Funding & Finance

Every Time You Eat, You Spend Land. Here Is How Much.

The Data and Where It Comes From

In 2018, researchers Joseph Poore and Thomas Nemecek published a study in Science that became the most cited reference in food sustainability research. They analyzed data from 570 studies covering roughly 38,700 farms across 119 countries and 40 food products. The goal was to build the first methodologically consistent comparison of food’s environmental footprint across multiple dimensions: greenhouse gas emissions, water use, and land use.

The land use figures in this article come from that study, processed and published by Our World in Data. The numbers measure square feet of agricultural land required to produce one pound of food, counting both grazing land and land used to grow animal feed. Understanding which foods use the most land starts with understanding why the numbers vary so dramatically across the list.

The range is not subtle. Lamb requires 1,806 square feet per pound. Wheat requires 19. That is a 95-fold difference between the top and bottom of the same dataset. Here is what is driving each number.

Lamb and Mutton: 1,806 ft² per lb

Lamb sits at the top of the list by a significant margin, and the reason is straightforward: sheep are ruminants with poor feed conversion efficiency, and they require extensive pasture to reach slaughter weight.

A ruminant animal digests food through a multi-stomach fermentation process that is well suited to grass but deeply inefficient at converting feed into body mass. Sheep need to eat a large amount of forage to produce a relatively small amount of meat. That forage requires land. On top of the grazing land, additional land is often used to grow supplementary feed. The result is a land cost per pound that is higher than any other food on this list.

Lamb also produces less meat per animal than cattle. A sheep yields far less edible weight than a cow, which means the land cost of raising the animal is spread across fewer pounds of final product.

Beef from Beef Herds: 1,593 ft² per lb

Beef from dedicated beef herds sits just below lamb, and for essentially the same reasons. Cattle are ruminants with low feed conversion efficiency, they require large areas of pasture, and the land required to grow their feed adds to the total.

The distinction between beef herd cattle and dairy herd cattle matters here. Beef herd cattle are raised specifically for meat, which means all of the land cost is attributed to that one product. Dairy herd cattle produce both milk and beef, so the land cost is shared across two outputs, which is why dairy beef appears much lower on the list at 211 ft² per pound.

It takes roughly 6 to 8 pounds of feed to produce one pound of beef. Most of that feed is grown on land. And because cattle take 18 to 24 months to reach slaughter weight, that land is occupied for a long time before it yields any food.

Cheese: 429 ft² per lb

Cheese drops sharply from beef, but its number is still significant. The land cost of cheese comes from two places: the land needed to feed the dairy cows producing the milk, and the volume of milk required to produce a pound of cheese.

It takes roughly 10 pounds of milk to make one pound of cheese. Each of those 10 pounds of milk required land to produce. The land cost multiplies through the concentration process. Dairy cows also need to eat, and that feed requires land. Add those two factors together and cheese ends up with a land footprint that is more than four times higher than any non-animal product on this list.

Dark Chocolate: 337 ft² per lb

Dark chocolate at 337 ft² per pound is the number that surprises most people. It is higher than pork, higher than poultry, higher than fish, eggs, milk, and several other animal products. For a plant-derived food, that is counterintuitive.

The explanation is two-part. First, cocoa trees have low yields per hectare compared to most staple crops. They are slow-growing, fussy about climate conditions, and only productive in a narrow tropical band. Getting a pound of cocoa beans out of a farm requires a relatively large amount of land.

Second, and more significantly, a large portion of cocoa’s land footprint comes from deforestation. Cocoa grows best on forested land. In Côte d’Ivoire, which produces roughly 40 percent of the world’s cocoa, the dominant farming model involves clearing forest, planting cocoa, extracting yields for five to ten years until the soil is depleted, and then moving on to clear more forest. Between 2000 and 2014, the land footprint of global cocoa plantations grew by 37 percent. That deforestation cost is embedded in the land use figure for dark chocolate.

Beef from Dairy Herds: 211 ft² per lb

Beef from dairy herd cattle is dramatically more land-efficient than beef from dedicated beef herds. The reason is cost allocation. Dairy cows produce milk as their primary output, and the land used to feed them is primarily attributed to the milk they generate. The beef that comes from culled dairy cows or their male calves is a secondary product, so only a fraction of the total land cost gets assigned to it.

This is one of the more important distinctions in the dataset. Not all beef has the same footprint. Where the animal comes from matters as much as the species.

Coffee: 106 ft² per lb

Coffee at 106 ft² per pound sits roughly midway through the list, and its land cost comes from the same dynamics that affect cocoa: low yields per hectare, slow-growing plants, and a history of expansion through deforestation in key producing regions.

Coffee trees take three to four years to produce their first harvest. They require specific altitude, temperature, and rainfall conditions. Yields per hectare are modest compared to grain crops. And like cocoa, coffee has historically expanded by clearing forest land in Latin America, East Africa, and Southeast Asia rather than by intensifying production on existing farmland.

