Diet and Population

Food production needs to roughly double by 2050. There is a spread of opinion between a 75% and a 150% increase. To clarify, this means food resources need to double, not calories consumed. The difference is important, because it is our resources that are constrained, and different diets produce calories with varying efficiency. A 30% increase in population from 2010 to 2050, and the dramatic increase in meat consumption, seen when families income moves through the US$500 to US$5000 bracket, gets you comfortably to a doubling. This assumes global population will reach 8.9 billion by 2050, from 6.9 billion today and peak at 9.2 billion in 2075, taking double the time to add the next 3 billion increment as it took to add the last.

The second factor, changing eating habits, is far more significant than population growth. For example Americans eat 125 kg of meat and 20 kg of fish a year and therefore use 1.5 hectares of agricultural land to feed each person, despite the highest grain yields per hectare in the world. The domestic animals in the US eat 7 times the amount of grain as the human population, which gives an indication of their land and water costs. In India, including the 50 million hectares of ‘forest land’ used for grazing, the area of land for food production is 0.2 hectares per capita and set to decrease to 0.13 with population growth by 2050. While land and its uses are far from homogenous, this gives just a feel for how an increase in meat consumption can have a multiplying effect on agricultural land requirements. It makes a doubling of the resources for food production, rather than the total calories, a reasonable assumption.

Nowhere has this shift in income and diet been more clearly illustrated than in China. China has taken its per capita meat eating from 3.6 kg in 1960 to 52 kg in 2000, doubling in the last decade. There has been a similar pattern with fish from 2.5kg to 21kg over the same period . Meat consumption and production was growing more than 10 times faster than population and correlates more closely to economic growth. Half the world’s pork is now consumed in China. In India meat consumption is now at just 4.5kg per capita. Cultural habits have put forecasts for 2050 at 11kg per capita and substantial growth in milk sales is expected, but this needs to be combined with the rapid population growth in that region. China is expected to slow to about 70 kg per capita by 2050. Despite these lower growth forecasts compared with 1960-2000, the meat consumption reflects a multiple effect on grain consumption and the land needed to grow it. US citizens therefore indirectly consume a ton of grain compared to 178 kilos per Indian. Hence the starting point of this report, which instigates a hunt for where to find this doubling of food resources.

Some reports talk about the doubling being achieved on even less land than we currently use, and further downward pressure on agricultural prices. Other’s talk about Armageddon, starvation and war on a global scale. In this report we will not address the extremely important issue of global soil degradation and its impact on available hectarage and yields. We assume, very optimistically that, globally, soil quality will remain constant. Views are simply too diverse, and data too limited, to make an estimate, But while some regions have seen agricultural soils improve in the last 50 years, no articles can be found that suggest globally we have seen an improvement. Likewise very optimistically we assume agricultural output is unaffected by climate change, that the benefits in new areas from more benign conditions equals out the areas where volatile weather and increasing temperatures reduce yields. This report focuses on yields and available land for agriculture.

Green Revolution

Before we look into how this doubling will be achieved, we need to look at how we managed to do it last time in 1960 to 2000, the so called ‘Green Revolution’. World population doubled from 3 billion to 6 billion. The population growth rate peaked in the 1960s at over 2% and declined to 1.35% by 2000. There was a very strong increase in food commodity supply. Global food production per person increased more than 20% and in developing nations by more than 50% per person. Global consumption of meat, on a per person basis, more than doubled. The global food resources, taking into account population, calories, and the meatier diet, were effectively tripled in 4 decades, an astonishing change.

In 2000 food commodities sold for less than 35% of what they did in 1960, taking into account inflation. This is part of a much longer term real price decline for food evident in an index constructed for wheat that shows a 90% decline from 1800 to 2000. Some 80% of the extra food came from improvements in the intensity and productivity of agriculture, and only 20% came from the development of new land.

1.5 billion hectares of land is now used for arable and permanent crops, around 11 percent of the world’s surface area. 3.5 billion hectares is pasture for cattle. An increase of approximately 180 million hectares of cropland over the period 1960-2000, and 370 million hectares for animal husbandry. This has been done while farmland in the US and EU dropped by 10% and Eastern European land use saw drastic reductions, by approximately 25%. So although there was a net increase of 550 million hectares there was actually a 630 million hectare increase in the developing world with the chief regions being Brazil, Sub-saharan Africa, Indonesia, and Malaysia.

