Guest Post by Geoff Russell. Geoff is a mathematician and computer programmer and is a member of Animal Liberation SA. His recently published book is CSIRO Perfidy. His previous article on BNC was: Feeding the billions on a hotter planet (Part II)
Welcome to Part III of my still presumptuously titled series on feeding the world in 2050.
I spent the first two parts of this series looking at global authorities like the FAO (United Nations Food and Agriculture Organisation) with its predictive obsession and its policy associate IFPRI (International Food Policy Research Institute) with its meat obsession. Writing in a similarly obsessed country with far more cattle than people, I felt compelled to add a special section on protein and to also quantify the place of meat, particularly sheep and cattle meat, on the world food stage. Cattle are a major player in climate change, biodiversity loss and general environmental destruction but both they and sheep are globally irrelevant to food security. But worse than being irrelevant, their net contribution may well be negative. Here are some of the negative impacts:
- Reductions in the productivity of the land that produces real food. These reductions are via physical soil damage, consumption of crop residues which protect the soil, the deliberate burning of areas that are croppable to maintain them as pasture.
- Fouling water. Lack of clean water is the second biggest cause of malnutrition.
- Acting as disease generators. I mentioned Cryptosporidium in the last post, but livestock are also major generators of novel rotavirus strains. Rotavirus kills a million children annually, with vaccination not always available in the developing world. We don’t need new strains.
- The direct sickening and killing of children and women via the use of animal dung as a fuel.
- The reduction in the global food supply by making feed production more profitable than food production. The last impact doesn’t always apply to sheep and cattle but is more general. People with the perspicacity to easily recognise this problem in the context of biofuels are almost universally blind to its existence elsewhere.
Today, in the last of the series, I want to look some standout scientific work that breaks the predictive meat obsessed mould. This is work by Jonathan Foley and Navin Ramankutty and a sizeable group of associated researchers. I’ll call this the “FR” work, but keep in mind that there are many other researchers involved.
This work breaks the mold because it isn’t concerned with mere prediction, like that of the FAO. Nor is it obsessed with meat as a food but rather it recognises meat’s central role in reducing global food Calories.
Step 1 and 2 and 3 … prepare the data
Central to the 2011 Foley-Ramankutty paper is the careful selection and preparation of global data sets over more than a decade. This combines large scale vision with obsessive attention to detail. Here are the major data sets that serve as input into their global modelling:
|Agricultural lands||Combined production and satellite data on 175 crops||I II|
|Cultivable land||Merging of satellite and GCM prediction data||I|
|Crop planting dates||Merging of regional data sets||I|
|Fertiliser and manure production and application||Merging of data sets||I|
|Nitrogen flows on croplands||Global dataset||I|
|Phosphorus balances||Global dataset||I|
|Monthly climate conditions||1961-90 average at 1km resolution||I|
|Soil attributes||soil carbon, wilting point, nitrogen, bulk density, thermal capacity||I|
I don’t know the extent to with these datasets are the product of independent effort or orchestration, but FR have assembled them to work as a unit together with assorted algorithms.
The first two of these datasets are particularly important. National governments collect data on agricultural output for various regions and then combine. Most people work with the combined data. But FR work as much as possible with the low level regional data. Australia, for statistical purposes, has 59 regions and Brazil has 5510! For some countries, the data is thin or suspect and FR use FAO national figures which are derived using all kinds of estimation techniques and expert knowledge. The FR team then combine this raw (or not so raw) data with two sets of satellite data. Satellites can identify different types of woodland, grassland and so on, but are not good at differentiating between different crops, you need the low level regional data for that. Combining the two types of data allows you to allocate production to geographical regions at a reasonable level of detail.
The resulting dataset allows you to answer many questions. FAO has national data of how much of which crops are used as feed, so if you combine this with knowledge of which crops and animal products are produced in different areas, then you can, for example, calculate the food Calories produced in those areas. You can combine this with climate records and predictions from Global Climate models to predict how much food might be produced with the same or different crops or animals in the same region. Knowing planting times and crop suitabilities for different areas allows you to calculate food production possibilities under any alternative agriculturally feasible scenario.
