ERL" href="http://images.iop.org/objects/erw/news/12/4/64/pic1.jpg">Maize yield estimates
ERL" href="http://images.iop.org/objects/erw/news/12/4/64/pic1.jpg">Maize yield estimates

The analysis quantified yield heterogeneity both within and between fields for maize and soybean in the US Midwest. It revealed that heterogeneity in maize yields is rising, with average yield differences between the best and worst soils more than doubling since the turn of the century.

That is both good and bad news, according to David Lobell of Stanford University, US. "It’s good that yields have kept rising, and it’s good that farmers are adopting precision agriculture," he said. "But there are potential problems with relying on a narrowing base of land for gains, just like in the economy there is a problem of relying on growth only [in the] top income brackets. Generally speaking, broad-based growth is more stable and sustainable."

Along with wheat and rice, maize and soybean are the most produced crops in the world, and the US is by far their largest producer. In recent decades the country’s yields of both crops have risen steadily but no-one is certain whether the US will continue to be able to achieve this. Studies have shown that the gap between actual yields in the US and the estimated maximum yields under perfect management has fallen to less than 25%, indicating that yield stagnation could be imminent.

"There are many places in the world where yields kept going up and then suddenly stagnated," Lobell said. "The question is whether we can count on a continuation of the trend."

To answer this, Lobell and his colleague George Azzari at Stanford tried the new approach of assessing yield heterogeneity by drawing on imagery from NASA’s Landsat satellites in conjunction with the Google Earth Engine. The researchers took images of each field in the US Corn Belt twice during the growing season every year since 2000, and predicted yields based on the colour of individual pixels.

"The novelty is that it works even though we don’t rely on any actual field data, so the method is easily applied in new places or years," said Lobell.

Lobell and Azzari’s discovery that yield heterogeneity has been rising for maize – though not for soybean – between and within fields came as a surprise. It suggests that farmers have been adopting precision-agriculture practices and denser sowing, which disproportionately raise yields on better soils. To continue to make gains, the researchers say, farmers should manage nutrients and other inputs, such as seeds, herbicides and pesticides, on a sub-field scale. This will also cut costs and environmental pollution, they add.

But a key message of the study, according to Lobell, is that the satellite data provide a "step change" in the ability to measure agriculture. "New measurements will open up lots of opportunities for growing food better, just like how microscopes opened up lots of possibilities for science and medicine," he said.

The team published the study in Environmental Research Letters (ERL).

Related links

Related stories