Pork: 85 ft² per lb

Pork drops significantly from coffee, and the reason is that pigs are monogastric animals, meaning they have a single-chamber stomach like humans. Unlike ruminants, pigs are efficient feed converters. They reach slaughter weight faster, on less feed, occupying less land.

Pigs do not graze pasture. Their land cost comes almost entirely from the grain used to produce their feed, primarily corn and soy. That grain requires cropland, which is why pork is not at the bottom of the list. But the efficiency of conversion means the total footprint is a fraction of lamb or beef.

Tree Nuts: 63 ft² per lb

Tree nuts sit below pork despite being a crop rather than an animal product, which reflects a complexity in how land is counted in this dataset.

Nut trees require land to grow, but they also store carbon and can be productive for decades on the same plot without depleting the soil the way annual crops do. The per-pound land footprint of nuts is influenced by the yield of the specific nut and the density at which trees can be planted. Almonds, for example, have a significant water footprint not captured in this land-use figure. Walnuts and cashews have different land profiles again.

Notably, some analyses of nut production show negative land use change, meaning nut orchards are replacing cropland in some regions and storing carbon in the process. The figure in this dataset represents the median across nut types and farming systems.

Poultry: 60 ft² per lb

Chickens are among the most land-efficient sources of animal protein. Like pigs, they are monogastric and convert feed to body weight more efficiently than ruminants. They reach slaughter weight in six to eight weeks. The land cost comes from the grain in their feed, not from pasture, and the short production cycle means that land turns over quickly.

Poultry also yields a high proportion of edible meat relative to live weight. The combination of feed efficiency, fast growth, and high carcass yield pushes poultry well down the land-use list compared to beef or lamb.

Groundnuts: 45 ft² per lb

Groundnuts, which include peanuts, are legumes that fix nitrogen in the soil as they grow. That means they require less fertilizer than many other crops, and can improve the quality of land used in rotation with other plants. Their land footprint is relatively low because yields per hectare are reasonable and the plant itself is not especially resource-intensive to grow.

Milk: 44 ft² per lb

Milk’s land cost, like dairy beef, benefits from the fact that it is one output of a system that produces multiple products. The land used to feed dairy cows gets allocated across milk, cheese, butter, cream, and eventually beef. When only milk is being counted, its per-pound footprint is far lower than cheese, even though both come from the same animal.

Fish: 41 ft² per lb

Farmed fish sits near the bottom of the list, which reflects the efficiency of aquaculture for certain species. Fish are cold-blooded, which means they use less energy maintaining body temperature and can convert more of their feed into body mass. Some farmed fish also eat feed that includes plant-based inputs, which have lower land costs than the animal products that go into the diets of ruminants or even pigs.

The land cost of fish in this dataset primarily reflects the land used to grow the plant-based components of aquaculture feed. Wild-caught fish have an even lower land footprint in some frameworks, though their sustainability picture involves different considerations around ocean health and stock depletion.

Peas: 36 ft² per lb

Peas, like groundnuts, are legumes that fix nitrogen. They also have high yields relative to the land they occupy, grow quickly, and require minimal inputs compared to most animal products. Peas are increasingly used as a base ingredient in plant-based protein products precisely because their production footprint is low.

Eggs: 31 ft² per lb

Eggs are more land-efficient than milk, poultry, or fish. Hens are efficient feed converters, and an egg represents a small but nutritionally dense output that does not require slaughtering the animal. The land cost comes from the grain fed to laying hens. Battery farming concentrates production and reduces land use per egg, though at welfare costs that this dataset does not capture.

Wheat: 19 ft² per lb

Wheat sits at the bottom of this list. It is a direct crop with no intermediary animal consuming it first. The land is used to grow the wheat itself, yields per hectare are high relative to most foods on this list, and there is no feed conversion inefficiency to multiply the footprint.

The 95-fold gap between lamb and wheat is essentially a measure of how much land is lost in the process of converting plant calories into animal calories. Every step up the food chain multiplies the land requirement. A cow eats grain. The grain requires land. The cow is inefficient at converting that grain into meat. The land cost of the grain gets passed through to every pound of beef, amplified by the inefficiency of the conversion.

Wheat skips that entire process. The land grows the food directly.

What the Gap Means

The foods that use the most land are almost all animal products, with dark chocolate and coffee as the notable plant-based exceptions. The difference between 1,806 and 19 is not a difference in taste or nutrition. It is a difference in how many steps exist between the land and the plate.

Every animal product on this list sits above every plant product except dark chocolate and coffee, both of which carry the specific burden of deforestation embedded in their production history. The closer a food is to its land origin, the less land it requires. The more biological conversion steps it passes through on the way to being eaten, the more land disappears in the process.

For businesses in the food sector, the land constraint is not abstract. Agricultural land is finite. Demand for food is growing. The gap between 1,806 and 19 is where a significant part of the resource competition of the next several decades is going to play out.