Next step … use it
So far, so good. The FAO and IFPRI modellers do similar things and indeed, FAO data is central to the FR work. But it’s what FR do with the data that separates them from the rest. They don’t stop with “mere” prediction and they don’t let prevailing meat industry biases block the science.
FR suggestions on ways to increase the global food supply begin with a basic constraint: agricultural expansion onto new land, particularly tropical forests, must be slowed and eventually cease. With a few important exceptions, tropical crop yields are below those in temperate regions, but carbon released during their deforestation is huge. This cessation of deforestation, particularly tropical deforestation, is a climate change imperative.
Within this already radical constraint, FR’s first two suggestions are straight from the “more of the same” play book of FAO and everybody else. These suggestions are to reduce the gaps that exist between the yields achieved by the best farmers and the worst and to increase resource use efficiency. Even in areas with similar soils, rainfall and climate, some farmers get far larger yields than others. There are many reasons for this, and current agricultural research and extension programs are designed to address such problems. While recognising and quantifying the room for improvement, the fact that we have doubled the global population without heavily increased rates of malnutrition is an indication that these parts of the current food supply system are working.
Plant scientists have achieved a myriad of small useful advances in the past few decades in addition to the larger and more heralded successes of the green revolution. They are also constantly battling plant diseases by deriving new resistant strains and treatment. Without those on-going efforts the global food supply would crash. Animal scientists on the other hand, have achieved increased output by pushing livestock well past physiological limits so that, for example, all but a few percent of the world’s 18 billion chickens at any point in time will be unable to walk normally in the final weeks of their 6 or 7 week life.
The next suggestion in the FR paper goes beyond business as usual and is the bombshell … shifting diets away from animal source foods. As we saw last time, IFPRI explicitly rejected any analysis of such an option for reasons that were simply rubbish. Similarly, the lead article in the special issue of “Science” on food security last year also rejected analysing the potential for such a shift, again using unsubstantiated claims. In the previous post I showed that the claim that grasslands produce significant amounts of meat is false under any reasonable definition of “significant”. The same lead article goes on to claim that: “… pigs and poultry are often fed on human food ‘waste’.” . This is also false under any reasonable definition of “often” and FR quantify the details.
One can reasonably argue that scientists should stay out of policy debates, but refusing to calculate or estimate the impact of a fairly plausible policy direction is not only entering the debate but taking a strong stand and showing a clear bias. It pre-empts policy decisions by failing to provide policy makers with proper data. This is unconscionable, sloppy and unprofessional.
Foley and Ramankutty grasp the nettle and calculate the impacts of a full shift to plant based foods. They estimate that this would add 49 percent to the world’s Caloric output. They naturally leave open the vexed issue of how big a shift in this direction is socially and politically possible. But their calculations allow policy makers to consider the issue. Not only do they calculate how many extra Calories we could produce but the article contains a map showing exactly where in the world these extra Calories could be generated.
And in a changing climate …?
The extensive datasets and modelling enable FR to calculate changes in potential food production under various climate scenarios. Early work on this was done in a 2002 paper and can be repeated as improvements are made in both climate models, auxiliary methodology and the datasets. In the 2002 paper, 3 climate models were used and in regions where the models were in substantial agreement, an index was derived to predict the change in suitability of the land for cropping.
Here is that early result. It shows large regions of increased suitability for crops in 2070-99 as well as regions of decreased suitability.
That 2002 study was aimed at mapping cultivable land … cropland. Unsurprisingly, there are vast croppable areas available in tropical regions, including Africa, but much of it is forested or grazed.