Many reports, in doing land use projections, stick to the same yield improvements per annum. In such a scenario the increased agricultural land would only have to supply 20% of the increase so we would ‘only’ need an additional 130 million hectares of cropland by 2050 and another 300 million hectares of pasture. This is a total of 430 million hectares for additional agricultural use. To put this in perspective the USA has a total of 155 million hectares of cropland (plus 16 million hectares fallow) and 234 million hectares of pasture. The agricultural land currently represents 46% of the USA’s entire landmass. So in this scenario, in 2050, we need to add a bit more than the entire farmlands of the USA in a scenario where 20% of the increase in production continues to come from additional agricultural land rather than yields.

While East Asia and the Indian subcontinent may be near their agricultural land capacity, theoretically there is another 50 million hectares in the US and Western Europe, that can come back into production, plus an additional 50 million hectares in the old USSR. After this you are looking at forest clearance, bringing wilderness under the plough. Argentina’s agricultural land is almost fully developed. What we have been seeing there in the past 2 decades is more pastureland being converted to crop growing. Brazil has a land mass of 850 million hectares with 250 million in agricultural use, 185 million of that for pasture. The most famous area for booming new agriculture has been the Cerrados. The Cerrados is one of Earth’s last remaining arable frontiers for the expansion of agriculture. Of its 200 million hectares there is still 80 million hectares available for further expansion. Brazil has almost as much farmland with more than 975mm of rain each year as the whole of Africa and more than a quarter of all such land in the world. According to the UN’s World Water Assessment Report of 2009, Brazil has more than 8,000 billion cubic kilometres of renewable water each year, easily more than any other country. Brazil alone (population: 190m) has as much renewable water as the whole of Asia (population: 4 billion). In the Amazon Basin targets set by the Brazilian government should limit the addition of cattle land to 400,000 hectares per year, so 16 million hectares by 2050. So, if we fully develop the Cerrado, gradually encroach on the Amazon, and bring all fallow land in the US and Europe into production we get to 200 million hectares. Perhaps we can piece together the other 230 million hectares from Africa and Indonesia?

It’s at this point that we need to be reminded that on top of this 430 million hectares we would still need to gain, through yield improvements the equivalent of another 500 million hectares of cropland and 1.2 billion hectares of pasture land. An area twice the size of China, not just its farmland, the entire country itself!

So before we go any further, let’s look in more detail at how realistic this assumption is, of avoiding the need for this additional 1.7 billion hectares of agricultural land. The assumption of a a repeat of the last 60 years ratio of 20% contribution coming from new agricultural land in the next 40 years appears somewhat arbitrary. Even if we can search the earth and piece it together with great difficulty we need to be confident the other 80% can be achieved with yields. So lets analyse the prospects for yields more carefully.

The Miracle of the US grain belt

During the ‘Green Revolution’ of 1960-2000 wheat use has grown faster than population. Since 1961 growing 2.5% per annum while population was growing at 1.7%. Looking further back, over the past 70 years the US grain belt has also increased corn yields 5 fold. Taking yields from 1.4 tons per hectare to over 8 tons. What has brought these increases about?

There are two main reasons for higher yields of corn in the past 50 years which reflect generally on all grain crops. Firstly, the introduction of hybrids with narrower denser root systems that extract high volumes of water to support the increased biomass generated by more efficient photosynthesis angles in the leaves of these hybrids. These improvements doubled yield, having a particularly dramatic effect on irrigated land or regions with excellent climatic conditions for the water demands of these new varieties. Yet yields have continued to grow with irrigated land yields going from 3.6T to 9T, and yields without irrigation rising from 1.5T to 4T. The additional density was also achieved by creating hybrids that responded to artificial fertilizer. This resulted in another doubling in yield. Pesticides were less impressive with estimates of impacts nearer to 10%. Other improvements are generally attributed to improved tillage practices, improved agricultural technology working with economies of scale.

So can we find another doubling of yield in the next 50 years? There appears to be no silver bullet on the horizon and some environmental factors are hampering progress further.

