Description:
About Mississippi tributary nitrate concentrations
Nutrients (fertilizers) applied to croplands in the Mississippi River drainage basin often make their way into rivers and streams, and ultimately are discharged into the Gulk of Mexico. Fertilizers in the Gulf of Mexico increase the production of marine algae, thereby decreasing the oxygen content of coastal waterways. Massive algal blooms following wet season runoffs can cause severe hypoxia, threatening a wide variety of marine organisms, including commercial fisheries.
Compounds of nitrogen (such as nitrates) are one of the most important fertilizers used on croplands in the central United States. As a result, about 1 million metric tons of nitrate are carried into the Gulf of Mexico every year by streams draining the Mississippi River basin. This value of nitrate flux is about 3 times higher than the flux 30 years ago, corresponding to increasing use of fertilizers to boost or maintain crop yields.
In order to quantitatively evaluate the effect of agricultural practices on water quality, Goolsby et al (1999) made a comprehensive analysis of 42 sub-basins within the Mississippi-Atchafalaya drainage system. Individual interior basins range in size from 7000 km2 to 237,000 km2. Each basin was measured for the percentage of land in row crops (corn, soybeans and/or sorghum), which ranged from less than 0.1% to 74% (see data table). The concentration of nitrate in river water discharging from the "outlet" of each of these basins was also measured repeatedly (from 40 to 300 times), and the mean nitrate concentration was calculated (see data table). Concentrations are reported in milligrams of nitrate per liter of water (mg/L), which is essentially equivalent to parts per million (ppm). Concentrations range from near zero to almost 7 mg/L.
The graph shows a scatterplot of % cropland versus nitrate concentration. There is a clear overall positive correlation between these two values, suggesting that agricultural practices are, indeed, influencing the runoff of nitrates. One could try fitting both a linear and exponential model to these data and comparing the two. Must the best fit regression pass through the origin? What is the maximum possible value for nitrate concentration, as predicted by forward extrapolation of a regression model?
Interior basin # 24, the Kaskaskia River basin in Illinois, has an anomalously low nitrate concentration. One explanation is that algal blooms in the reservoirs behind two large dams on the Kaskaskia strip out nitrates, which are incorporated into bottom sediments when the algae die. Should more dams be built to trap nitrates? Interior basin #35, the St. Francis basin in Arkansas, is mostly planted in soybeans, a nitrogen fixer, which requires very little nitrogen from fertilizers.
