An industrial revolution for agriculture


Depending on what is included in the calculation, agriculture accounts for up to a quarter of the world’s greenhouse gases. Emissions include methane arising from the livestock production process, nitrous oxide from the use of fertilisers and, third, the cutting down of forests. Deforestation, driven by the need to expand the land area devoted to the production of food for humans and for animals, adds to carbon dioxide to the atmosphere.

The world has the rough outline of a realistic plan for cutting fossil fuel use across the global economy. It will increase the amount of renewable electricity produced and switch transport to battery cars and lorries. No such scheme exists for agriculture, even in the vaguest outline. The issue is rarely discussed. But unless emissions from global farming are curtailed, all long-term targets for greenhouse gas reductions are unattainable. 

Take one example. The average Briton eats about 18 kilos of beef a year. The emissions from the production of this single food add about 4% to his or her carbon footprint. And nothing in the government’s plans proposes to reduce this, even though we are increasingly aware that net emissions may need to fall to zero within a few decades to meet a 2 degree temperature change target. It is simply politically impossible to push for a reduction in meat consumption. So the problem is ignored.

The livestock production chain is the most important cause of agricultural emissions. A move to a one hundred percent vegan diet would reduce emissions by 50% or more. Although veganism is growing sharply in some places around the world, the switch to conventional plant-based foods will almost certainly be too slow to provide the speed of reductions required.  And even vegan diets have high carbon footprints and land use requirements.

In this article, I suggest that the only way to achieve substantial greenhouse gas cuts is to move as much agriculture as possible out of fields and into factories. This will directly reduce emissions but also cut greenhouse gases by decreasing the pressure to switch forests to agricultural land.

More specifically, we first need to shift to artificial meat. The need to stop farming beef cattle is particularly urgent; Silicon Valley startup Impossible Foods is a highly plausible contender in the race to create acceptable meat substitutes.

A burger from Impossible Foods

A burger from Impossible Foods

My second suggestion is to grow many products in indoor hydroponic systems rather than soil. This saves land, fertiliser, pesticides and reduces greenhouse gas emissions. Hydroponic techniques for growing leafy crops and some berries are also advancing fast. Fellow Silicon Valley company Plenty Farms is showing how the world might get many of its micronutrients from indoor farming.

A wall of lettuce at Plenty Farms

A wall of lettuce at Plenty Farms

Root crops, such as potatoes and carrots, will stay outdoors for some time. The world’s main sources of human calories – wheat and rice – will also be difficult, but not impossible, to transfer to hydroponic systems. But, despite their importance to human nutrition, these high calorie crops do not occupy much of the world’s usable land area.

We need a new industrial agriculture to reduce emissions and to allow much of the world’s land area to return to forest. Unfortunately, organic agriculture, often seen as a crucial part of reducing emissions, seems unlikely to assist in rapid decarbonisation. It may even increase emissions. As someone who grows and buys organic produce, it saddens me to say this but It is a distraction from the difficult task of feeding up to 10 billion people with the lowest possible carbon footprint.

This is a long article, for which I apologise. I wanted to demonstrate both that we need an industrial revolution in agriculture and that the raw technologies for higher productivity with low carbon impacts are already in place.

The background

Global emissions of greenhouse gases are about 50 billion tonnes a year.[1] Although the figures are considerably more uncertain than for fossil fuel combustion, agriculture and changes in land use contribute about 10-12 billion tonnes of this. The IPCC’s 2014 assessment suggests that agricultural production accounts for slightly over half the total figure, with land use change slightly less.[2] (Land use change in this context is predominantly the conversion of wooded land or wet peatlands to arable or grazing land, a process which results in the emission of CO2 and methane back to the atmosphere).

Agriculture itself directly creates emissions in three main ways. First, some animals, particularly cows and sheep, emit methane from the digestive tract as they break down the complex molecules in the grassy diet. This alone may result in over 2 billion tonnes of equivalent CO2 emissions a year (about 4% of global emissions). Second, animal manure rots down, giving off methane and nitrous oxide. Third, artificial fertiliser applied to fields produces nitrous oxide emissions.

