Tuesday, December 12, 2017

Confused indeed

Comments to the report Grazed and confused from the Food Climate Research Network.

I had high expectations on the report Grazed and confused, developed under the lead of Tara Garnett from the Food Climate Research Network. I have been impressed by her previous research on many aspects of the food system and her capacity to go further than using lifecycle analyses to provide the Truth. Unfortunately this report doesn’t live up to my expectations. At all.

By and large, the conditions differ so much in different parts of the world, that the kind of generalisation this reports tries to make is rather pointless. It is like making a global report for “arable farming” and drawing general conclusions that farming is good or bad.

The introduction of the report is quite promising and balanced, but that balance is unfortunately lost in what follows. The authors state in the introduction that the report will only focus on the effects on the climate and not other aspects, good or bad, of livestock or grazing. Fair enough, but they mostly forget this when they report something slightly positive about grazing animals. That is often dismissed or relativized with other arguments about alternative land use or something else.

In the report one can read statements such as “while well-managed systems that are not implicated in deforestation certainly exist….” giving the impression that deforestation would be the norm. Such statements must be considered extremely biased since most grasslands are not the result of deforestation at least for hundred years.  

Admittedly definitions are difficult, and the report tries to clarify some things reasonably well in the start. But later on, the report is not helping with a clear line, but mixes discussions which are relevant for grass grown in croplands or dramatically altered pastures with native grasslands. They are very different and what is correct for one is not necessarily correct for the other. Grasslands in crop rotations (with grains etc.) can only meaningfully be discussed as part of such a cropping system.

The nature of grasslands differ enormously. By and large, most grasslands are located in places where the natural conditions are harsh. And a lot of the grasslands of the planet are not even grazed by domestic livestock. The alternative use of this grassland is not obvious. A recent assessment found that 2 billion hectares, i.e. less than 60% of the world’s grasslands, are grazed by domesticated livestock. Meanwhile there are certainly examples of very productive and very intensively used grasslands.

While this report has already been used by advocates as a support for that carbon sequestration in grassland is negligible, it is worth noting that the report clearly states that carbon sequestration can be significant, that grazing can increase sequestration and that under certain conditions grazed lands may sequester more carbon than forests.


Now to the more disturbing “facts” in the report.

The methane calculations don't consider the research of the Oxford colleague Myles Allen (2016)[i] which shows very clearly that the calculations for expressing methane in carbon dioxide equivalents hides a lot of information. For short-lived greenhouse gases the comparison with carbon dioxide based on a pulse of emissions gives a reasonably correct result only in a time span of a few decades. In the longer term (which this is all about), the more correct comparison is between a pulse of carbon dioxide and a constant rate in methane emissions. This means that “to achieve a balance between sources and sinks of greenhouse gases in the very long term, net emissions of cumulative pollutants such as CO2 need to be reduced to zero, while emissions of SLCPs [of which methane is one, my comment] simply need to be stabilised.“ (quote from the research of Myles Allen and colleagues).  The authors do acknowledge this in the report, but they don’t seem to consider it in most parts of the report where they just use the extremely simplified ways of expressing methane in carbon equivalents.

This should be combined with

- that pastures are not expanding on the planet, as a matter of fact they have decreased last fifteen years (FAOSTAT).

- the number of grazing cattle have most likely also not increased. 

- the carbon sequestration In grasslands have most likely taken place over hundreds of years and even much longer. For example the prairies was converted from mineral gravel to thick fertile soils over a period of 10 000 years.  A lot of the stable carbon in grasslands is many thousand years old.[ii]

The analysis in the report of the balance between carbon sequestration and methane emissions is based on that there is a (ungrazed) grassland and now let us put a cow that didn't exist before on that grassland. Then we add methane emissions into the calculation and now we start to measure carbon sequestration. But the reality is that the cow was there before and its "pulse" of methane emissions already has happened; the same number of cows over a period of several decades don’t add methane to the atmosphere as equal amounts are released and broken down every year. And the carbon sequestration has been ongoing for hundreds of years.  Those things together makes the calculations in Grazed and confused -- just confused.   

In addition, the discussion on carbon sequestration is based on two other dubious and unsubstantiated assumptions.

A central argument in the report, and the basis for a most of the subsequent calculations, is that over time carbon sequestration will diminish and reach an equilibrium after “perhaps 30-70 years”. It does seems like a plausible assumption that sequestration will diminish over time, but the hard evidence of this is lacking (only one example is cited and this example has not even measured the actual carbon content), and even more so that the decline in the rate of sequestration would be so rapid as claimed. The report also contradicts itself by claiming that observed rates of carbon sequestration could be legacy effects of the lands much earlier conversion from arable to grassland. “Much earlier” must be a lot more than “perhaps 30-70 years”.

The report makes a big issue of the fact that carbon can be lost from the system if exposed to draught, fire or flooding. Of course it can, but this can happen regardless of if there are cows grazing or not. As a matter of fact, grazing animals can very often reduce the incidence of wildfires in dry landscapes. Many (also this report) suggest that grasslands can and should be converted to forests, but forests are certainly exposed to even more such events with fires and storm felling (75 million M3 of trees fell in the storm Gudrun in Sweden 2005). The fact that grasslands today have such great pool of carbon, more or less the same as forests[iii], shows that this objection carries little weight.

The calculations of carbon sequestration in the report are based on several weaknesses.

- Measurements of “carbon” are mostly only in the upper layer of soil, often the top 30 centimetres and sometimes as little as the top 10 cm. Most studies which have supplied the data for the report have been interested in changes in soil organic matter as a measure of soil fertility and not the potential for carbon sequestration. For that purpose the shallow measurements are quite OK. But the stable carbon fractions which can be stored over a long time are found deeper down. Admittedly the carbon content deeper down is lower, but it is a lot more stable. Grassland has a very high proportion of belowground biomass and deep roots, therefore the carbon stored by grasslands is likely to be distributed deeper than in arable land. According to Jackson et al 2017[iv], native grasslands allocate around 60 % of primary production to roots, croplands 10 % and forests 20 %.  

- Measuring carbon content in a specific layer of the soil and using that as a measure of carbon sequestration, omits what happens with the soil in total. Soil can grow “deeper”, by root activity and “higher” by accumulation of litter and dust on the top. This is how “soil” is made in the first place. One can very well sequester a lot of carbon in a soil while the carbon content 0-30 cm remains constant.  

- Measurements are mostly on “carbon” not differentiating between different forms of carbon the soil, while it is well known that there are only some fractions of carbon (humic acids) which are stable.

- The authors claim that carbon sequestration is most likely to take place in degraded soils, but the evidence for this is not at all conclusive. Many of the best soils in the world have been accumulating carbon for a very long term, and many of the poor soils have been poor for a very long time. From a practical farming perspective the experience is rather the opposite. A good soil can go on accumulating carbon for ages, while it is a lot harder, but not impossible, to increase soil organic matter in poor soils in poor climates. 

- Carbon sequestration is often, and by the authors, seen as having a direct relationship to nitrogen (N) availability. For instance, the data in figure 7 is calculated from N values and not from actual measurements of carbon. This assumption of some kind of simple relationship between N and C fluxes is unsubstantiated. For example, in analysing total N and C fluxes over 75 year Sochorova et al (2016) found that C content in soil in unfertilized hayfields was significantly higher than in plots where N fertilizer had been used.[v] In general, while N stimulate growth, it also stimulate above ground growth at the expense of below ground growth, and carbon below ground is much more likely to be stable in the long run. Increased N availability can also stimulate decomposition of carbon rich materials, something everybody making composts can easily witness.


The report deals very lightly with the topic of ruminant/grass interactions and gives the impression that the existence of ruminants in grasslands has little significance compared to non-grazed grasslands. The role of the manure falling on the field is mostly seen as a potential source of pollution. But this is a far too simplistic view of the very complex interactions in nature. For example, dung beetles and mycorrhiza play an important role, “In pastured livestock operations, particularly in the tropics, dung beetles help mitigate greenhouse gas emissions and aid carbon sequestration in part by increasing grass growth, aerating soil, and delivering manure carbon to mineral surfaces (Slade et al. 2016)”[vi].

The figure 10 shows the theoretical relationships between stocking rates, methane emission and carbon sequestration. It shows that with a stocking rate of 0,5 cows long term net sequestration is possible (also with all the limitations discussed above), while it is impossible for higher stocking rates of 1 and 2 animals per hectare. If we divide the global grazing livestock (say of half a billion livestock units, which is probably a big exaggeration) with global grasslands of say 2.6 billion hectares, we get a stocking rate in the range of 0.2. Such low stocking rate would, with the authors own calculation, have the potential to sequester more carbon than the methane emissions. The authors include no such graph but instead shows the result for 0.5, 1 and 2 livestock units per hectare. But 2 livestock units per hectare is a density ten times higher than even a high estimate of average global grazing intensity.

