agriculture Organization of the United Nations, International Monetary Fund & Organisation for Economic Co-operation and Development, World Bank System of Environmental-Economic Accounting 2012-Central Framework (United Nations, 2014).

  9. 9.

    Digitalization and Energy (ed. Technology and Outlooks Directorate of Sustainability) (International Energy Agency, 2017).

  10. 10.

    Pimentel, D. & Pimentel, M. Food, Energy, and Society (CRC Press, 2008).

  11. 11.

    IPCC Climate Change 2014: Mitigation of Climate Change (eds Edenhofer, O. et al.) (Cambridge Univ. Press, 2014).

  12. 12.

    Rockström, J. et al. Ambio 46, 4–17 (2017).

    Article  Google Scholar 

  13. 13.

    Sachs, J. et al. Nature 466, 558–560 (2010).

    CAS  Article  Google Scholar 

  14. 14.

    Mader, P. et al. Science 296, 1694–1697 (2002).

    CAS  Article  Google Scholar 

  15. 15.

    Stolze, M., Piorr, A., Haring, A. & Dabbert, S. The Environmental Impact of Organic Farming in Europe Vol. 6 (Hago Druck & Medien, Univ. Hohenheim, 2000).

  16. 16.

    Pretty, J. et al. Nat. Sustain. 1, 441–446 (2018).

    Article  Google Scholar 

  17. 17.

    Pretty, J. et al. Environ. Sci. Technol. 40, 1114–1119 (2006).

    CAS  Article  Google Scholar 

  18. 18.

    Tilman, D., Balzer, C., Hill, J. & Befort, B. L. Proc. Natl Acad. Sci. USA 108, 20260–20264 (2011).

    CAS  Article  Google Scholar 

  19. 19.

    Rockström, J. et al. Nature 461, 472–475 (2009).

    Article  Google Scholar 

  20. 20.

    Grubler, A. et al. Nat. Energy 3, 515–527 (2018).

    Article  Google Scholar 

Download references


We thank C. A. S. Hall for his insightful guidance during the development of this Comment.

Author information



Corresponding author

Correspondence to John R. Schramski.

Ethics declarations

Competing interests

The authors declare no competing interests.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Schramski, J.R., Woodson, C.B. & Brown, J.H. Energy use and the sustainability of intensifying food production. Nat Sustain 3, 257–259 (2020).

Download citation

Further reading


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing