Corresponding author: Nathaniel Chu



Over the last 200,000 years, 100 billion people have lived and died on Earth.1 And although only some of them got to play with fire, books, and e-bikes, every single one of them needed to eat.

Food is our first and most important industry. Most major early human inventions (tools, fire, language, writing, agriculture) served to get more food.2 Our quest for ever-greater amounts of food reshaped the planet: we ate all the large animals3 and turned half the world’s habitable land into farms.4 What we eat is also a primary cause of climate change.


Food is responsible for more greenhouse gas emissions than all fossil fuels burned to date

In our current “holy crap” climate moment, we tend to focus on energy for machines: Greenhouse gases from burning fossil fuels are altering climate patterns.5 To feed the energy needs of an industrializing world, we have over all of history released about 1.5 trillion tons of CO2 into the atmosphere from burning fossil fuels: 500 billion tons of CO2 from oil, 250 billion from natural gas, and 700 billion from coal.6

But this focus misses the fact that food is energy for humans. Yes, burning fossil fuels releases greenhouse gases; what fewer people realize, however, is that agriculture—particularly land-use change—does the same. If we broaden our view of climate change to account for the footprint of food and agriculture, we find two surprises:

  1. Carbon released from agriculture-driven land-use change (most of which occurred since the industrial revolution) outstrips all carbon released from fossil fuels to date
  2. Reforestation of that land represents an equally huge opportunity for carbon sequestration

Although our use of fossil fuels is rightfully a focal point in the sustainability fight, it’s easy to forget that we used those fossil fuels in part to dramatically change the face of the planet. Those changes set us up for our climate crisis.

Today, agriculture directly contributes 17–34% of annual greenhouse gases (including methane from cow burps and ammonia production).7 But this present-day activity is dwarfed by the historical greenhouse gases released from converting half the world’s habitable land into fields.

When we convert a forest or grassland into farms or pasture, the carbon that was previously stored in those trees and soils is released into the atmosphere. Either the living matter is burned (slash-and-burn agriculture8), or it decomposes. Both processes release CO2.

Since the agricultural revolution, agricultural land-use changes have released 1.7 trillion tons of CO2.9 The vast majority of that land-use change came after the 1700s, as globalization and fossil fuels created the economic systems and industrial machines that led to rapid forest-to-farm conversion.10 This spread of agriculture and the green revolution of the 20th century11 fueled the dramatic increase in population from 600 million people in 1700 to 8 billion today. We effectively doubled our emissions by using fossil fuels to cut down trees.

The upshot: Food alone is responsible for more carbon emissions than all fossil fuels burned to date12 (Figure 1).

Figure 1. Agriculture is responsible for more carbon emissions than all fossil fuels burned to date across industries, largely because of land-use change. These values (1.7 trillion tons of CO2 from land use change and 0.1 trillion tons from fossil fuels burned for agriculture) don’t even account for the majority of current day agricultural emissions (see left panel “Agricultural processes”): methane from cows and waste decomposition. Allowing land to return to forests and prairies represents the potential to sequester the same enormous amount of carbon emissions. See footnotes for references, calculations, and caveats.

The silver lining of these emissions is that the process is reversible: Allowing agricultural land to return to its forested state could absorb the same staggering amount of CO2. Converting to renewable energy can prevent emissions (carbon neutral); changing the way we eat could actually absorb emissions (carbon negative). So while we can’t stop growing food, improving land-use efficiency is one of our most powerful tools to combat climate change. Not to mention that none of the above gets at the inherent value of biodiversity and natural ecosystems.13


Cows (beef and cheese) are the least efficient food

Of the different foods we eat, cows are uniquely inefficient. In every metric (land-use, carbon emissions, water, pollution), food products from cows are often the worst by a large margin.14 They are especially land-intensive, with 41% of the land area of the United States used for cow pasture or feed—land that might instead be forests or grasslands.15

Cows also produce greenhouse gases directly. They are one of the two largest sources of anthropogenic methane emissions16—have you ever admired the smells on I-5 near Coalinga, CA?—and, along with other livestock, contribute 15% of annual global greenhouse emissions—more than all of transportation.17 Cows are inefficient as food because they require a lot of land, grow slowly, and use 83% of the energy they get from plants to stay warm and walk around instead of making beef or milk.18 Cows are so inefficient that, after beef and lamb, our most inefficient food is cheese,13 not even another meat.

We also raise a lot of them—more than 1.5 billion globally.19 Cows outweigh humans.20 They also outweigh chickens by 12 times.21

Compare this inefficiency with plants: The entire world’s protein needs could be met by a soybean farm the size of 1% of global habitable land22 or less than eight times the size of Iowa. Satisfying this need with cheese would require 120% of global habitable land or a farm covering the entire continents of Asia, Africa, North America, South America, and Europe (Figure 2). To do this with beef alone, we’d need 955% of habitable land or nine more planets.23 What’s more, soy and many other plants24 don’t emit methane and require less water, machinery, fertilizers and pesticides simply because it’s more efficient to grow a pound of soybeans than to grow 7–10 lbs of soybeans, feed them to a cow, and get a pound of beef or cheese.

Figure 2. Cows are an inefficient way to create protein from land that could be forests or grasslands.

We need to stop eating cows and start eating plants. But there’s one problem: cows and their milk are delicious. The sizzle of a steak, the juiciness of a burger, the stretch of pizza: no one can deny that cows make fantastic foods from plants that are way less fun to eat. For much of history, they provided critical nutrition in nonindustrialized food systems where calories and easily digestible protein were rare, and they often did it by grazing on land no longer suitable for agriculture.25 For these and other reasons, the meat and dairy industry grew to its current astonishing scale, forcing us to ask an uncomfortable question: will fighting climate change force us to give up foods we love?


Plonts uses microbes to make anything delicious

In food, only two things matter—taste and cost.26 Without both, a food cannot be globally dominant.

The cheapest and most sustainable foods we make today are plants. We farm corn and soy in such volumes not because we love to eat them but because they are efficient, and efficiency—doing more with less—is the heart of sustainability. A big problem with corn and soy is that they aren’t that tasty. So we use inefficient cows to convert that corn and soy into something we actually want to eat: beef and cheese.

At Plonts, we take a different approach: We take the most sustainable, inexpensive plants available and make them delicious. We do this with microbes.

Fermentation was humanity’s first biotechnology, providing a simple way to transform cheap but not particularly tasty ingredients like milk and wheat into delicacies like cheese and beer. During fermentation—which is distinct from precision fermentation,27 which targets a specific molecular product—microbes digest proteins, carbohydrates, and fats, enhancing the underlying ingredients with three amazing benefits:

  1. Better nutrition by increasing digestibility, providing additional vitamins and minerals, and acting as probiotics
  2. Preserving foods by creating conditions (e.g., high acid) that inhibit spoilage and pathogens
  3. Rich flavors from the small molecules resulting from microbial digestion (e.g., the nuttiness of parmesan)

We are modernizing the discovery of fermented foods. For the last 10,000 years, humans have discovered fermented foods by accident (“Dude, I dare you to eat that”).28 This innovation-by-accident means that the possible delicious pairings between ingredients and microbes have been severely under-researched. Because of the essentially infinite diversity of microbial ecosystems, we focus on finding the microbes that transform whatever ingredient is most sustainable, cheap, and nutritious into something delicious. Our searches may result in familiar foods from unfamiliar ingredients (e.g., cheese made from cheap plants) or entirely new categories of fermented products.


We're starting with 🧀

We aim first to make a cheese from sustainable plants that tastes better and costs less than commodity dairy cheese. We chose to start with cheese for five reasons:

  1. Cheese is the third-most inefficient food behind beef and lamb
  2. Despite advances in plant-based milk (thanks Silk and Oatly!) and beef (thanks Impossible!), plant-based cheese remains unsatisfying29
  3. Cheese is a fermented food, giving us a path to use microbes to make something tastier and cheaper
  4. Seventy percent of the global population is lactose intolerant,30 which means that a delicious nondairy cheese would have a motivated customer base
  5. We love eating cheese

Cheese was one of the first fermented foods. The flavors that people love in cheese—the funk of a camembert or the sharpness of a cheddar—don’t come from milk, which is bland. They come from the metabolism of microbes. Cheese is not a dairy product; it is a microbial product. Thus, our approach is to discover microbial ecosystems that transform plants into the flavors we love in cheese. We are creating a new category of cheese, akin to the creation of cheese from cow or goat milk.

In addition to flavor, texture is the other pillar of taste.31 Instead of matching cheese’s ooey-gooey texture by replicating the precise molecules in cheese—which would be expensive—we aim to replicate those melting and stretching functions by biochemically and physically manipulating inexpensive ingredients. Many molecules from the natural world melt and stretch because of specific molecular structures, and we use these rules and patterns to our advantage.