Corn yield growth rates peaked at an annual-average rate of 3%–5% in the 1960s, 124.5 kg increases per hectare per year. But have steadily declined to a relative rate of 0.78%  or 49.2 kg per hectare per year during the 1990s. A general inverse relationship between increasing corn yield and decreasing yield growth rates was noted after yields reached 4 Tons per hectare, suggesting that widespread, significant increases in corn yield are not likely to take place in the future, particularly on irrigated land, without a second agricultural revolution.

Nearly half of the value of all crops sold comes from the 16 percent of harvested cropland that is irrigated. Perhaps further irrigation could bring in the tonnage? As it brings in an average of 3 tons per hectare in additional yield, with a maximum difference between irrigation and rain fed of 6.4T in dry regions but as little as 1.3T in wet areas.

The Basin and Range aquifers which are the main source of irrigation in the Southwestern States are under some pressure, but nothing to panic about. Only about 5%, or a few centimetres of total rainfall, is recharged back into the aquifer system a year. There have been meaningful attempts to manage these two important resources. Most notably in 1993 a 540 kilometre canal was completed that runs from the Colorado River to just south of Tucson. This canal took twenty years to build and was built to recharge and slow the uses of water getting pulled from the Basin and Range aquifers in Arizona.

The High Plains Aquifer, is a vast yet shallow underground aquifier located beneath the Great Plains. One of the world’s largest aquifers, it covers an area of approximately 450,000 km² in portions of the eight states of South Dakota, Nebraska, Wyoming, Colorado, Kansas, Oklahoma, New Mexico and Texas. About 27 percent of the irrigated land in the United States overlies this aquifer system, which yields about 30 percent of the nation’s ground water used for irrigation. In addition, the aquifer system provides drinking water to 82 percent of the people who live within the aquifer boundary.

It was only in the 1940’s that affordable technology became available to substantially extract water. This transformed the High Plains into one of the most agriculturally productive regions in the world. During the early years, this source of water was thought to be inexhaustible, and its hydrology a mystery. But, because the rate of extraction exceeds the rate of recharge, water level elevations are decreasing. The water table has been drained in some places, such as Northern Texas. The USGS estimated that total water storage was about 3,608 km³ in 2005. This is a decline of about 312 km³, or 9% since substantial ground-water irrigation development began, in the 1950s.^[

The situation is by no means a disaster waiting to happen but, importantly, there is not much room for a significant increase in irrigation. 1 ton of grain requires roughly 1000 tons of water.

Increased fertilizer usage isn’t a solution. Yields peak at an optimum spreading of kilos per hectare. Additional fertilizer can actually reduce yields. Fertilizer usage in the US has been relatively static in the past 20 years.

Looking at the much heralded and demonized GM crops, 90% of GM crops in use today relate to pest resistance and herbicide tolerance and the benefits in yield are argued back and forth between 6% – 11%. No meaningful increases in biomass, only reduced pest infestation and reduced labour for weeding.

China and India 1960-2010

So let’s look at developing countries where yields per hectare are lower than the US, and assess the viability of yields there rising to US/Argentina levels. This represents a potential 60% increase in the yields in China and India’s grain belts if they could achieve the same outputs as US grain producers per hectare.

China and India represent 44% of the mouths to feed on this planet. India has 20% of the global population living on 4% of the world’s land. With no room to add additional agricultural land, additional volume has to come from yields. With wheat, India is currently managing rates of 5 tons per hectare in Punjab and pockets of Uttar Pradesh. The spectacular improvements in yields in India to-date have occurred for all the same reasons as in the US. The difference is that irrigation represents 60% of cropland in India compared with 16% in the USA. Tubewell use in India increased by more than 100-fold between 1960 and 1985, on the back of a policy of free electricity for farmers. For example, in 1950 India had an irrigation potential of 22.6 million hectares. By 1993 this had grown to 86 million hectares. Irrigation can double yields on one hybrid seeded crop, but in areas of India used to one poor crop a year this allowed them 3 planting seasons a year, so annual yields increased from 100% to as much as 400%. This was the Green Revolution in India, free on-tap ground water combined with new hybrids that took full advantage of this plentiful water supply.