Of the nearly 6 billion tonnes of emissions coming directly from agriculture, perhaps half arises from livestock, or around 3 billion tonnes. Cows are by far the most important source, but the role of pork is increasing fast.

If we add in the impact of deforestation that occurs as a result of increasing meat production, the impact is far greater.

Of the decarbonisation challenges facing the world, this clearly ranks in the top division. In addition, agriculture is the dominant use of fresh water around the world, likely to be an increasingly scarce commodity, and fertilisers and other chemicals applied to fields are important sources of watercourse and coastal zone pollution.

The livestock problem is made worse by the central role livestock production plays in stretching the world’s land resources. About 50% of the world’s habitable land (not glaciated or completely barren) is given over to agriculture. Of this, over three quarters is devoted to livestock and to growing the crops that help feed that livestock. But this land only results in the production of about 17% of the food calorie that humans consume. Land growing crops is - on average – about fifteen times as productive in terms of calories as land given over the animals.

This isn’t an entirely fair comparison because animals are often kept on land that would produce very little grain or other planted crop. But, more realistically, a farmer putting crops instead of pigs into lowland and reasonably fertile field might get five times as many food calories as she did from the animals. Twenty calories of grain fed to a cow will result in about one calorie of usable meat when the cow is slaughtered.

This gross inefficiency is sometimes justified by saying that humans need the proteins provided by meat. This is incorrect; cows and other animals typically eat much more protein, often in foods made from beans, than they actually provide in meat. Mosa Meat, one of the pioneers of artificial meat, says that a cow or pig will transform only 15% of vegetable proteins into edible animal proteins. Animals therefore reduce the net amount of this important food constituent available to humans.

As the now famous Food and Agriculture Organisation (FAO) report said, livestock has a ‘long shadow’. Perhaps surprisingly, relatively few people know this, particularly compared to the increasing numbers aware of the climate impact of travel and energy use. In one survey less than 30% of respondents reported that they believed that meat and dairy production had a major impact on climate change. The figure for transport was twice this.

But not only is the global effect of agriculture as large as all transport emissions, it is far more difficult to reduce. Changes in cultivation practices may marginally reduce the climate-changing effect of cows and sheep. Keeping animals in intensive feed lots probably reduces emissions, but at a cost to welfare that many people regard as too high.

Reducing the land area given over to animals and for the growing of their food would enable arable crops to be grown, at least in some places. The total amount of available food would rise. The world’s extra calorie needs to cope with as many as 3 billion more people in 2050 could be accommodated without further deforestation.

The problem is that as people get more prosperous, they tend to consume more meat. So growing wealth will result in more animals, more land devoted to growing food for those animals and not for humans. Inevitably, the threat to the world’s forests will increase although it is worth pointing out that global deforestation rates have probably been tending to decline for several years. Growing agricultural productivity has keep the land requirements for animal cultivation lower than they otherwise would have been. (One important piece of recent research questions whether forest loss has indeed declined).

Global beef consumption is up about 10% since the turn of the century. We might have expected the number to be higher but increased incomes have generally arisen in countries that do not consume much beef, such as China and India. But rich countries with high beef sales, including the US, might see further growth in consumption. Overall, the US Department of Agriculture sees the average American increasing the amount of meat eaten by 5% between 2015 and 2025, some portion of which will be greater beef purchases.

The second, and very welcome, impact of prosperity is often improved access to high quality foodstuffs that are expensive to produce. Green vegetables, herbs and leafy crops add variety, fibre and important micronutrients to a grain-based diet. Fruits such as berries are attractive to eat and probably good for health. The problem is that these products require far more land for each calorie of food value than the rice or wheat that forms the backbone of most people’s diet. A hectare of spinach, a valuable source of vitamins and metal ions, might produce 5 million calories of food. The same area given over to rice could give six or seven times as much. So as global population expands and people get better off, we can expect more pressure on land use for this reason as well.