The report claims that most extensive grazing systems are important sources of fossil fuel derived CO2 emissions. There are pastoralist systems that use no fossil fuel derived inputs whatsoever, apart from the odd veterinary medicine. Of course there are others which use fences and motorcycles. But it is hard to imagine that this could have any significance compared the massive amounts of fossil fuel used in agriculture for machinery, pumps, fertilizers, dryers etc.


We live in a world where the whole agriculture system is unsustainable, including the production of staple foods such as wheat and potatoes. By and large the whole system is driven by global markets and competition, massive use of chemical fertilizers, fossil fuels and pesticides, which in turn has created an enormous overproduction of agriculture crops. This unsustainable system is driving deforestation for palm oil, cattle grazing, cocoa, coffee or soybean cultivation. This unsustainable system is driving enormous waste, a rapid increase in chicken and pork consumption and obesity. This unsustainable system leads to massive emissions of greenhouse gases, massive pollution and loss of bio-diversity.

Grazing animals are not the big problem in the food system. As a matter of fact, grazing and other forms of traditional livestock management are endangered forms of agriculture in many parts of the world, simply because they can’t compete with industrial farming practices.  

A sustainable agriculture and food system will look different in different locations; after all, local adaptation is a key characteristic of any sustainable system. Therefore, the number of livestock and the ways they are integrated in the food system (and how much animals products that can be consumed) will differ enormously in the same way as it has done historically. Trying to conclude that pasturing is “bad” or that eating meat is “not sustainable” is pointless on a ´global level.

Both grasslands and forests can play an important role for carbon sequestration. Even arable land can sequester carbon, but currently they are mostly doing the opposite. There are big knowledge gaps about which methods work under which conditions, even if there are many pointers as to which processes are most efficient. There are also most likely trade-offs where carbon sequestration in the soil will be in conflict with carbon for use as food, paper, timber, feed. I wish the Food Climate Research Network would direct its energy into looking into how to improve carbon sequestration while enhancing bio-diversity without reducing the carbon available for use by human kind.  

[i] Allen MR, Fuglestvedt JS, Shine KP, Reisinger A, Pierrehumbert RT, Forster PM 2016:
New use of global warming potentials to compare cumulative and short-lived climate pollutants. Nature Climate Change  doi: 10.1038/nclimate2998

[iii] Harden, Jennifer W. et al. Networking our science to characterize the state, vulnerabilities, and management opportunities of soil organic matter, Glob Change Biol. 2017;1–14.

[iv] Jackson B. Jackson et al, The Ecology of Soil Carbon: Pools, Vulnerabilities, and Biotic and Abiotic Controls Annu. Rev. Ecol. Evol. Syst. 2017. 48:419–45

[v] Sochorova et al,  Long-term agricultural management maximizing hay production can significantly reduce belowground C storage, Agriculture, Ecosystems and Environment 220 (2016) 104–114

[vi] Jackson B. Jackson et al, The Ecology of Soil Carbon: Pools, Vulnerabilities, and Biotic and Abiotic ControlsAnnu. Rev. Ecol. Evol. Syst. 2017. 48:419–45

Thursday, November 23, 2017

Are pigs eating our food?

We are told that increased meat consumption is driving environmental destruction. But what if it is the other way round?

For each person on the planet, about 1500 kcal of plants is used daily to feed animals. In return we get just some 500 kcal per capita per day in the form of animal foodstuffs. The 1000 kcal that is “lost” in this transformation would be enough to feed another 3 billion people. Isn’t it a no-brainer that we rather should eat plants than animals? One could think so. But the reality is a bit more complicated than that. A recent paper from Anne Mottet, Livestock Development Officer, and colleagues from the FAO, Livestock: On our plates or eating at our table?, goes far in giving a better big picture understanding of the role of livestock in our food system.

“I came to realize that people are continually exposed to incorrect information about livestock and the environment that is repeated without being challenged, in particular about livestock feed,” says Anne Mottet. “There is currently no official and complete international database on what livestock eat. This study contributes to fill this gap and to provide peer-reviewed evidence to better inform policy makers and the public.”

The researchers have calculated the total feed intake of domesticated livestock (measured in dry matter) and as can be seen from the pie chart 46% of all feed are grass and leaves. The second most important feed are crop-residues (straw, stover etc.) with 19%, while grains (maize, wheat, barley etc.) account for 13% of the global livestock dry matter intake.  Other by-products from the food chain (in particular from the milling, drink and food industry) supply 5% of the feed, while oil seed cakes contribute another 5%. Notably, oil seed cake is partly a by-product as well. Global consumption of vegetable oil has skyrocketed after WW II and almost all vegetable oils have a waste production in the form of oil seed cake. Large scale cultivation of soybeans was first driven by soy oil demand, and still today soy oil is the second most important oil in the world and contributes one third of the income from soy growing.
Around 30% of the total acreage, 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-production is as energy, industrial raw materials or fertilizers and not food for humans. 