However while we have seen some minor impacts on US aquifiers, the situation in India is much, much more drastic. In the Punjab, India’s equivalent of the US Grain Belt, the past two decades have seen groundwater levels dropping at 25-30 cm per year. At groundwater depths below 15 meters, the commonly used tubewells will not function, and a well must be abandoned. The percentage of land where the water table is below 10 meters has increased from 3 percent to 46 percent between 1973 and 1994. This overuse of groundwater threatens the future of the area and the national goal of food security.

15% of agriculture and 80% of urban water is from depleting ground water supplies. The World Bank, I believe rather overdramatically, forecasts India’s ground water will be used up by 2050. But it does point to there being no sustainable room for increased harvests. Even India’s  current production appears to be a global disaster waiting to happen within the next half century.

Without the billions of tons of water, no amount of fertilizer, GM, or hybridization can build biomass.

Let’s take a look at China. China has 22% of world population and 7% of the fresh water. Falling water tables are also already adversely affecting harvests in China, the world’s largest grain producer. A groundwater survey released in Beijing in 2001 revealed that the water table under the North China Plain, which produces over half of that country’s wheat and a third of its corn, is falling faster than earlier reported. Overpumping has largely depleted the shallow aquifer, forcing well drillers to turn to the region’s deep fossil aquifer, which is not replenishable. The survey, conducted by the Geological Environmental Monitoring Institute (GEMI) in Beijing, reported that under Hebei Province in the heart of the North China Plain, the average level of the deep aquifer was dropping nearly 3 meters per year. Around some cities in the province, it was falling twice as fast.

A World Bank report notes that there is evidence of deep wells around Beijing now having to reach 1,000 meters to tap fresh water, adding dramatically to the cost of supply. In unusually strong language for a Bank report, it foresees “catastrophic consequences for future generations” unless water use and supply can quickly be brought back into balance. Wheat farmers in some areas are now pumping from a depth of 300 meters. Pumping water from this far down raises pumping costs so high that farmers are often forced to abandon irrigation and return to less productive dryland farming.

After peaking at 123 million tons in 1997, the wheat harvest has fallen in five of the last eight years, coming in at 95 million tons in 2005, a drop of 23 percent. The recent decline in rice production is partly a result of water shortages. After peaking at 140 million tons in 1997, the harvest dropped in four of the following eight years, falling to an estimated 127 million tons in 2005. Only corn, China’s third major grain, has thus far avoided a decline.

Overall, China’s grain production has fallen from its historical peak of 392 million tons in 1998 to an estimated 358 million tons in 2005. For perspective, this drop of 34 million tons exceeds the annual Canadian wheat harvest. China largely covered the drop-off in production by drawing down its once vast stocks, taking stocks down from 121 million tons in 1998 to 38 million tons in 2006. While stocks have now climbed to 60 million tons China began importing grain in 2004.

China, like India, has wheat yields averaging 5 tons in its northern grain belt. But the shortage of fresh water limits any ability to push this higher. China is also the largest global user of fertilizer at 40 million tons, double that of the USA. I see no clear road ahead for significant increased yields in China.

 

 

Looking forward to 2050

Looking through the regions first separately and setting aside surpluses for regions with deficits. The US population is expected to grow to 450 million, a 40% increase. Currently the US is exporting 10% of its agricultural produce and has 50 million hectares up its sleeve. It should be able to feed itself but will nolonger be a net exporter.

India’s population is expected to reach 1.7 Billion a 500 million increase. Combine this with Pakistan and Bangladesh and you have 2.15 Billion, a 770 million increase. The problem with the research reports in India from the provincial waters departments is they look at all the net water available in India (available monsoon run off and ground water) and distribute this evenly to India’s farming requirements in 2025 and 2050. Their views are summarized in their Water Report and is quoted here.

“The average annual precipitation over the country is estimated at 4000 BCM of which a part goes towards increasing ground water storage, a part is lost as evapo-transpiration and the remaining appears as surface water. The water resources potential of the country which occurs as natural run off in the rivers is estimated at about 1869 BCM., considering both surface and ground water as one system. Due to various constraints of topography, uneven distribution of resource over space and time, and geography only about 1122 BCM of the total potential can be put to beneficial use, 690 BCM through surface water resources and 432 BCM by ground water, total 1122 BCM. Estimated usage in 2000 was 634 BCM, estimates for 2025 and 2050 are 1093BCM and 1447BCM respectively.”