What do we do?

1, Replace meat

The most important challenge is to reduce the amount of farmed meat that is consumed. Currently the world uses about 270 million tonnes a year, or just over 30 kilos a person. (These amounts vary enormously, and not necessarily in a way obviously tied to income; the people of Uruguay and Argentine - both middle income countries - eat about 50 kilos of beef alone).

Veganism, or the conscious avoidance of any form of animal product including, is growing strongly in many places around the world. The world leader is probably Israel, with perhaps one in twenty adults saying that they avoid all animal products. About 1% of UK adults now self-identify as vegans - up from not much more than a quarter of this a few years ago - and the percentage is probably double this level in parts of northern Europe. Vegetarianism, its milder alternative, might have gained the affiliation of 10% of Swedes and around 3% of French people, for example.

A rapid worldwide switch to a meat-free diet, preferably with no dairy products either, might be possible but seems very unlikely. Although young adults are restricting their meat intake in richer countries as a deliberate choice, their switch is generally not being matched by the middle-aged and older.

So the world needs to find an acceptable alternative to meat. Plant-based alternatives have historically been poor at copying the texture and full taste of meat. Most vegetarian burgers, for example, may be very acceptable foods to many people but they don’t really mimic minced beef.

We’re left with two main options: trying to improve meat substitutes or making meat in the laboratories. Both routes are being pursued by commercial enterprises, mostly in the US. For what it is worth, I think it is going to be easier and quicker to get good substitutes for meat down to a competitive price than growing similarly inexpensive meat in the lab.

Meat is approximately a trillion dollar industry (c.1% of world GDP) and global capital circles the companies in the artificial segment knowing that the successful businesses will become very valuable entities indeed. And the venture funds putting cash into these companies seem also to be very aware that climate change pressures will be likely to make farmed foods more expensive in decades to come, improving the economics of artificial alternatives.

Meat grown from cultures.

Memphis Meats is one of the leaders in lab-grown meat. Like other companies at the forefront of animal meat replacement, it has attracted investment from well-known investors. Bill Gates holds shares, as does the global agricultural commodity trading firm Cargill. (Gates is an investor in several of the best-known companies in the meat replacement market).

Memphis makes a beef and a chicken meat from animal stem cells cultured from an animal foetus. The company believes it will eventually rely on entirely self-reproducing cells and will not need to extract them from animals. These cells are fed with a cocktail of vitamins, sugars and minerals and over a period of weeks in a bioreactor become meat.

Southern Fried Chicken from Memphis Meats

Southern Fried Chicken from Memphis Meats

The cost is still thousands of dollars per kilo and the company won’t start commercial sales until 2021. Memphis Meats provides figures suggesting that the worldwide average price for meat is about $4 a kilo, and it knows it will have to compete with this figure. In fact, one recent survey I saw showed that consumers actually expect to buy artificial meat at a discount to the farm grown product.

Is it possible for Memphis Meats to get costs down to $4 a kilo? It seems a huge challenge to me, given the length of time it will take growing the meat and the expenditure on nutrients but the company points to the huge energy savings possible from lab-grown meat. It says that one calorie of its beef consumes about three calories of ‘food’ compared to the 20+ that a cow would need to make the same quantity of beef. Memphis Meats also stresses the savings in water and land, saying that its technology may cut 90% from the requirements of conventional agriculture. It puts greenhouse gas reductions at a similar level.

Professor Mark Post set up a lab-grown meat startup after creating the world’s first artificial burger in 2013. Mosa Meat is attempting the same task as Memphis Meats, using some cow cells and encouraging them to replicate in a bioreactor filled with nutrients. It has fallen behind and only says that commercial lab-grown meat might be available within ’10 to 20 years’, not the 2021 promised by Memphis Meats.

