Anyone whose social or working life is spent partly among environmental campaigners and partly among small-scale organic farmers has to adjust to two contrasting diets.
At functions put on by the green movement, you can expect a vegetarian, or even a vegan table, with a high proportion of pulses, grains, nuts, colourful vegetables, olive oil, tofu, margarine and soya milk. Often a significant proportion of the food is imported.
At a farmer's do, on the other hand, you are more likely to be served home-reared beef or lamb, home-cured ham or homemade sausages. The vegetables will be of a more local complexion, and you can expect to find cheese, butter and full-cream milk, rather than their soy-based substitutes.
Aside from any health or animal welfare considerations, the environmentalist's diet is based on the principle that grains and pulses have, on average, a lower environmental impact than meat or dairy products. It is hard to dispute this, and for six years, when I lived in the city, I too was vegetarian, partly for this reason.
The trouble is that if you leave the city and move on to a farm, the vegan or vegetarian diet can start to make less sense, especially if the farm consists mainly of the UK's most abundant crop: grass. It is not unusual for urban vegetarians to revert to meat-eating, as I did, within a year or two of moving to the country.
For some years I have been trying to work out why two groups of people, united in their opposition to industrial matter is a complex one because farming and land management are a complex affair.
There are, however, two simplistic and exaggerated statistical clichés, disseminated by some promoters of vegetarian or vegan diets, which, while useful because they hone in on the two most crucial environmental problems associated with meat, have also muddies the picture considerably.
Ten to One
In 1813, in his Vindication of Natural Diet, Percy Shelley wrote: ‘The quantity of nutritious vegetable matter consumed in fattening the carcase of an ox would afford 10 times the sustenance if gathered immediately from the bosom of the earth’. Shelley’s observation is one of the earliest expressions of a concern that rightly worries all who think about the ethics and economics of what we eat.
Animals convert plant protein and energy into meat protein and energy notoriously inefficiently. This means anyone who consumes large amounts of meat – and that includes most inhabitants of the USA, the UK, and other industrialised countries – may be consuming disproportionate amounts of the world’s available nutrients, in a world where 800 million people do not have enough food to eat.
The feed-to-food conversion ratio of 10:1 has continued to echo down through the years. Peter Singer, in his book Animal Liberation, writes: ‘most estimates conclude that plant food yields about 10 times as much protein per acre as meat does’ – and I have a collection of similar ‘10-to-one’ quotations from a number of writers, including Jeremy Rifkin, Marvin Harris and Patrick Whitefield.
As some other writers have acknowledged, however, the 10:1 figure is an exaggeration for a number of reasons:
• A conversion ratio as ineffi cient as 10: 1 is only ever achieved in meat from ruminants, which mostly means beef cows; this is because their digestive systems are designed to digest high-fi bre materials such as grass, not high-protein feeds such as grains. Pigs, poultry and dairy cows all convert feed into human food at a ratio of between 3:1 and 5:1.
• The feed-to-food ratio ignores other animal products, such as wool, leather, industrial products, nutrient accumulation through manure and animal traction – the last two being a matter of such importance to many farmers in the developing world that meat for them is a mere byproduct.
• It is argued by some nutritionists, and sensed by many meat-eaters, that animal protein (or a proportion in the
diet) is of more nutritional value than pure vegetable protein.
• Organic crops, grown without animal manure, require on average approximately 33 per cent more land (to supply green manure) than crops grown with animal manure.
These factors probably serve to reduce the average feed-to-food conversion ratio of meat and dairy to 3:1 or so. This is still inefficient, but doesn’t account for the most important element of all: a considerable amount of what is or could be fed to livestock can’t be eaten by humans anyway.
How large is this amount? According to University of California food analyst J G Fadel, the processing byproducts of seven industries worldwide (vegetable oil, sugar beet, grain milling, distilling, citrus fruits, almonds and cotton) in 1993 amounted to about a quarter of a billion tonnes of dry matter. If fed to animals it would be sufficient to support the production of 435 million metric tonnes of milk – more than the entire world’s milk supply at the time, or approximately a third of all the nutrients provided by livestock in the world (according to 1997 figures).
