Variation in Riverine Nitrate Flux and Fall Nitrogen Fertilizer Application in East‐Central Illinois

In east‐central Illinois, fertilizer sales during the past 20 yr suggest that approximately half of the fertilizer nitrogen (N) applied to corn ( Zea mays L.) occurs in the fall; however, fall fertilizer N sales were greatly reduced in 2009 as wet soil conditions restricted fall fieldwork, including fertilizer N applications. In 2010, we observed unusually low flow‐weighted nitrate concentrations (approximately 40% below the long‐term average) in two east‐central Illinois rivers (5.7 mg N L −1 in the Embarras River and 5.6 mg N L −1 in the Lake Fork of the Kaskaskia River). Using long‐term river nitrate data sets (1993–2012 for the Embarras and 1997–2012 for the Kaskaskia), we examined nitrate concentrations and developed regression models to estimate the association between fall fertilizer N application on riverine nitrate yields in these tile‐drained watersheds. During these periods of record, annual riverine nitrate yields ranged from 8 to 57 kg N ha −1 yr −1 (30 kg N ha −1 yr −1 average) for the Embarras River and 2.6 to 59 kg N ha −1 yr −1 (32 kg N ha −1 yr −1 average) for the Kaskaskia. Multivariate linear regression relationships with the current and previous year's annual water yields, previous year's corn yield, and nine‐county fall fertilizer sales accounted for 96% of the annual variation in nitrate yield in both watersheds. Running the regression models with fall fertilizer sales set to the 2009 amount suggests that the average reduction in nitrate yield (for the period of record) would be 17 and 20% for the Embarras and Kaskaskia Rivers, respectively. These data suggest that shifting fertilizer N application to the spring can be detected in watersheds as large as 481 km 2 .