Other entrants into the race for artificial meat also lag the Californian company. SuperMeat, an Israeli venture trying to tap into the large vegan population in the country, has struggled to crowd-fund its activities. For the moment, Memphis Meats looks like the early winner. But even it has fallen well behind Winston Churchill’s 1931 prediction.[3]

Fifty years hence, we shall escape the absurdity of growing a whole chicken in order to eat the breast or wing by growing these parts separately under a suitable medium.

Better meat substitutes

The race is to make a burger that tastes like real meat. In other words, the customer can get the environmental advantages of plants with the desirable experience of eating beef. Two companies lead the field. Beyond Meat and Impossible Foods.

Beyond Meat has its products in stores across the US, having finally cracked the wary scepticism of the supermarket buyers early last year. The product is made from pea protein and soya and made to taste either like chicken and beef. The company proudly flags the fact that its products are now in the meat section of the conventional large supermarkets, not tucked away in a section of the store catering for vegetarians. And Beyond Meat sells its product as ‘clean’, ‘healthy’, ‘light’ but full of protein. It is attempting to capture the millennial wish for a food that has the supposed virtues of meat (protein) alongside the health-giving advantages of plant foods and their lower calorific content. (However the product does contain titanium dioxide, a chemical that some people think is potentially carcinogenic in the nano-scale form used by Beyond Meat).

Beyond Meat burger patties in a US store

Beyond Meat burger patties in a US store

The cost is not yet competitive with conventional meats; stores are selling the burgers at three or four times the price of minced beef. The price premium didn’t stop the US meat giant Tyson Foods investing in the company at the end of last year. And a few weeks ago Beyond Meat announced it had added Leo Di Caprio to its list of shareholders. To capture attention from both is an impressive achievement.

Opinions vary as to whether Beyond Meat does really taste like the products it is emulating. There appear to be few such concerns about the burgers made by Impossible Foods. Impossible also makes its products from the protein of plants including wheat and potato. Its key extra ingredient is ‘heme’, an iron compound richly present in meats (and also to a lesser extent in plants). Impossible Foods uses genetically modified yeasts, specially engineered by the addition of a gene from soya beans to express heme. The compound gives the burger a meat-like taste and texture. It also seems to be the reason that an Impossible Burger seems to bleed a red liquid when cooked.

An Impossible Burger with the centre still red from the heme

An Impossible Burger with the centre still red from the heme

At present the Impossible beef patty is sold only to upmarket restaurants. I looked at the menu of one chain and the burger sells for about twice the price of the lowest cost conventional equivalent.  But the difference between the Impossible product and the more expensive parts of the burger menu was not large. In two or three years the cost of the meat alternative will be the same as conventional ground beef, the company claims.

Impossible Foods recently opened a factory in Oakland, California. It’s a big establishment but the world would need over fifty thousand factories of the same size to produce all the meat the globe eats. However it claims that its understanding of the effect of heme on flavour and texture means that it can move from beef to other ersatz meats. And, by 2035 to ‘completely replace animals as a food production technology’, in the brave words of the CEO. You can say things in Silicon Valley, thank goodness, that would be dismissed as ludicrously optimistic in other parts of the world. 

As with Memphis Meats, Bill Gates is a shareholder in the company, which has raised about $200m so far. Its CEO and founder, an idealistic but firmly commercial former biochemistry professor at Stanford, is himself a vegan and says that he started the business explicitly for environmental reasons. Its beef replacement product has a tiny footprint of greenhouse gases, land, water and fertiliser pollution compared to conventional meat.

On the question of the eventual cost of the product, this is what CEO Pat Brown said about the product in an interview in August 2017.

..The economics are very tilted in our favor because the way we produce it is so much more resource efficient. We use a quarter of the water, 1/20th the land, 1/8th the greenhouse gas emissions, way less fertilizer and pesticides and stuff like that. That translates into cheaper production cost. When we look at the technology we have today and project it at scale, there’s a clear trajectory to being able to produce this product and basically all the meats that are in our pipeline at prices that are at or below the cost of the cheapest meats on the market.

