By Fred Pearce

Government chief scientist Sir John Beddington calls it 'the perfect storm'. Soaring world population, coupled with climate change, is set to create a world food crisis and leave billions starving.

'We are at a unique moment in history,' he said recently, while launching a report from his Government think-tank, Foresight.

The Foresight project, Global Food And Farming Futures, says only a revolution in the way the world grows its food can save us. Clearly, David Cameron's top boffin wants to kick-start that revolution.

The world's population will reach seven billion this year and may peak at nine billion by mid-century. There are plenty of things wrong with the world's food system. But the amount of food it produces isn't one of them.

We already grow enough food to nourish nine billion people, probably 15billion people, in fact, for we eat only about one third of those crops.

Much of the global harvest feeds livestock  -  an inefficient route for delivering our nutrition, since it takes eight calories of grain to produce one calorie of meat.

Plenty more is diverted to make biofuels. An African could live for a year on the corn needed to fill one gas-guzzling SUV fuel tank with ethanol.

That's not all. In the developing world, an estimated 30 per cent of the harvest is eaten by rats and insects, or rots in grain silos. We in the First World are better at preventing losses, but then we throw about 25 per cent our food away, uneaten.

The truth is that the world's farmers could probably double the amount of food they grow  -  using GM crops and other technologies  -  and still people would go hungry. This is ultimately not about production or about human numbers, it is about poverty.

Every time there is a famine, it turns out later that someone, usually just down the road, was hoarding food for sale. The problem is that the hungry families didn't have the cash to buy it.

Every few years we get news reports that there are only so many days' supply of grain in the world's warehouses. If the warehouses are full, prices fall and farmers stop producing. When they start to empty, prices rise, farmers start planting and soon the warehouses are full again.

Beddington's 'perfect storm' is the operation of a perfect market. Does this mis-diagnosis matter? Even if we grow enough food, surely growing more can't hurt.

Well, yes, it does matter. Because Beddington's planned revolution stands a good chance of making the poor poorer. It could mean we have both more food and more famines. This is because most of the methods he suggests to increase food production are about big farms and big investment.

Beddington wants to plough up vast tracts of African cattle pastures and amalgamate the smallholdings of millions of peasant farmers to create giant, high-tech farms. His blueprint will take land away from the rural poor.

Last month, I watched this scenario playing out on the edge of the Sahara desert in Mali. The government there has recruited foreign experts to help it invest in agriculture. Western aid agencies are building irrigation projects to boost production of rice.

Libya's Colonel Gaddafi, Mali's biggest sugar daddy, has just dug a 25-mile canal to irrigate an area of dry scrub three times the size of the Isle of Wight.

The trouble is that these projects will take water out of the River Niger. They will empty fertile wet pastures just downstream, where one million of Mali's poorest people currently live by catching fish and grazing their cattle. They fear the plans will create desert.

Most of the rice from the new fields will go to feed Libyans. Meanwhile, the poor of the Niger wetlands are likely to join the Al Qaeda groups already penetrating the country's desert borders.

Beddington is right that farming needs investment. But it has to be the right investment. Perhaps he should have a word with another of the Government's scientific advisers, Professor Robert Watson, the real Whitehall food expert.

He is currently chief scientist at the Department for Environment Food and Rural Affairs (DEFRA). Three years ago he chaired an international report on the future of the world's farming.

Watson reached rather different conclusions from Beddington. He said African smallholder farmers should be backed, not stripped of their land; that local knowledge of crops would often work better than high-tech methods; and that fighting poverty was the key to feeding the world.

Watson told me: 'It's not a technical challenge; it's a rural development challenge. Small farmers will remain the predominant producers. The question is how to help them.'

Beddington sees the spread of Western farming methods and giant food and seed companies as the solution to the food problem.

Watson sees it as part of the problem. Beddington's report says: 'We need to make agriculture more efficient.'

But more efficient for whom? For agribusiness and its bottom line? Or for farmers and consumers? In an age where the smart investment banks are putting their cash into biofuels rather than bread, and where large corporations are buying farms across the developing world to grow cotton for cash rather than food for people, the two are not the same thing.

Beddington's report chastises countries such as India, which imposed bans on food exports during the food price crisis in early 2008 in an effort to keep their people fed.

He blames them for 'undoubtedly exacerbating' the crisis, and says such protectionist actions should be banned. He has no such strictures for the speculators who caused the soaring prices.

Surely if we've learned anything over the past couple of years, it is that unbridled markets can bring chaos, and speculators are a menace. It was bad enough letting the financial markets run riot. But if the food markets run riot we will have empty bellies as well as empty pockets.