China's peasants look to the skies

But the science of yields is unyielding

May 20th 2010 | From The Economist print edition

Another mineshaft

IF THE Andhra Pradesh principles point the way to a reasonably equable future, they will have to be adopted not only throughout South Asia but also in China, where the water available to each person is only a quarter of the world average. In the rain-starved north, the availability per person is only a quarter of that in the south. Yet this is where almost half China’s population lives, and where most of its maize, wheat and vegetables are grown.

Water scarcity is hardly new in China, whose irrigation records go back 4,000 years, but the use of groundwater is. In the 1950s this was virtually unknown in the north. Today there are more wells there than anywhere else in the world, and they are relentlessly pumped, with alarming results. For instance, in the Hai river basin, in which both Beijing and Tianjin lie, shallow water tables have dropped by up to 50 metres, deep ones by up to 90. These will not quickly be put right.

Chinese governments have usually responded to shortages with canals, dykes, storage ponds and so on. The 1,800km Grand Canal, started in 486BC, was built chiefly to move grain to the capital, but will now become part of the great South-North Water-Transfer Project, intended to slake the thirst of China’s arid regions. Dams and canals appeal to the engineers who are disproportionately represented in China’s government. And the country’s engineers are still taught that the way to “save” water is to improve the way it is delivered—by lining irrigation canals, for instance, or laying pipes—to reduce the water that is “lost” by seeping into the soil.

In truth, though, such water is not all lost: much of it returns to the aquifers below, from where it can be pumped up again. There is a cost to this, in energy and therefore cash, but not in water. The only water truly lost in a hydrologic system is through evapotranspiration, since no one can make further use of it once it is in the atmosphere. If genuine savings are to be made, either evaporation must be cut (for example, by storing water underground, or by delivering it to plants’ roots under the surface of the soil); or food must be produced with less transpiration.

The trouble with efficiency savings
Almost all China’s (and others’) attempts at using groundwater more efficiently so far have foundered on a failure to grasp these facts. The water “saved” by sprinklers, lined canals and other forms of seepage control has simply been used to expand the area under irrigation. Over the past 30 years this has gone up by 8m hectares, allowing food production to increase even though the amount of pumped water has remained much the same. But ET has risen, so aquifer depletion has continued.

Sometimes, say Frank Ward, of New Mexico State University, and Manuel Pulido-Velázquez, of the Technical University of Valencia, policies aimed at reducing water use can actually increase groundwater depletion. This has happened in the Upper Rio Grande basin shared by the United States and Mexico, where measures designed to achieve more efficient irrigation have led to an increase in yields upstream; this in turn has increased ET, leaving less water available for aquifer recharge.

Such discoveries increase the attractiveness of demand management, and that is being tried in China as well as in Andhra Pradesh. In a project that covers several parts of arid and semi-arid China—Beijing, Hebei, Qingdao and Shenyang, as well as the Hai basin and the smaller Turpan basin—the World Bank has been promoting water conservation. Elements of the approach are similar to that in Andhra Pradesh: farmers gather in water-users’ associations to plan and operate irrigation services, for example. But the aim here is specifically to reduce ET, at the same time increasing farmers’ incomes without depleting the groundwater.

This is high-tech stuff that involves not just drip irrigation and condensation-trapping greenhouses, but remote sensing by satellites which provide ET readings for areas of 30 by 30 metres. This tells farmers how much water they can consume without adversely affecting the ecosystems in their river basin. If the project is successful, as a pilot has been, it will also establish the use of an internet-based management system, mitigate losses from flooding and increase the supply of water to industry.

If such practices were extended across Asia, groundwater depletions might well be arrested. With luck, farmers too would be better off. But would they produce enough food for the extra 2.5 billion people expected by 2050 in today’s developing countries? The constraints seem to be set by science, and they are tight.

Growing more crops over a wider area leads to more ET. The yield of a crop can be increased a bit by giving plants only as much water as they need and no more; but, says Dr Perry, the water accountant, the productivity gains are unlikely to exceed 10%. Increases in biomass—total vegetative matter—are matched almost proportionately by increases in transpiration, unless humidity or nutrients are changed or the plant is modified genetically. But so far, he notes, “the fundamental relationship between biomass and transpiration has not been changed.”