How long will this take? ‘Maybe three years or so’. If he is right, then we have a potential way forward to eventually reduce the scale of meat’s impact on the global environment.

I was particularly struck by one of the many other comments the company makes about the effect of switching to meat-free meat. It says that eating just one of its burgers rather than a conventional equivalent will save 75 square feet of land (seven square metres), principally because of the reduced need to grow feed for cattle. This space could, for example, be used for reforestation that will capture carbon. If Impossible Foods is right that one burger saves seven square metres of land for reforestation, then the typical British person switching to their product for all his or her beef consumption would sequester almost a tonne of CO2 a year. This is over ten per cent of that individual’s current footprint. 

2, Grow as much as we can hydroponically and inside buildings or greenhouses

The other move I hope we see is away from field horticulture towards hydroponic techniques. This is particularly useful for leafy vegetables and some berries. Even cucumbers can be grown this way.

Hydroponics avoids the need for soil. Seeds or seedlings are placed with their roots in a channel of water. Plants grow by feeding off the nutrients in the rich broth of water flowing past. Or in some cases the roots of the plant grow directly into air through which a dense mist of nutrient-laden water passes. (This is usually called aeroponics, rather than hydroponics).

Hydroponic techniques can be used indoors or outdoors. If indoors, climate can be more easily controlled. But the ‘farm’ needs to use LED lights to provide the energy the plants need. As LEDs fall in price and gain in efficiency, this is becoming more financially feasible every month. But, it needs to be added, many of the initial hydroponic ventures  havfailed, in part because of high electricity costs for lighting and for cooling but also because they operated at a scale insufficient to cover high fixed costs.

Hydroponics can deliver huge increases in yield per square metre of space. If the plants are stacked vertically in trays, proponents claim a hundred-fold greater output of food. Water use is also dramatically reduced, by up to 99% according to Plenty Farms. Since about 70% of the world’s frash water is used for agricultural purposes, this matters.

Pesticides are either not needed or can be employed in tiny quantities. Weedkillers are unnecessary. Fertiliser consumption can be at least halved. As importantly, very little, if any, fertiliser pollution gets into watercourses. The third most important source of greenhouse gases from agriculture is from the breaking down of ammonia based fertilisers partly to nitrous oxide, a particularly virulent cause of global warming. Hydroponics reduces this source almost to nothing.

Not all plants can be successfully grown in hydroponic systems. But those crops that can be cultivated often have a relatively low yield of calories per hectare out in the field. So they occupy more space than would be needed for high yielding crops such as rice or potatoes of the same food value. More prosperous people not only eat more meat but also prefer to consume larger quantities of green vegetables. As the world’s population grows and average incomes increase, the need for indoor hydroponic cultivation becomes ever more obvious.

The best-funded hydroponic growers also tend to share an ambition to make the production of lettuces and other greens more local. That is, instead of shipping the product from a remote location, perhaps California or Spain, to the main urban markets of the US or Europe, they want to locate the hydroponic farms next to centres of population. The food is much fresher and thus its nutritional content is likely to be better. Incidentally, it also reduces the carbon footprint of the greens or berries because of reduced diesel emissions.

Does a lettuce grown hydroponically taste as good as a fresh lettuce from a local farmer’s market? The prevalent opinion is a confident ‘yes’. The companies trying to take hydroponics to a much larger scale in industrial countries employ scientists who focus exclusively on improving the mix of nutrients in the broth and the spectrum of light directed at the plants. (Most indoor hydroponic growers use a light that appears very pink to human eyes).

Two companies look as though they have solved most of the early problems with large-scale hydroponic growing and say they are ready to roll out their industrial farms to large cities around the world. Plenty Farms uses vertical hydroponic towers which face the light sources; AeroFarms uses an aquaponic technique combined with stacked trays of growing plants.

