Monday, May 3, 2021

We can't all live on grass-fed beef - but we can eat more of it

Those that claim that grazing animals can only make a marginal contribution to human nutrition are mistaken. Those that claim that we can feed the global population with grass-fed beef are equally wrong.

It certainly not possible to sustain the global population on a diet that is dominated by products from grazed animals, but grazing ruminants make a very significant contribution to human nutrition.

It is not likely that current total meat consumption (i.e. including also poultry and pork) could be sustained just from grazed animals.

It is likely possible to substantially increase global production of meat and milk from grazing animals to reach a quantity similar to the current consumption of beef and milk.


Those of you that follow my writings have probably noted that I question the opinion that veganism is superior from an environmental and agronomic perspective. I will not repeat what I have written in many articles (here and here are two examples). Here, I will, however, partly turn my attention to a much smaller, but still very vocal group of “carnivores” who see a meat-heavy or total animal diet as the best possible, for the body and for the human environment. This is often coupled to the promotion of regenerative grazing. I will not enter into the discussion of the health benefits of various diets (my own humble opinion is that humans seem to be a very adaptable species when it comes to diet, ranging from largely vegetarian diets to animal heavy diets).

I think it is quite obvious that livestock can be raised in a very sustainable way, of which well adapted grazing is one of the best examples. It is equally obvious that today livestock is often not raised in an environmentally sustainable way, not to speak about the ethical aspects of livestock rearing. Instead my focus here is on the possibility (regardless if it is desirable or not) of supplying a human population of 8 billion (which will be reached in a few years) with livestock products from (regenerative) grazing on a scale that would allow a diet dominated by meat and milk.

Let’s first establish some baselines. The human body needs some 2100 calories of energy per day as well as some 50-100 g of protein and some 50-80 g of fat according to most nutritional guidelines. I will here stick to the lower numbers for protein and fat, recognizing that the actual protein consumption in many countries is way higher. Animal sourced food contributes around 18% of the calories, 40% of the protein and 45% of the fat in the global human diet. Notably, the production of meat from poultry and pork is almost three times higher than the production of meat from the grazing animals. Average per capita consumption is 11 kg of cattle, buffalo, camel, sheep and goat meat together, 16 kg of pork, 15 kg of chicken and 90 liter of milk.

Because of the dominance of cattle (including buffalo) among the grazers I will focus on beef as the meat source and milk from cows. Whenever relevant, my calculations will include other animals that are predominantly grazing. The dietary composition of beef carcass varies much depending on race, age, sex and feed. Here I use 18 % protein and 10% fat and 1600 kcal per kg of carcass. Retail cuts will have considerably lower fat content as quite a lot of the tallow will not be present in the cuts and protein content will be higher as most of the bones and some of the fat are not included in the cuts. From a nutritional perspective I think it is best to keep to the carcass composition assuming that all the fat and more of the bones are actually integrated in the food chain. Carcass weight doesn’t include offal, head, and other parts of the body of which some could be used as food. On the other hand, in rich countries consumers eat only part of the carcass weight (in Sweden estimated to only 60%), so it seems like a reasonable limitation. The composition of milk also varies, but not as much as it does from meat. Here I use 1 liter of milk with 4.5 % fat and 3.5% protein and 700 kcal of energy.

Comparing the protein and energy values of beef and milk with human needs:



Energy (kcal)

Annual total need  per person

18 kg


Quantity needed for total supply: Beef

100 kg

470 kg

Quantity needed for total supply: Milk

514 kg

1094 kg


As can be noticed, there is a huge gap between the possibility for beef to supply us with protein and the ability to sustain us with energy. Globally, milk is a more important protein source than beef and milk is even more important as a source of energy and fat. This needs to be kept in mind, as the debate often is stuck with meat.

Many of the grazing animals are not grazing anymore or get only part of their ration from grass. According to the FAO GLEAM database pure grassland systems produce around one third of the combined protein production from cattle. Most production is in mixed systems where cattle graze to some extent and get additional feed. Even though not well reflected in this classification I would venture that there is a substantial production of milk which is quite similar to the feedlot system, where cows never graze and get a high share grains and protein feed. Nevertheless, grazing is estimated to contribute 55 % of ruminant feed  according to Mottet et al 2017.

If we look into the grasslands and meat production of the ten countries having most grasslands in the world according to FAO statistics we see that the average production of meat per hectare of grasslands is between 2 and 60 kg per hectare and year. But, in many of these countries feedlots, with farmed crops provide a large share of the meat produced. I would say that among the countries listed only Mongolia, Kazakhstan and Saudi Arabia have a production that is totally dominated by grazing. Feed lots, on the other hand, are often linked to cow-calf operations where mother cows are kept on grass and their offspring is brought to feed lots at the age of 6-9 months. This means that also in the USA, the homeland of the feed lots, grazing still plays an important role. 

Grasslands and production of beef, mutton and camel 2018

million hectare

million kg cw


China, mainland








United States of America












Saudi Arabia












Russian Federation




These ten countries have almost half of the grasslands of the world and they produce around half of the total quantity of meat from the species that are adapted to grazing. The average production is 21 kg of meat per hectare. According to estimates around 55% of the total feed of ruminants originates from grasslands.* If we use the FAO figure of 3.6 billion hectares and multiply that with 12 kg per hectare we will get some 5,5 kg (cw) of beef per hectare. Making a similar exercise for milk, but a lower share of grazing we will get some 40 liters of milk from grasslands. This adds up to just around 2,4 kg protein per year, or 6-7 grams per person and day, or 12%-14% of protein needs. This is a valuable contribution to the protein supply, but very far from making it possible to “feed” humans with meat and milk from the world’s grasslands. Counted on an energy basis (calories) livestock products from grasslands contribute only 5% of the calories needed – no wonder that Mongolian herders complement their animal heavy diet with potatoes and wheat.

Even with (heroic) assumptions of double productivity with improved grazing methods, such as AMP or holistic grazing, we are very far from “feeding the world”. However, such improvements would make it possible to produce beef and milk on current consumption levels from grasslands only. While I am a fan of regenerative grazing, I believe there are far too many open questions for us to assume that a doubling of productivity is possible, but I also don’t think there is a basis for ruling it out.

Notably, figures on global grasslands are very variable. A recent article in Nature by Jinfeng Chang and colleagues estimates that there is almost 5 billion hectares of grasslands in the world, but that only 1.6 billion hectares are grazed by domestic animals (i.e. only half of the area in the calculations above, but presumably the grasslands used are those most productive). In addition, the same study estimates that the grazing intensity, expressed as how big share of the above ground net primary production that is actually consumed by livestock is around 16 %, indicating a potential for increased production even without other productivity increasing management interventions. In Europe, Russia and USA utilization of grasslands have shrunk considerably the last two decades.

We should, of course, keep space for wild life in grasslands as well, so even if we could increase productivity and increase area of utilized grassland we can’t assume that we can use all that productivity for our domestic grazers. Even though the grazing of sheep or cattle in many cases has led to the extermination of wildlife (both other grazers and even more so predators), there are many examples of relatively peaceful co-existence between grazing animals and wild life. In our own small farm, deer, hare, foxes, voles, badgers, mice, flocks of geese and other birds, the occasional moose, lynx and many species of predatory birds also feed on our grasslands, not to speak about all the insects. Increased stocks of wild animals could also make a significant contribution to human diet as well (this depends mostly on how much ecological space grant wolves, cheetahs and lions).

But what if we also used arable land for grazing?  Most grasslands of the world have very low productivity. In part it is because of management but mostly it is an effect of a climate that is too dry or to cold and sometimes because of very poor soils. Arable land is mostly developed where the conditions are much better and where one can expect yields which are much, much higher than from permanent grasslands. In addition, arable land is often irrigated and fertilized. Let us further assume that we would take half of the current arable land and use it for grazing. Let us also assume that we can produce 250 kg of beef (cw) per hectare on those 0.75 billion hectares land. It is certainly possible in many cases to produce even more, but also there are many places where such a yield would be out of reach, especially if we envision a production without chemical fertilizers and no import of nutrients via feed. In such a scenario, we would get another 24 kg of beef per person contributing one fourth of annual protein need. If the production was oriented to milk its contribution to human nutrition could be even higher. There would be little land available for growing pig or chicken feed in this scenario, so the total contribution of livestock to human nutrition would not increase compared to the current situation. In some situations, grazing animals on arable lands may not infringe on the production of crops as animals can graze stubble, cover crops etc. To integrate grazing livestock in many crop farming systems would have many other benefits as witnessed by many practitioners.

The overall conclusion is that it is not possible to “feed the world” with a diet dominated by meat and milk from grazed animals. However, meat and milk from grazed animals make a substantial contribution to human nutrition and it is very likely that it could increase substantially, through better grassland management and increased grazing in crop farming systems, without jeopardizing human food supply. Having said that, it is essential to understand that these kinds of global averages are not a good guide for local food systems, neither for production not consumption. Food systems should be adapted to the local conditions and the possible role of ruminants (and other livestock) varies and should vary depending on the local conditions.


Finally, my apologies to chicken and pigs. They can also graze and they should be allowed to roam outside. There are many examples of pigs that are fed almost entirely on grazing and eating nuts and roots. Pigs are also particularly well adapted to the conversion of left-overs and both chicken and pigs can convert waste to valuable food. They have a role to play in most agro-food-systems apart from pure pastoralism, and under some conditions they are even more important than the ruminants. This is also well reflected in traditional diets.



*The nutritional quality of grass is often lower than many of the other feed stuffs, so the estimate is probably giving a bit too high weight for the contribution of grasses, in particular in milk production, which is more dependent on high quality feed.


Wednesday, March 17, 2021

Don't throw the cow out with the feedlot

While there are many environmental and ethical reasons to criticize industrial forms of livestock production, there are no valid reasons to shun the rearing of livestock in general. One can of course discuss how large populations of various animals should be as well as how a responsible and sustainable production and consumption can look like. But an intelligent such discussion must be based on a proper analysis of the strength and shortcomings of livestock keeping, and must be context specific as the agricultural conditions vary enormously. What is sustainable in one place may be a disaster in another place.

The two most frequent misconceptions about livestock in general and and ruminants (cattle, sheep, goats etc,) in particular are their climate impact and their land use. I have discussed the climate impact of livestock many times. An article on Our World in Data made me write to them and point out significant errors about land-use by livestock:


To Our World in data

Comments on the article, ”If the world adopteda plant-based diet we would reduce global agricultural land use from 4 to 1billion hectares” by Hannah Ritchie, March 04, 2021

I believe the article is very misleading for a number of reasons:  

1.      In the text and graphic there is some distinction made between arable land and grasslands. But overall most readers will understand “agriculture land” as croplands.

2.      The figure used for “pasture”, 2,8 billion hectares, exaggerates the actual land use by domestic livestock. Jinfeng Chang et al 2021 Climate warming from managed grasslands cancels the cooling effect of carbon sinks in sparsely grazed and natural grasslands estimate the area grazed grasslands to 1,65 billion hectares.

3.      The average grazing intensity of the managed grasslands was estimated by Chang et al (2021) to just 16%, meaning the there is plenty of space for other species. Many of the managed grasslands are important habitats for other (endangered) species and grasslands provide many important ecosystem functions. This is also a very important difference between croplands and grasslands. To describe grasslands as being primarily production areas for livestock is comparable to describing coastal waters as fishing grounds, disregarding all other important functions of those areas.

4.      The graph of global land use claims that 43% of the global crop land is used to produce animal feed. This figure is misleading. I believe the best estimate is found in Mottet et al 2017 Livestock: On our plates or eating at our table? A new analysis of the feed/food debate. Some 210 million hectares of cereal production is used for livestock feed, in addition 66 million hectares are used for the cultivation of hay, silage and other fodder crops. Together that makes 275 million hectares of land used for crops dedicated to feed production, slightly less than 20% of all arable land. If you add the proportion of feed coming from other crops (e.g. the straw or chaff from grain grown for food or oil seed cakes) one could consider that another 260 million hectares are used for feed production. But in the context of “people could eat what is eaten by animals” it makes little sense to add these as the alternative use of these residues and by-product is as energy, industrial raw materials or fertilizers and not food for humans. It would also be appropriate to indicate the important nutritional contribution of animal sourced foods, which contribute around 32% of all proteins and 37% of all fat in the global food systems.  

5.      The graph also fails to describe that approximately 10 percent of global cropland is used for bio-energy, fiber and other industrial uses.

6.      Apart from the factual shortcomings in the figures, the statement that “80% of all agricultural land is used for meat and dairy production.” disregards the many other roles of livestock. Many hundred million heads of cattle, buffaloes and camels are used for carrying loads and plowing the land, increasing the ability of poor people to grow food. Livestock manure is used as fuel and construction material. Wool, leather, many different kinds of hair are very important materials found in everything from world cup footballs to cashmere scarfs. Livestock also provide banking services and has cultural and religious values.

7.      The claim that land use could be reduced by 75 % in a vegan scenario is not properly demonstrated by Poore and Nemecek (or anyone else).

a.      The proposed diet (supplementary table 14) is not realistic. Suggested consumption of pulses is for example much higher than the current consumption in almost any country. The huge consumption of nuts, tofu and soy milk is a purely theoretical construct, also not taking into consideration allergy, culture or affordability of such foods.

b.      The socio-economic impact of making even more people dependent on oil crops and soy produced in a few regions in the world and the loss of livelihood caused by this is not analyzed at all.

c.      Climate, soils, pests, nutrients and markets determine to a large extent what farmers can and will produce. Crop rotations are required to control many disease.  There is no analysis of whether the global agriculture system actually can produce the proposed diet.

d.      The diet has many foods which have co-products which currently are used for animal feed. Currently land use is allocated to the animals eating them as in the case of rape seed cakes, wheat bran, soy meal, dregs and many other products (see point 4). But if there are no animals the whole land use must be allocated to human food. The most striking example is that in the proposed diet there is calculated with an increase in soybean oil consumption. This means that soy production must expand even further and no land is saved for soy production.

e.      There is not a proper analysis of if the proposed vegan diet will be nutritionally satisfactory.

8.   There are, on the contrary, several research articles showing that land use will be less with a certain (rather small) proportion of animal foods. See for example Zanten, Hannah et al 2018, Defining a land boundary for sustainable livestock consumption and  Peters, Christian J. et al 2016, Carrying capacity of US agricultural land, ten diet scenarios. The reason for this is the efficient use of by-products by animals. 

Finally, it should be noted that the scenario by Poore and Nemecek 2018 includes the abandonment of all pastureland, which means the prohibition of pastoralism as a livelihood. Pastoralism constitute the livelihood  and identity of hundreds of million people and the land has almost no alternative food producing use. A scenario without livestock would functionally amount to genocide. It is irresponsible of Our World in Data to promote such a scenario. 


There are obviously many more things to say about land use and livestock. A frequent criticism is that land is cleared in the rainforest for grazing cattle. While one can discuss exactly how much cattle is to blame for deforestation, it is a fact that global grassland area has not expanded at all since 1860: the expansion of grasslands in some parts of the world is offset by reductions in other parts. 

Lifecycle analysis, the method used for the calculation by Our World in Data and their main source, is fairly useful for analysis of industrial production but largely unfit for analyzing the complexities of food and agriculture. Recent research in Europe demonstrate, for instance, that while ruminants according to life cycle analyses  use much more land than chicken, a circular food system in Europe can have a much higher share of ruminants and pigs than of chicken, because chicken mostly eat farmed crops. 

From a bigger picture perspective it is probably a bigger ecological shortcoming  that there are not enough herbivores grazing in the grasslands of the world. Preferably a lot of them should be wild ones (deer, bison, buffaloes etc.) as there are far too few of them left. The interaction between wildlife and domestic livestock is complex and undoubtedly there are cases where they are in direct competition, but often they are complementary, so there need to be no conflict between them (the question of predators is even more complex and complicated). In my own farm, our few cattle live in a peaceful co-existance with flocks of geese, roedeer, wild boars, foxes, hares, a lynx and the odd elk, not to mention the legions of much smaller herbivores (grasshoppers and alike). There are a much more conflicts between wildlife and our cultivation of vegetables and apples trees.

The notion that the best for the climate and biodiversity would be to plant grasslands with trees is in general unfounded. For sure there are such instances, but most grasslands are grasslands because the conditions are not conducive to trees. There is also no reason to make such a strict distinction between grasslands and forests, there are many biomes in between a pure grassland and a dense forest. A wooded savanna or a European mosaic landscape is as natural as a steppe, a prairie, the taiga or the rainforest. 

P.S. I have got no reply from Our World in Data.....