Admittedly, some 15 per cent of the above figure comes from the meal left over after making soya bean oil – which is more co-product than byproduct – but Fadel also calculated that the crop residues from wheat, rice, barley, maize and sugar cane (bagasse) totalled a further three-quarters of a billion tonnes of dry matter. Mainly fibrous straws, these are ften not fed to animals when better food is available, but in poorer countries they often are, and worldwide they provide enough energy to support nearly 60 per cent of the world’s milk supply, and enough protein to supply 27 per cent.
On top of that is all the consumer food waste. In the US, according to figures from the US Department of Agriculture, consumers throw away 44 per cent of available food, while in the UK, Lord Haskins estimates that 16 million tonnes of food are thrown out every year from homes, restaurants, shops, schools and the like. Much of this waste used to be given to pigs, and still would be were it not for the EU’s hysterical ban on pigswill in the aftermath of the 2001 outbreak of foot-and-mouth in the UK. Observed conversion ratios for swill into meat suggest feeding this 16 million tonnes to pigs would provide at least 660,000 tonnes of pork, about 45 per cent of all the pork consumed in the UK.
Finally, there is grass. An estimated 25 per cent of the world’s land is classified as rangeland or permanent pasture unsuitable for growing crops. In 1997, when the UN Food and Agriculture Organization (FAO) published a comprehensive analysis of global livestock figures, farming systems based solely on grazing produced approximately 9 per cent of the world’s meat and 8 per cent of the world’s milk. In addition, there is a considerable amount of grazing of non-arable lands on the mixed farms that, at the time, produced 54 per cent of the world’s meat and 92 per cent of its milk.
I have been unable to find an up-to-date figure for the total amount of meat, dairy, leather, pet food, traction and other commodities provided by livestock fed on food we cannot eat or land we cannot cultivate. According to Dutch food analyst Sanderine Nonhebel, writing in 2003, 70 per cent of livestock feed in the Netherlands consists of residues, though in other countries the figure is lower.
In 1979, David and Marcia Pimentel cited figures showing that 60 per cent of the animal protein consumed by humans in meat and dairy products was derived from grasses and forages that can’t be used by humans. In 1997, the FAO estimated that, worldwide, animals consumed 74 million tonnes of human edible protein and provided 54 million tonnes. That’s an overall conversion ratio of 1.4:1 – a far cry from the proverbial 10:1.
Taken together all the above figures suggest that today somewhere between a third and two-thirds of all the world’s meat, dairy products and leather could be provided without using any arable land that could otherwise be used to grow human food crops.
Eighteen per cent
In recent years, the livestock industry has come in for criticism on another score: meat and dairy produce are responsible for a significant proportion of global greenhouse gas emissions. The concerned British consumer can be forgiven for being confused by two apparently contradictory ranges of figures.
On the one hand, campaigners such as Jonathon Porritt and Caroline Lucas have claimed that livestock are responsible for 18 per cent of global greenhouse gas emissions – ‘more than the whole of the transport sector’, according to a campaign leaflet published by Compassion in World Farming and undersigned by Porritt. This slant comes from a 2006 report entitled ‘Livestock’s Long Shadow’, published by the FAO, whose widely publicised press release announced that livestock’s 18 per cent of global greenhouse gas emissions was ‘a bigger share than that of transport’.
On the other hand, almost all other estimates put the figure for livestock emissions far lower. The World Resources Institute assigns just 5.1 per cent of global greenhouse gas emissions to ‘livestock and manure’, whereas it holds transport responsible for 13.8 per cent of emissions, a figure that only covers fuel and does not take into account roadbuilding, vehicle manufacture and other transport infrastructure.
The Government reports that direct emissions from livestock represent just 4.5 per cent of total UK emissions, while Tara Garnett of the Food Climate Research Network estimates them at 6 per cent or so – 8 per cent if we include food imports. These lower estimates are in line with similar figures for other industrialised countries. It is worth examining how, and also perhaps why, the FAO arrives at figures for livestock emissions that are so much higher than anyone else’s. The main way it does this is to include CO2 emissions for Amazon deforestation.