Plenty Farms

Recently funded with an additional $200m by investors including Jeff Bezos’s foundation and Eric Schmidt of Alphabet, Plenty Farms aims to develop farms on the edges of every major city in the world. The underlying technology it uses is the ZipGrow Tower, a 6 metre high moulded white plastic square tube into which a black plastic spongy material full of air slides. Lettuces and other plants are germinated and grow to small seedlings in a separate area and are then inserted into the black sponge.

A wall of basil with LED lights on the left

A wall of basil with LED lights on the left

After being filled with seedlings, the tower is moved to a vertical position alongside other units and water containing the right nutrients is dripped down through the sponge. The roots of the young plants capture the nutrients and the water itself. In the right conditions, plants will grow perhaps twice as fast as they do out in the fields.

At harvest time the plants such as lettuces are simply cut away from the matrix and transported to where they are to be sold.

Plenty’s first farm is in South San Francisco. It is about 0.4 hectares in size and claims to produce about 900 tonnes a year of lettuces, herbs (particularly basil) and other crops. The intention is to double the size for future urban farms. The average US resident apparently eats about 10 kilos of lettuce a year, meaning that one of Plenty’s new sites will cover the demand from nearly 200,000 people if focuses entirely on this crop. That means the London area might need 50 farms, totalling about 40 hectares, though they would logically be placed right next to the main supermarket distribution centres rather than in the city itself.

Can this new form of farming offer produce at prices that compete with conventionally produced greens? Most existing urban hydroponic farms offer lettuces and other greens that are priced at a multiple of grocery store prices. (Think $5 for a 30 gram box). But Plenty is convinced that parity is possible. Electricity costs may be high but labour and other operating expenses should be lower.

How much land will be saved by each Plenty indoor farm? My rough calculations suggest about 50 hectares. (By the way, this isn’t consistent with the claim that Plenty and Aerofarms increase land productivity by at least a hundred-fold. The company’s own figures suggest an actual figure of around sixty times). Each one is therefore not hugely significant, but many thousands of farms around the world will make a difference.


At the other side of the country in New Jersey, Aerofarms does things a bit differently. But the aims are the same: very dense production of leafy vegetables in unused buildings. Aerofarms sprays a mist of nutrients rounds the roots of plants, stacked in trays between which sit LED lights.

The trays in an Aerofarm indoor growing farm

The trays in an Aerofarm indoor growing farm

It makes similar claims to Plenty about the reduction in water use, fertilisers and pesticides. It says that it avoids 98% of transport emissions from shipping greens across the country. Aerofarms has also dipped heavily into the pools of venture capital looking at agriculture and has raised more than $100m. It has now put farms into other countries, including in the Middle East, and wants 25 within five years. Like Plenty, it recognises the importance of producing its crops at a price no higher than ordinary greens. (Though this still looks very high compared to European prices).

Both technologies use the power of the Internet of things to gather data from cameras and sensors to achieve the best yields and product quality. These are true factory farms.

A final thought

Many people in Europe romanticise farming, particularly if it is of the small-scale family kind. We think that the more agriculture resembles the farming of half a century ago, the more environmentally benign it is. This is wholly wrong. The damage that livestock production and intensive field agriculture is doing to the soil, to watercourses and to the climate is huge, but almost entirely unseen.

As much as possible of our food production needs must be fully industrialised as soon as possible. That means food creation needs to go indoors and agricultural land returned to the wild and to forest. Without this change, the battle against climate change is unwinnable. We already have the outlines of the technologies to make the shift.


[1] This includes carbon dioxide and the other greenhouse gases weighted according to their global warming potential.

[2] IPPC Fifth Assessment report, 2014, p811 et seq.

[3] I saw this quotation on the website of an organisation that funds research into artificial food –

embed Block
Add an embed URL or code. Learn more.