EPIBuilding a Sustainable Future
Lester R. Brown, Janet Larsen, and Bernie Fischlowitz-Roberts

Part 1. Assessing the Food Prospect: The Changing Food Economy

The biggest prospective structural changes are in the animal protein sector. During the second half of the last century, a time when population was more than doubling and incomes were nearly tripling, the world demand for animal protein was climbing. For much of this period, growth in animal protein was satisfied by turning to rangelands and oceanic fisheries. Between 1950 and 1990, beef and mutton production more than doubled and the oceanic fish catch expanded fivefold. Since then, growth in the output of these two natural systems has slowed or leveled off as demands have pressed against their sustainable-yield limits. Indeed, in many cases demand has far exceeded sustainable yields, leading to the desertification of rangelands and collapsing fisheries. Today overgrazing and overfishing are the rule, not the exception. 34

Even as these natural systems were approaching their limits, the demand for animal protein was growing at a record rate. As it did so, the world turned to grain-fed beef, pork, poultry, eggs, milk, and farmed fish. Population growth and the strong desire to move up the food chain has driven the demand for meat steadily higher. Indeed, except for the recession year of 1960, the world demand for meat has climbed to a new high every year since 1950. (See Figure 1–3.) Meat consumption per person more than doubled, climbing from 17 kilograms in 1950 to 39 kilograms in 2001. 35

Wherever incomes rise, people try to diversify their diets, reducing their overwhelming dependence on a starchy food staple, such as rice, and augmenting it with meat, eggs, and milk products. This desire to move up the food chain as incomes rise appears to be innate, perhaps a genetic legacy of 4 million years as hunter-gatherers.

Given the rising demand for animal protein in diets, now principally in developing countries, the question is how to satisfy that demand most efficiently. At the first level, the advantage goes to ruminants that can convert roughage into edible forms of animal protein. The roughage may come from rangelands or from crop residues. Once the use of roughage is fully exploited, then the advantage goes to those grain-dependent forms of animal protein that are most efficient. This shift to more grain-efficient, lower-cost, animal protein sources is already under way.

There are some encouraging success stories in efficiently satisfying the hunger for animal protein. In India, for example, milk production has soared over the last few decades, spurred by local dairy cooperatives that provide a marketing link between villagers, who typically have only two to three cows, and consumers in other villages and nearby cities. Milk production in India, which has overtaken that of the United States, the longstanding leader, is based almost entirely on the use of local forage and crop residues. Little grain is fed to cattle in India. 36

China is using a similar approach to expand beef production. In areas that produce grain, particularly those that double-crop grains, such as winter wheat and corn in east-central China, there are large amounts of crop residues—either straw from wheat or rice or the stalks from corn—that can be fed to cattle. Cattle, being ruminants, can easily convert crop residues into protein, leaving the manure to fertilize fields. The amount of beef now produced in this manner in the east-central provinces greatly exceeds that being produced on rangelands in the overgrazed northwest. 37

The pattern of animal protein production worldwide has shifted substantially over the last decade. The growth in beef and mutton production, most of which comes from rangelands, was less than 1 percent a year from 1990 to 2001. Pork grew by nearly 3 percent, poultry production by over 4 percent. But the most rapid growth of all was in aquaculture, which expanded by 10 percent a year. (See Table 1–5.)

The variation in growth rates is explained largely by the efficiency with which various animals convert grain into protein. With cattle in feedlots, it takes roughly 7 kilograms of grain to produce a 1-kilogram gain in live weight. Growth of feedlots is now minimal. For pork, the figure is close to 4 kilograms per kilogram of weight gain, for poultry it is just over 2, and for herbivorous species of farmed fish, such as carp, tilapia, and catfish, it is less than 2. The market is shifting production to the animals that convert grain most efficiently, thus lightening the pressure on soil and water resources. Health concerns are also helping to shift consumption from beef and pork to poultry and fish. 38

Egg production is growing fast, again because laying hens can convert grain into protein rather efficiently. In addition, eggs are a popular source of animal protein in developing-country villages where there is no refrigeration.

Once the potential for relying on ruminants to convert roughage into food products that are edible by humans is fully exploited, then the question is how to satisfy the additional need for high-quality protein. One way of doing this is to convert grain into animal protein at varying degrees of efficiency. Another way is to supplement grain with various beans, peas, and other leguminous crops that contain high-quality protein. This can be seen, for example, in the corn-and-beans diet of Mexico or the wheat-and-lentils combination of northern India. The basic choice is whether to use the land to produce leguminous crops for direct consumption or to produce grain and convert it into animal protein.