These are a matter of huge concern, but to attribute them solely to the global meat industry distorts the picture in several ways: first, because approximately 99 per cent of meat and dairy products are not rainforest beef; second, because it is debatable to what extent cattle ranching is the driving force behind deforestation; and third, because emissions from deforestation reflect expansion of the industry, not production. If Amazon deforestation were halted tomorrow, we could still be getting beef from areas already cleared, without any further emissions. If you take rainforest beef out of the equation, the FAO’s figure drops from 18 per cent to 13.5 per cent of global emissions, or 12 per cent if you discount Amazon soya as well.
Another fifth of the FAO’s 18 per cent is derived from nitrous oxide emissions emanating from manure when applied to soils – whereas the World Resources Institute assigns these emissions to ‘agricultural soils’, not to livestock. At first sight, the FAO might seem to be correct in attributing these emissions to livestock, but in fact it introduces a distortion. The FAO’s figures only take into account emissions from nitrogen fertiliser applied to animal feed crops, yet include emissions from manure when applied both to animal feed and human food crops.
Since, in the absence of a livestock industry, any manure formerly applied to food crops would have to be replaced by synthetic nitrogen or green manures, with broadly equivalent rates of nitrogen loss in the soil, these emissions are the consequence not of livestock manure, but of our need for food crops. Elsewhere I have described other ways in which the FAO authors puffed up the emissions from livestock, but they do not put forward their elevated figure as a reason for reducing meat-eating. Far from it: they advocate a doubling of meat consumption by 2050. Their line of argument is as follows: The bulk of the emissions from rainforest beef, manure use and other factors that the FAO has introduced to swell the figures can be blamed on extensive farmers – namely peasants, graziers and organic farmers. This enables the FAO to conclude that ‘by far the largest share of emissions come from more extensive systems’. It is then a simple step to conclude ‘intensification and perhaps industrialisation of livestock is the inevitable long-term outcome’.
Includes ‘a relative expansion of concentrate-based production systems, in particular chicken and other poultry’.
This is not a view that the FAO has suddenly come to as a result of its investigation into greenhouse gas emissions; rather its conclusions about these emissions conveniently reflect an ideology that FAO economists have held for many years, and which the Ecologist vigorously opposed in its special issue on the FAO in 1989.
This is Henning Steinfeld, the main author of ‘Livestock’s Long Shadow’, writing in 1998, before the FAO began to treat global warming seriously: ‘We cannot afford the common nostalgic desire to maintain or revive mixed farming systems with closed nutrient and energy cycles... To avoid overuse of immediate natural resources, mixed farmers and pastoral people alike need to substitute them with external inputs’ [meaning fertilisers, feeds and pesticides].
The trend of further intensification and specialisation is inescapable.’ Of course, this trend isn’t inescapable – humanity has the power to make choices – but the authors of ‘Livestock’s Long Shadow’ are no doubt amused that a report advocating factory farming should find favour with so many environmentalists and vegans. And the transport lobby, which the Government states is responsible for 30 per cent of all UK greenhouse gas emissions, must be delighted that so many people are now being persuaded that meat-eating is worse for the environment than driving a car.
Default livestock farming
One paragraph in ‘Livestock’s Long Shadow’ is particularly helpful in pinpointing the difference between traditional livestock husbandry and the intensive stock-rearing the FAO promotes: ‘Livestock are moving from a “default land user strategy” (i.e. as the only way to harness biomass from marginal lands, residues and interstitial areas) to an “active land user strategy” (i.e. competing with other sectors for the establishment of feedcrops, intensive pastures and production units). This process leads to efficiency gains in the use of resources.’
Quite how growing corn on grade-one arable land and feeding it to livestock at a conversion ratio of 3:1, 5:1 or 8:1 can be construed as an ‘efficiency gain’ is a mystery that the authors do not unveil. To those of us whose capacity for logic has not been twisted by excessive enthusiasm for free-market economics, it is abundantly clear that default livestock farming – keeping animals only to graze marginal land and mop up residues, wastes and surpluses – must be more efficient than intensive farming because no human food is fed to animals; and more efficient than stockless farming in that it keeps in the food chain wastes and residues that non meat-eaters can only burn for energy or return to the soil.
