June 4, 2009

Scientists from the Agricultural Research Service (ARS) and colleagues have developed new mathematical models that could help farmers use climate patterns to predict corn yields.

 

Farmers could use this information, which indicates yield cycles of about two years, to adjust their production practices. For example, crops grown in low-yield years may require less fertiliser.

 

The adjustments could help reduce the flow of excess nitrate from crop fertilisers into surrounding watershed, which may help control hypoxia downstream in the Gulf of Mexico.

 

Corn yield variability affects nitrate loss because small changes in corn yield may have greater effects on nitrate loss in fields with subsurface tile drainage systems.

 

Agricultural engineer Rob Malone and team gathered more than 50 years of data on corn production from six high-yield corn-producing counties in Iowa to see if they could identify key correlations among yield, weather conditions and climate indices.

 

Malone's modelling results indicated that high surface radiation and low temperature early in the growing season often produce high yields when followed by sufficient rainfall later in the growing season. This model accounted for 89 percent of the variation in annual corn yields.

 

Changes in these weather variables are often associated with long-term climate trends. So the team used established climate indices derived from the large-scale flow of high- and low-pressure air masses and equatorial stratospheric winds to develop models that accounted for the variability in corn yields. This model detected an average difference between high- and low-yielding years of 19 percent and identified an approximate two-year cycle between high- and low-yielding years.

 

Malone's research helps explain the combined effect of several long-term climate trends on long-term US corn yields. In addition, these results provide information about how the annual variation in ground-level solar radiation during the growing season affects long-term corn yield in the US.