They appear unaware that soils and therefore fertility vary greatly across the country. The US Grain belt that produces over 50% of all the grain in the country is in an area that is 15% of the total country land mass, which has optimum soil and water. It is quite similar in India. This belt in the northern strip of the IndoGangetic Plain is about 15% of the country and produces 50% of the country’s grain. The belt runs through Uttar Pradesh, Bihar, Haryana, and the Punjab. Punjab and Haryana’s groundwater problems are undeniable. Uttar Pradesh and Bihar’s water problems are disputed. Fresh data needs to be gathered. Alfisols, good fertile soil also stretches down the east coast, a secondary belt that supplies about 20% of the grain. West Bengal, Andra Pradesh, Tamil Nadu, and Orissa. Water problems here are well documented.

The Agricultural Departments calmly say that India needs another 500 Billion cubic metres of irrigation water per year by 2030, nearly double the current national level. To put this in perspective the capacity of all India’s dams is 200 Billion cubic metres, but they actually hold only 78 Billion cubic meters after the monsoon and 14 Billion cubic meters at the end of the dry season.

Reports indicate the 4 states in the primary grain belt have 120 Billion cubic meters of replenishable ground water. The 5 states in the secondary grain belt have 90 Billion cubic meters. 70% of India’s additional grain needs to come from these two areas which is an additional 350 Billion cubic meters of water. This extra requirement is more than double the current TOTAL replenishable groundwater of 175 billion cubic meters. Will the Indian subcontinent be able to feed itself? The answer appears to be No, No, and No. Only by using non replenishing groundwater, can they stave off the need to become a massive importer. Where is the food going to come from? India cannot sustain its current agricultural outputs and yields, let alone increase it based on available water and farm land.

In China, while the population will only increase 12%, we can expect some further increases in meat and calorie consumption. Current Chinese agricultural output is likewise restricted by fresh water limitations and we can expect it to contract slightly by 10% from its current levels. China, atleast has a strategy, the transfer of water from South to North (88 Billion cubic metres per year) and buying land and influence in Africa and South America. With drastic water rationing, significant food imports, and large farming interests abroad, China may muddle through. Unlike India which faces a simply hopeless task.

In Africa the situation is very different.  Yields have been trailing the rest of the world significantly. Sub- Saharan Africa should, from a hectare, soil and rainfall perspective be able to feed even a population double its size if it could double yields to still modest international standards, This requires no technology, hybrids, or fertilizers that don’t already exist today. However If we look at the Arab nations separately, with forecast population also expected to double to 645 million, they will be forced to continue to rely on more than 90% of their food coming from imports, chiefly from sub-Saharan Africa. Africa’s yield problems are socio-political and as such it is anyone’s guess if they can become efficient farmers. Going on past history and China and the Middle East’s strategic ambitions it is likely that famine and malnutrition will remain a part of Africa’s future.

The big question is can Brazil and Europe feed the Indian subcontinent and China with its surplus? Brazil’s emergence as a major player in food exports is well documented and set to continue over the coming 2 decades. Europe currently exports as much food stuffs as it imports. I believe Europe, in twenty years will be a significant player in food exports. Food prices will feed back into agricultural land prices but there are other issues relating to food pricing that have become more relevant since I first wrote this in 2007.

Prices and social stability

This story will not unfold in one year or one decade. We will have good global harvests one year which will ease concern and briefly dampen prices, but the trend is for prices of grain to return to 1960 levels (triple) in the coming 15 years. In 2007 – 2011 we already are half way there. We can then expect prices to continue to increase in the 15 years to follow that. How this plays out politically will be very interesting, looking at Tunisia and other social unrest during recent price spikes. Where 50% of your income is spent on food, tripling of food prices has untenable consequences. Without being over melodramatic, previous collapses of regional civilizations and empires did not occur due to an inability to feed themselves. They collapsed before that, when price escalations led to civil strife, which in turn led to the collapse of distribution networks and war. For the last 50 years we have lived with astonishingly cheap food, which has meant, in Europe, that working class families spend just 12% of their income on food. This period seems to be coming to an end and the consequences for the 80% of the world’s population who spend over half their income on food will be hard to predict.