As the future warms and more northern regions become cropable, there is a disturbing potential for the north-south food and wealth gap to grow. What will the northern countries do? Will they clear and crop or clear and graze newly productive land? Will they clear and crop and feedlot and refuse to sell cheap food but ship frozen meat? Will some dark forested areas become more reflective under wheat and will this increased albedo compensate for the loss of carbon? There are many questions.
Foley and Ramankutty have quantified what is qualitatively obvious. The explosion of global food production over the past few decades would have wiped out malnutrition if had been used to feed people instead of livestock.
But malnutrition is still with us. It has changed its face a little with a billion overfed people facing chronic diseases being added to a billion underfed people facing the more pressing daily diseases of poverty. As the wealthy in developing countries adopt the diets of the dominant western culture they contract the same diseases and demand the health care facilities that go with the diet. You can get a triple heart bypass in Cairo or Nairobi despite 30 percent of the children in Egypt and Kenya being stunted.
The same policy path which FR has shown has the potential to tackle the biggest component of malnutrition, not enough food, will also enable both a cessation of deforestation and an increase in reforestation. James Hansen’s Target atmospheric CO2: Where should humanity aim? regards both rolling back 200 years of deforestation and slashing non-CO2 climate forcings as both essential to get to a stable level of atmospheric forcings. Changing to a plant based diet tackles both problems. It isn’t a substitue for rebuilding our energy infrastructure, but it is both essential and complementary. A 2009 paper also showed huge financial benefits in fighting climate change using dietary change.
We know we need to change the global food system for a myriad of reasons and the work of Foley, Ramankutty and their colleagues has quantified the evidence and so strengthened the case.
Any move to reduce livestock and increase plant foods will be resisted by powerful and well connected forces. Marion Nestle in Food Politics and T. Colin Campbell in The China Study described the subversion of the nutritional advice system in the US by the livestock industries. They did it as well connected Professors inside that system.
In Australia, the CSIRO has pushed a high meat diet in a bestselling book owned by 1 in 7 Australian households, effectively thumbing its nose at its climate scientists and sending a very effective message to the Australian public that climate change isn’t worth doing anything about. Similarly, the recently released NHMRC Draft Dietary Guidelines have done no more than pay lipservice to environmental issues in general and climate change in particular. We don’t have a whistle blowing insider (yet), but we have all the external public evidence of the subversion of the system. Elsewhere, the signs of progress away from meat obsession are reasonable in the UK and parts of Europe. India is holding the line and Chinese livestock growth has at least plateaued.
It will be an uphill battle.
And from left field …
A long crippled theory which frequently gets a run in discussions on food security received another hammering recently. Fish and other seafood are globally irrelevant to food security. Together they provide about 1 percent of Calories and mostly to the rich with some 63 percent of the ocean’s fish going to high income countries having just 18 percent of the world’s population. Fish can of course be locally important in developing countries like Indonesia and we rich countries know exactly what to do when people who need food try to take it from people who don’t … we burn their bloody boats and put them in jail.
The dodgy theory that drives the demand for fish and fish oil in rich countries is that we need to eat them for their long chain polyunsaturated fatty acids (LCPUFAs), because our brains can’t make them efficiently.
The existence of millions of children with normal brains dispite no LCPUFAs in their bottled infant formula for decades (it is still only “optional”) should have been enough to kill this myth, but it keeps popping up. However recent work in the Maasai in Kenya has done the hard work of measuring everything carefully (easier said than done) and showing that people can indeed make ample LCPUFAs, far more than theory predicts. The Maasai have almost no sources of LCPUFAS in their diets, but have perfectly normal levels in their red blood cells.
The study also contains some interesting and highly relevant observations about the value of cattle to the Maasai. The cattle have little food value. They rarely eat them, but cattle are traded for maize. The exchange rate is said to work out at about 1000 Calories of beef for 8000 Calories of maize. The argument that cattle are important for food security because of their economic value is very different from the argument that they are important nutritionally. It is one of a myriad of economic and social problems associated with formulating plans for dietary change.