How then would a strict default livestock land-use strategy perform in respect of global warming? We have already established that between a third and two-thirds of our current global production of meat, dairy and so on could be provided without using arable land that could otherwise be used to grow human food crops. This suggests, even by the FAO’s own methodology, that the 18 per cent might be roughly halved.
Reduced demand for soya protein (though admittedly not for soya oil) might ease pressure upon the Amazon and make it easier to stop the deforestation. Nitrous oxide emissions would be disproportionately reduced because all emissions related to the production of dedicated animal feed would be eliminated, while – as the authors of ‘Livestock’s Long Shadow’ admit – ‘ineffi ciency is irrelevant in broader terms as long as the animals are totally grass-fed, or raised primarily on crop and foodprocessing residues’.
The only element of the FAO’s 18 per cent that would survive to a significant degree in a default livestock scenario would be methane emissions, mainly from ruminants’ stomachs, which might remain equivalent to perhaps 4 per cent of the world’s greenhouse gas emissions. These could partially be eliminated if we got rid of dairy animals and sheep, and reared only pigs, poultry and fish. This is not an attractive option, however, since ruminants are by far the most efficient way we have of turning grass, straw, leaves and other human-indigestible fibres into human food, and methane emissions are arguably a worthwhile price to pay for this invaluable service.
Many farmers in the developing world are dependent upon grass-fed ruminants for their traction, manure, fuel and milk, while the world’s largest dairy industry, in India, is nourished largely through crop residues, wastes and wayside roughage. Getting rid of these cows in order to reduce livestock methane emissions would result in widespread hunger. On the other hand, getting rid of the US feedlot system, in which beef steers really are fattened on a grain-based diet at a conversion rate of close to 10:1, would be a benefi t to everyone. A return to a default livestock economy might result in something in the order of a 50 per cent reduction in meat and dairy products, and a proportionately larger reduction in greenhouse gas emissions from livestock. Moreover, the majority of livestock reductions would be in industrialised countries, where obesity, not hunger, is the problem, and less in developing nations, where many livestock are already fed on a default basis.
A move to a wholly vegan diet (if it could be introduced and enforced), however, would have severe repercussions on the millions of rural and urban poor who supplement their diets with milk or meat derived from animals fed on wastes, residues and unwanted biomass. I don’t advocate default livestock farming as a strict ideology to which all humanity should conform. There are a number of circumstances in which we might find it necessary or sensible to feed grain to animals – to act as a ‘feed buffer’, for example, a reserve of grain to fall back on in the event of a poor global harvest. But default livestock farming does present a clearly defi ned and easily understandable middle road that steers a course between the rigours of total veganism and the greed, pollution, dependence upon fossil fuels and contempt for animal welfare that characterises intensive, factory farming.
Simon Fairlie is editor of the The Land magazine
Everything we eat has an ecological footprint, but few food products have been subjected to high-quality life cycle analyses that quantify all the energy inputs from farm to fork. Many of the analyses that do exist are inadequate, failing to take account of things like logistics, packaging, refrigeration, cooking and processing. When all these factors are taken into account, the results can challenge some close-held assumptions about the balance between omnivorous and vegetarian diets.
In 2002, when Swedish researchers took a comprehensive look at the energy inputs of 150 different foods, meat – especially from ruminants – had predictably large inputs. But vegetarian food flown in by plane, vegetables that are deep-frozen and those grown in heated greenhouses were all less environmentally friendly than locally produced organic meat.
The study found that grains, pasta and fresh, minimally processed fruit and vegetables generally required the lowest energy inputs per kg. Surprisingly, the most energy-intensive foods included some mainstays of vegetarian eating, and consisted of (in approximate descending order):
1. Shrimp, without shells
2. Tropical fruits, fresh, flown in
3. Cod, fresh, cooked
4. Salmon, farmed
5. Beef, fresh and frozen
6. Tomatoes, grown in greenhouses
7. Cheese, all types
8. French fries, single portion
9. Milk powder
10. Tuna, canned
12. Lamb, chicken and pork, most types
14. Baked cereals
15. Apples,commercially dried
16. Mackerel, fresh, cooked
17. Herbs and spices, commercially dried
19. Potatoes, baked
20. Strawberries, fresh, flown in from Middle East
This article first appeared in the Ecologist October 2008