LTHO EXAMPLE: EFFECT OF STREAM CHANNEL SIZE ON THE DELIVERY OF NITROGEN TO THE GULF OF MEXICO
Based on the manuscript: “Effect of stream channel size on the delivery of nitrogen to the Gulf of Mexico”, by R. B. Alexander, R. A. Smith, and G. E. Schwarz, Nature, V 403, pp 758-761, 2000.
PROBLEM STATEMENT
· What are the locations of the major sources of nitrogen in the Mississippi River Basin that are delivered to the Gulf of Mexico?
· How does nitrogen removal in streams resulting from denitrification and particulate storage affect the delivery of nitrogen to the Gulf from sources in the Mississippi Basin?
· How do the rates of nitrogen loss vary from headwater streams to large rivers?
· What are the principal hydrologic and physical factors governing spatial variations in these rates?
BACKGROUND
An increase in the flux of nitrogen from the Mississippi river during the latter half of the twentieth century has caused eutrophication and chronic seasonal hypoxia in the shallow waters of the Louisiana shelf in the northern Gulf of Mexico. This has led to reductions in species diversity, mortality of benthic communities and stress in fishery resources. There is evidence for a predominantly anthropogenic origin of the increased nitrogen flux, but the location of the most significant sources in the Mississippi basin responsible for the delivery of nitrogen to the Gulf of Mexico have not been clearly identified, because the
parameters influencing nitrogen-loss rates in rivers are not well known.
LIMITS TO PROGRESS
· The problem of tracing nitrogen through large watersheds stems from the difficulty of establishing a spatially continuous mass balance between three rate variables:
o the in-stream flux of nitrogen
o the rate of nitrogen supply from atmospheric and terrestrial sources
o rate of removal due to denitrification and storage on the landscape and in stream channels
· Much of the controlled study of supply and removal processes has taken place in relatively small watersheds, where landscape and channel conditions are less variable; few measurements of nitrogen-loss rates are available for the relatively heterogeneous basins typical of large river channels.
· The reported range of nitrogen-loss rates for stream and river channels exceeds two orders of magnitude, and few explanations for this large variability have emerged; little has been reported about how the loss rates of nitrogen vary with river channel size.
· In the absence of systematic knowledge of nitrogen-loss rates in streams, no accepted method has emerged for predicting nitrogen transport over long channel distances, such as those found in the Mississippi River Basin.
APPROACH
A recently developed mass-balance method (SPARROW; SPAtially-Referenced Regression On Watershed attributes) was used to estimate nitrogen flux through the interior watersheds of the Mississippi basin. The method correlates observations of stream nitrogen flux (i.e., the response variable) with spatially referenced explanatory data on nitrogen source inputs (e.g., fertilizer use, atmospheric deposition) and factors controlling nitrogen transport in watersheds, including physical characteristics of the landscape (e.g., soil permeability) and stream channels (e.g., channel size, water velocity). The model provides separate empirical estimates of the rates of terrestrial and in-stream removal of nitrogen. The response variable in the model is mean stream nitrogen flux computed from water-column measurements of total nitrogen concentration (TN; sum of nitrate-nitrite and kjeldahl nitrogen-ammonia plus organic nitrogen) and daily flow measurements at 374 river locations in the United States. These monitoring locations include a subset of 123 stations in the watersheds of six major tributaries to the Mississippi river (figure 1).
RESULTS
· The statistically estimated SPARROW model coefficients indicate that the first-order rate of nitrogen removal in streams (i.e., the fraction of nitrogen removed per unit water travel time) declines rapidly with increasing channel depth (figure 2), a result that is consistent with theoretical models of nutrient dynamics in streams. The rates of nitrogen removal via denitrification and particulate storage are expected to be smaller in deeper channels where stream waters have less contact with the benthic sediment.
· When re-expressed as a loss rate per unit water travel time and plotted as a function of depth, the available literature estimates of percentage nitrogen removal agree well with the magnitude and declining monotonic pattern of the SPARROW rates of in-stream loss over a wide range of channel depths (figure 2). Larger variability is observed in nitrogen loss rates in shallow streams, which is likely caused by variability in other stream conditions responsible for nitrogen removal, such as the hyporheic exchange of waters and the organic and oxygen content of sediment.
· The SPARROW estimates of nitrogen loss, in conjunction with river network data on channel size and water travel time, quantify the percentage of nitrogen sources that is delivered by streams to the Gulf of Mexico from each of the interior watersheds in the Mississippi Basin (figure 3). The region of high nitrogen delivery to the Gulf is dendritic in shape, and extends far upstream along the major tributaries in the eastern and central portions of the basin (e.g., Ohio, Tennessee, Lower Missouri, Lower Arkansas and Upper Mississippi rivers). Despite the long water travel times, many watersheds located on large rivers more than 2,500 km from the Gulf deliver much larger fractions of their exported nitrogen (some more than 90%) to coastal waters than watersheds located on smaller streams less than a few hundred kilometers from the Gulf. Nitrogen deliveries from many arid watersheds in the more distant drainages of the western Mississippi basin (i.e., the western portions of the Missouri and Arkansas/Red regions) are uniformly small because of the effect of the typically shallow rivers with high nitrogen-loss rates and the lengthy water travel times to the Gulf.
· The quantities of nitrogen delivered to the Gulf of Mexico from interior watersheds (i.e., delivered yield; figure 4) reflect spatial differences in both the supply and processing of nitrogen on the landscape and in the rivers of the Mississippi basin. The largest quantities delivered to the Gulf originate in the northeastern and north central portions of the Mississippi basin, areas containing large amounts of corn and soybean acreage, livestock, atmospheric deposition, and municipal wastes. Two regions, the Central Mississippi and Ohio/Tennessee, account for about 70% of the nitrogen delivered to the Gulf although they contain less than 30% of the total area of the Mississippi basin.
CONCLUSIONS
· The first-order rate of in-stream nitrogen removal varies over more than two orders of magnitude and declines rapidly with increases in channel size.
· As a result, the length of time that surface waters spend in channels of a given size greatly influences the quantities of nitrogen delivered to coastal waters.
· The delivery of nitrogen to coastal systems from point and diffuse sources is not a simple function of the distance of these sources from coastal waters.
· Instead, the proximity of sources to large streams and rivers (as measured by the length of time that surface waters must travel through smaller tributaries) is a major determinant of their downstream transport to marine systems.
3