Meat and dairy are some of the most expensive commodity foods, despite massive subsidies from national governments, which keep them artificially cheap.32 Eating plants is cheaper than feeding those same plants to cows and then eating the cow or its milk. By transforming supremely inexpensive plants—like soy and corn—we can deeply undercut dairy cheese on the cost of ingredients (Figure 3).33

Figure 3. We aim to beat dairy on cost by 2X at scale using inexpensive, sustainable, nutritious plants. Despite large government subsidies for dairy, our approach allows us to beat dairy cheese on ingredient costs, something that cannot be said for cheese made with nuts or engineered proteins.

Although we focus on taste and cost as the drivers of food decisions, they are only proxies for the fundamental purpose of food—nutrition. Our approach provides all the health benefits of fermented foods while avoiding the health concerns of dairy. Making a delicious nondairy cheese would mean that many people wouldn’t have to choose between eating cheesy goodness and feeling like garbage an hour later.

We also love cheese—its depth of flavor, its comforting richness, its diversity across cultures, its stockpiling in a cave by the U.S. government,34 its application in fart jokes.35 Given the calamitous topics discussed here, we might most relish our ability to laugh at ourselves. We’re not saving the world; we just make cheese.


Feed the next trillion with us

By pairing new food ingredients with the right microbes, we accelerate the discovery of delicious, cheap, and nutritious fermented foods and create aged cheese at a fraction of the cost of dairy.

We fed the last 100 billion people using inefficient tools like rocks tied to sticks and animal agriculture. Join us to invent smarter ways to feed the next trillion.