Contrary to popular opinion, the latter may sometimes represent a more efficient use of land simply because the yield per hectare of soybeans, lentils, chickpeas, and other leguminous crops is so low compared with grain. In the United States, for example, which produces roughly 40 percent of the world corn and soybean harvests, the ratio of corn yields, at 8.7 tons per hectare, to soybean yields, at 2.7 tons per hectare, is 3.2 to 1. (The United States offers an ideal comparison between corn and soybean yields because they are grown on the same land, often in a two-year rotation.) 39

If land is used to produce corn that is fed to a herbivorous species of fish, such as carp in China or catfish in the United States, which convert 1.5–2 kilograms of grain into 1 kilogram of live weight, or if it is fed to chickens, which use about 2 kilograms of grain to produce a kilogram of live weight, it may yield more high-quality protein than land planted to soybeans for direct consumption, for example, as tofu. In summary, the more grain-efficient forms of animal protein may not require any more land or water resources per unit of protein than legumes. At this point, whether someone consumes tofu, lentils, carp, catfish, or chicken may be less a question of the efficiency of land and water-resource use and more a question of taste. 40

If the alternative is producing beef in feedlots, then the 7-to-1 conversion of grain to live weight of cattle is much less efficient than using land to produce beans for direct consumption. If the option is pork production and the pork is produced with table waste, as it often is in villages in China, the advantage goes to pork. But if pork is produced with grain, as is the case elsewhere and, increasingly, in China, then consuming beans directly would be more efficient. 41

Perhaps the most impressive growth of any animal protein-producing enterprise has been fish farming in China, where a carp-polyculture has been highly successful. Over the last two decades China’s aquacultural output, consisting largely of carp, has expanded from 3 million tons per year to 25 million tons. Indeed, fish-farm output in China is now double the U.S. beef output of 12 million tons. 42

As the growth in animal protein production has shifted over the last decade or so from largely oceanic fisheries and rangelands to primarily pork, poultry, and fish farming, the pressure on croplands has intensified. Expanding protein production by feeding animals, whether fish, poultry, or pigs, means expanding grain use. At the same time, land is required by these enterprises themselves. For example, in China 5 million hectares are now devoted to fish ponds, an area equal to 6 percent of China’s harvested grainland. In the United States, catfish ponds, the dominant source of U.S. farmed fish, occupy nearly 50,000 hectares (190 square miles) of land, most of it concentrated in Mississippi. 43

As animal protein production shifts to more grain-efficient sources, it is automatically shifting to the more water-efficient sources, helping to lower the water deficit. This interaction between the expanding demand for animal protein and the need for greater efficiency in the use of grain and water is reshaping the food economy.


Table 1-5. Annual Growth in World Animal Protein Production, by Source, 1990-2001
Annual Growth
(million tons)
(million tons)


Oceanic Fish Catch1
Aquacultural Output1
1Latest figures available for oceanic fish catch and aquacultural production are for 2000.
Source: Based on FAO, 1948-1985 World Crop and Livestock Statistics (Rome: 1987); FAO, FAOSTATS Statistics Database, updated 28 May 2002; FAO, Yearbook of Fishery Statistics: Capture Production and Aquaculture Production (various years).



Previous Table of Contents Next

34. Beef and mutton from FAO, Crop and Livestock Statistics, op. cit. note 5; FAO, FAOSTAT Statistics Database, op. cit. note 5, with meat production updated 28 May 2002; fish from FAO, Fishery Statistics, op. cit. note 5, and from FAO, Aquaculture Production (various years).

35. Figure 1–3 and data from FAO, FAOSTAT Statistics Database, op. cit. note 5.

36. A. Banerjee, “Dairying Systems in India,” World Animal Review, vol. 79/2 (Rome: FAO, 1994); S. C. Dhall and Meena Dhall, “Dairy Industry—India’s Strength Is in Its Livestock,” Business Line, Internet Edition of Financial Daily from The Hindu group of publications, at <www.indiaserver.com/businessline/1997/11/07/stories/03070311.htm>, 7 November 1997; milk production data from FAO, FAOSTAT Statistics Database, op. cit. note 5, updated 28 May 2002.

37. China’s crop residue production and use from Gao Tengyun, “Treatment and Utilization of Crop Straw and Stover in China,” Livestock Research for Rural Development, February 2000; USDA, ERS, “China’s Beef Economy: Production, Marketing, Consumption, and Foreign Trade,” International Agriculture and Trade Reports: China (Washington, DC: July 1998), p. 28.

38. Conversion ratio for grain to beef based on Allen Baker, Feed Situation and Outlook staff, ERS, USDA, Washington, DC, discussion with author, 27 April 1992; pork conversion data from Leland Southard, Livestock and Poultry Situation and Outlook Staff, ERS, USDA, Washington, DC, discussion with author, 27 April 1992; feed-to-poultry conversion ratio derived from data in Robert V. Bishop et al., The World Poultry Market—Government Intervention and Multilateral Policy Reform (Washington, DC: USDA, 1990); conversion ratio for fish from USDA, ERS, “China’s Aquatic Products Economy: Production, Marketing, Consumption, and Foreign Trade,” International Agriculture and Trade Reports: China (Washington, DC: July 1998), p. 45.

39. USDA, op. cit. note 12.

40. Fish feed requirements from Rosamond L. Naylor et al., “Effect of Aquaculture on World Fish Supplies,” Nature, 29 June 2000, p. 1019; poultry feed requirements from Bishop et al., op. cit. note 38.

41. Beef conversion from Baker, op. cit. note 38; grain to pork conversion from Southard, op. cit. note 38.

42. Aquaculture from FAO, op. cit. note 34; beef from FAO, FAOSTAT Statistics Database, op. cit. note 5, with meat updated 28 May 2002.

43. China’s fish farms from K. J. Rana, “China,” in Review of the State of World Aquaculture, FAO Fisheries Circular No. 886 (Rome: 1997); China’s grain area from USDA, op. cit. note 11; U.S. catfish farms from USDA, ERS, National Agriculture Statistics Service, Catfish Production (Washington, DC: July 2000), p.


Copyright © 2002 Earth Policy Institute