The study found that an average of only 3 kg of cereals are needed to produce 1 kg of meat at global level. But it also shows important differences between production systems and species. For example, because they rely on grazing and forages, cattle need only 0.6 kg of protein from human edible feed to produce 1 kg of protein in milk and meat, which is of higher nutritional quality. For backyard chicken and pigs only 0.1 kg of human edible protein is used to produce 1 kg of human edible protein while for industrial  systems between 2.9 and 5.2 kg of human-edible protein is used per kg of protein. By and large, traditional animal raising systems are an efficient way of producing high value food, which actually is quite self-evident as our ancestors certainly couldn’t afford any large scale waste of valuable food.

For ruminants, around one third of the protein in livestock products comes from pure grazing systems and 62% comes from mixed systems (system where animals graze to some extent, and where the feed has a considerable proportion of roughage) while only 4% of the ruminant protein come from feed lot systems. For poultry the situation is very different, where only 4% are estimated to be kept in backyard systems, i.e. small scale systems where animals are fed scraps, waste or find their own food here and there. One fourth of the pigs are raised in backyard systems and 56% in pure industrial systems. So poultry certainly eat food that people could eat, pigs as well to some extent, but ruminants are hardly guilty on a global scale.

While almost half of all livestock feed comes from grazing, because of low productivity of most grasslands huge areas are used for grazing. The statistics are notoriously unreliable, but the researchers state that the global area of grasslands is 3.5 billion hectares. Only 2 billion hectares, i.e. less than 60%, are grazed by domesticated livestock, however, the remainder is only grazed by wildlife, or not at all. The researchers claim that one third of this could possibly be converted to arable land. In the rest of the land the natural limitations are too big to allow for crop production. When I look at their map, however, I think their assessment is very optimistic. For example the map indicates that most of the southern part of Sweden has grasslands that could be converted to crop fields. While there certainly are such lands, my assessment is that most of what is classified as permanent grasslands in Sweden (not arable land which is grazed) is not convertible to cropping. 

Grassland, top ten countries 1000 hectares
China, mainland 392833
Australia 358962
United States of America 251000
Brazil 196000
Kazakhstan 187465
Saudi Arabia 170000
Mongolia 112364
Argentina 108500
Russian Federation 93000
South Africa 83928

It is important to consider the wider value of animal products in the diet. On a calorific basis they are “wasteful”, less so for protein as we have seen above. 40% of all dietary protein comes from animals (34% from livestock and 6% from fish and seafood). Around 45% of all dietary fat comes from animals and many of those fatty acids are unique for animal fat, and here there is no waste at all in feed conversion as animals get very little fat in their diets. Recent research in the USA, concluded that if the US food system changed into a vegan diet, there would be a huge increase in supply of calories and protein, but shortage in some essential fatty acids, vitamin B12 and vitamin A as well as Calcium.

While the focus of the paper is the role of animals in providing food, the researchers also point out that animals add to agricultural production through manure production and drought power. Further, keeping livestock provides a secure source of income for over 500 million poor people in many in rural areas. In addition, animal products are an important source of fuel and as an industrial raw material.  
We are told that increased meat consumption is driving environmental destruction. But what if it is the other way round? 
It is the constant increase in productivity of the worlds’ farmers, with the assistance of chemical fertilizers, pesticides, fossil fuels, breeding and technology that has created an enormous surplus of agriculture raw materials, in particular wheat, soy, maize, sugar and palm oil. 

There are no additional 3 billion people out there eating all that extra output. The surplus has driven prices down and made it possible to use grain for purposes which earlier were rare, such as bio fuels and animal feed. Under industrial farming conditions it is simply more profitable to use grown plants for feeding pigs and chicken than to have backyard systems. In almost all rich countries the acreage of grasslands are shrinking because it is “cheaper” and “more efficient” to give cows grown fodder and grain, despite the fact that grazing has very low direct cost and very low resource use. Since 2000 the global acreage of grassland has decreased with some 3% according to FAOSTAT.

By and large, the industrialisation of farming is the main driver of the expansion of livestock production. Under Swedish conditions a worker had to work eight times longer sixty years ago to earn the price of a chicken. No wonder global chicken consumption has increased 12 times in the same period!