CHAMPAIGN, Ill. A computational study on nitrogen inputs to the Mississippi River Basin from the 1950s to the 1990s suggests that better use of the fertilizer such as not over-applying it could substantially reduce the amount of nitrates flowing down river without compromising crop yields.
The study, appearing in the Nov. 8 issue of the journal Nature, concluded that had there been a 12 percent reduction in nitrogen fertilizer use in the last two decades, there could have been a 33 percent reduction in the nitrate flux in the Mississippi River and Gulf of Mexico. In the gulf, excess nitrogen has been suspected as a major contributor of seasonal dead zones where oxygen levels are depleted and marine life is reduced.
"An earlier study estimated that a 24 percent reduction in fertilizer use would be needed to achieve the same level of reduction of nitrate movement to the gulf, but our results indicate that increasing the efficiency of fertilizer use may have a greater impact than previously thought," said Gregory McIsaac, a professor of environmental sciences at the University of Illinois and lead author of the Nature article.
Although the precise cause-and-effect relationship between fertilizer use and the hypoxic zones is still uncertain, McIsaac said, the fact remains that nitrogen going into to the Mississippi River Basin increased faster than the amount of nitrogen harvested in crops in the 1960s and 1970s. Nitrogen that is not taken up by plants becomes available to leach into groundwater and rivers.
As the difference between nitrogen inputs and outputs onto the land became larger, so have nitrate concentrations in the lower Mississippi River. But the UI model produced a surprising result.
"As net nitrogen inputs increased in the basin, the percentage that appeared in the river as nitrate also increased," said study collaborator Mark David, a UI professor of biogeochemistry. "It seems that when the capacity of the land to use or store nitrogen is exceeded, further nitrogen inputs dramatically increase losses to streams."
"So our model," McIsaac said, "also suggests the reverse that a relatively modest reduction in nitrogen fertilizer use, while maintaining crop yields, could substantially reduce the amount of nitrate found in the Mississippi River."
Although there has been improvement in efficiency since 1988, he said, data collected as recently as last year show that some Illinois corn growers could still reduce nitrogen use without reducing their crop yields.
"In a survey conducted in 2000, about 30 percent of Illinois farmers indicated that they apply more nitrogen than is recommended for economically optimum crop production. Eliminating that over-application will maintain yields, reduce costs and, according to our analysis, reduce the nitrate in the Mississippi River," McIsaac said.
The UI model accounts for 95 percent of the annual variation in nitrate-nitrogen delivered to the Gulf of Mexico as measured by the U.S. Geological Survey. It also incorporates the state's level crop production statistics and estimates of nitrogen input from the atmosphere in rainfall. Most of the data were compiled under the leadership of Donald Goolsby, a scientist with the U.S. Geological Survey, another collaborator.
"We conducted a very comprehensive uncertainty assessment of the model and its inputs," said George Gertner, UI professor of biometrics and co-author. "The empirical model proved to be very robust."
"A main difference between our model and some others is that we assumed a steady state of soil organic nitrogen," David said. "In other words, we assumed that there has not been a large change in soil organic nitrogen over the years.
The assumption is supported by long-term soil monitoring data from the UI Morrow Plots and the Sanborn Field in Missouri. That data produced a more accurate correlation between net nitrogen inputs to the watershed and river nitrogen, David said. However, correlation studies do not prove cause and effect, and more studies are needed, the researchers stressed.
"Even with the uncertainty of correlations, it makes sense to continue to improve the efficiency of fertilizer use by following university recommended practices," David said. Among those, he added, are fertilizing for five-year average corn yields not maximum yield and counting all forms of applied nitrogen such as manure and credit from soybeans and other legume crops.
The full effect of reducing nitrogen fertilizer application to UI-recommended levels would depend on weather conditions and could take up to nine years to be observed.
"Our model provides some insights, but we need to continue to refine our understanding of nitrogen processes in soils and rivers by monitoring to see what happens in response to changes in fertilizer use and weather," McIsaac said.
Funding for the study was provided by the University of Illinois and the Illinois Agricultural Experiment Station. Much of the data had been compiled in a previous study funded by the U.S. Geological Survey.