1. How Many People Have Ever Lived on Earth?, Population Reference Bureau
2. Only once freed from looking for food did we have time to invent other stuff, like money, vaccines, the internet, jorts.
3. Thereby depriving us the joy of seeing a giant armadillo. Did humans cause the Quaternary megafauna extinction?, Our World in Data. Late Pleistocene extinctions, Wikipedia
4. And how were we able to turn all that land into farms? With fossil-fuel powered machines. Half of the world’s habitable land is used for agriculture, Our World in Data.
5. Although active release of new carbon into the atmosphere is important—and are the typical numbers used to talk about causes of climate change—what actually warms the planet is simply the total amount of carbon in the atmosphere.
6. Calculated based on fossil fuel usage rate over time and the conversion of each type of fossil fuel to carbon dioxide (Energy Production and Consumption, Our World in Data). For calculations see “All-time fossil fuel usage” in Feeding the next trillion data - 20240419.
7. Food systems are responsible for a third of global anthropogenic GHG emissions, Nature. Emissions due to agriculture, Food and Agriculture Organization of the United Nations.
8. Slash-and-burn Wikipedia.
9. Although numbers of this type and size are inherently noisy, many lines of evidence suggest that these values (carbon emissions from land-use and fossil fuels) are similar in magnitude. Unexpectedly large impact of forest management and grazing on global vegetation biomass, Nature. The carbon opportunity cost of animal-sourced food production on land, Nature. What are the carbon opportunity costs of our food?, Our World in Data. Analysis: Which countries are historically responsible for climate change? Carbon Brief.
10. Estimates of the total land used for agriculture suggest that agricultural land use increased more than 6X from 1700 to 1992 and 10X from 1400 to 1992. A reconstruction of global agricultural areas and land cover for the last millennium, Global Biogeochemical Cycles. Land Use, Our World in Data. The values are partly based on and also follow estimates of human populations, which are enabled by access to food. How has world population growth changed over time?, Our World in Data.
11. Green Revolution, Wikipedia.
12. For calculations see “Fig1 base data” in Feeding the next trillion data - 20240419. In short, we calculated the sum of all greenhouse gas emissions from fossil fuels and land-use change over time. We split the fossil fuel emissions by sector using the data from Emissions by sector: where do greenhouse gases come from?, Our World in Data and Sector by sector: where do global greenhouse gas emissions come from?, Our World in Data. We make the false assumption that emissions prior to 1990 continued in the same fractions as in 1990. Although this is a flawed assumption, we felt it was suitable for illustrating the dominance of the agricultural sector as a source of emissions when you factor in land-use change over human history.
A critical caveat is that timescales matter for carbon cycling. The incredible amount of carbon emissions released from agriculture occurred over 12,000 years, while the emissions from fossil fuels have been released mostly since 1945. Although there are many implications for this difference, it does mean that emissions from land-use change had more time to cycled from the atmosphere back into the rest of the carbon cycle, while emissions from fossil fuels are rapidly accumulating in the atmosphere. But would the climate effects of that carbon be as great if land-use change hadn’t already (1) primed the pump with carbon emissions and (2) dramatically cut down the forests that are one of the fastest-acting and largest carbon sinks?
13. But frankly, not that many people care, and talking about it sure would make this narrative less coherent.
14. Reducing food’s environmental impacts through producers and consumers, Science.
15. Here’s How America Uses Its Land, Bloomberg.
16. Global Methane Assessment: Benefits and Costs of Mitigating Methane Emissions, United Nations Environment Programme. Methane is 28X more potent as a greenhouse gas than carbon dioxide (Understanding Global Warming Potentials, United States Environmental Protection Agency). Fugitive emissions—the leakage of methane from the extraction and transport of fossil fuels, much of which is actual methane (a.k.a., natural gas)—account for 35% of methane, while cows account for 32%.
17. Livestock solutions for climate change, Food and Agriculture Organization of the United Nations. Ref
18. Compare this to chickens, which only spend about 30% of the energy they eat on staying warm and walking around. Feed conversion ratio, Wikipedia.
19. Number of cattle, Our World In Data.
20. The biomass distribution on Earth, Proceedings of the National Academy of Sciences.
21. The global biomass of wild mammals, Proceedings of the National Academy of Sciences.
22. For calculations see “Figure 2 data soy” in Feeding the next trillion data - 20240419. Key assumptions are a global population of 8 billion people, daily protein intake of 50 grams, and based on numbers from US soy production. Similar estimates could be made for fats and carbohydrates, the other macronutrient components of soybeans. Soybeans are a complete protein, so no other protein sources would be required.
23. For calculations see “Figure 2 data cows” in Feeding the next trillion data - 20240419. Global protein need assumptions are the same as for soy. Key assumptions are that land use is equal for dairy and beef cows, thus their share of the total land used in the US for cow pasture and feed splits along that fraction (24.5% of adult cows are dairy cows). We used data from the US dairy and beef industry because it is readily available.
24. Except for rice, which because it needs to grow in a swamp releases a lot of methane.
25. Most of the marginal land unsuitable for agriculture is the result of extractive agriculture or human-caused erosion. If left alone—or actively managed—much of this land would return to forests or grasslands. Climate-smart forestry through innovative wood products and commercial afforestation and reforestation on marginal land, Proceedings of the National Academy of Sciences.
26. Nutrition is also important, but what is considered nutritious changes every four years. Furthermore, for the vast majority of human history, the most delicious things were also the most nutritious, since this is why our taste evolved. Only in the twentieth century did industrial agriculture break the link between taste and health.
27. Precision fermentation is when a specific molecular output is the product. Often, this approach genetically engineers an organism to create a protein or chemical, which then must be laboriously isolated from the bulk fermentation mass. This method is distinct from traditional fermentation, where the entire fermented mass (tankers of beer, blocks of cheese, drums of sauerkraut) is the final product.
28. History of cheese, bread, wine.
29. This opinion is of course subjective and is largely leveled at the mass-market plant-based cheeses that try but fail to replicate cheddar, mozzarella, and parmesan.
30. Review article: lactose intolerance in clinical practice--myths and realities, Alimentary Pharmacology and Therapeutics.
31. Cheese food porn anyone?
32. There are many government programs that support the production of dairy, meat, and the corn and soy used to feed those animals (Dairy Programs of the USDA). We could not find strong analyses of the real market effects of those subsidies.
U.S. Federal and State Subsidies to Agriculture is an analysis by Grey, Clark, Shih and Associates, a consulting group focusing on international trade. They estimate that US dairy subsidies might account for 73% of dairy market returns. Saving the Planet The Market for Sustainable Meat Alternatives (UC Berkeley Sutardja Center for Entrepreneurship & Technology) is an analysis that estimates that the unsubsidized price of ground meat would be $30/lb in 2015 (the commodity price for US ground beef in 2015 was $5/lb).
33. See Feeding the next trillion data - 20240419 for data.
34. Yes, the Government Really Does Stash Billions of Pounds of Cheese in Missouri Caves, Modern Farmer.
35. Bud Light "Cut The Cheese" secret TV spot not on Super Bowl, Youtube.