5
Sources of Nutrient Inputs to Estuaries and Coastal Waters

KEY POINTS IN CHAPTER 5

This chapter reviews the sources and amounts of nutrients supplied to coastal water bodies and finds:

  • Globally, human activity has dramatically increased the flux of phosphorus (by a factor of almost 3) to the world’s oceans. There has been an even more dramatic increase in nitrogen flux, especially in the last 40 years, with the greatest flux adjacent to areas of highest population density. Human activity has increased the flux of nitrogen in the Mississippi River by some 4-fold, in the rivers in the northeastern United States by some 8-fold, and in the rivers draining to the North Sea by more than 10-fold.

  • Although point source nutrients are the major problem for small watersheds adjacent to major population centers, these inputs are relatively easy to minimize with tertiary wastewater treatment processes. In contrast, nutrients from nonpoint sources have become the dominant and least easily controlled component of nutrients transported into coastal waters from large watersheds, and especially from watersheds with extensive agricultural activity or atmospheric nitrogen pollution.

  • Phosphorus flux to estuaries is dominantly derived from agricultural activities as particle-bound forms mobilized in runoff. In some areas, groundwater transported phosphorus is also important.

  • Nitrogen input to estuaries is derived from both agricultural activity (e.g., dominant in the Mississippi River) and fossil-fuel combustion (e.g., dominant in the northeastern United States). Animal feeding operations have become a major contributor to nitrogen exports.

  • It is likely that the atmospheric component of nitrogen flux into estuaries has previously been under-estimated. This component is derived from fossil-fuel



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution 5 Sources of Nutrient Inputs to Estuaries and Coastal Waters KEY POINTS IN CHAPTER 5 This chapter reviews the sources and amounts of nutrients supplied to coastal water bodies and finds: Globally, human activity has dramatically increased the flux of phosphorus (by a factor of almost 3) to the world’s oceans. There has been an even more dramatic increase in nitrogen flux, especially in the last 40 years, with the greatest flux adjacent to areas of highest population density. Human activity has increased the flux of nitrogen in the Mississippi River by some 4-fold, in the rivers in the northeastern United States by some 8-fold, and in the rivers draining to the North Sea by more than 10-fold. Although point source nutrients are the major problem for small watersheds adjacent to major population centers, these inputs are relatively easy to minimize with tertiary wastewater treatment processes. In contrast, nutrients from nonpoint sources have become the dominant and least easily controlled component of nutrients transported into coastal waters from large watersheds, and especially from watersheds with extensive agricultural activity or atmospheric nitrogen pollution. Phosphorus flux to estuaries is dominantly derived from agricultural activities as particle-bound forms mobilized in runoff. In some areas, groundwater transported phosphorus is also important. Nitrogen input to estuaries is derived from both agricultural activity (e.g., dominant in the Mississippi River) and fossil-fuel combustion (e.g., dominant in the northeastern United States). Animal feeding operations have become a major contributor to nitrogen exports. It is likely that the atmospheric component of nitrogen flux into estuaries has previously been under-estimated. This component is derived from fossil-fuel

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution combustion and from animal feedlots and other agricultural sources, and is both deposited directly into estuaries and also deposited initially onto the land surface and then carried into estuaries by runoff. Human activity has an enormous influence on the global cycling of nutrients, especially on the movement of nutrients to estuaries and other coastal waters. For phosphorus, global fluxes are dominated by the essentially one-way flow of phosphorus carried in eroded materials and wastewater from the land to the oceans, where it is ultimately buried in ocean sediments (Hedley and Sharpley 1998). The size of this flux is currently estimated at 22 Tg P yr−1 (Howarth et al. 1995). Prior to increased human agricultural and industrial activity, the flow is estimated to have been around 8 Tg P yr−1 (Howarth et al. 1995). Thus, current human activities cause an extra 14 Tg of phosphorus to flow into the ocean sediment sink each year, or approximately the same as the amount of phosphorus fertilizer (16 Tg P) applied to agricultural land each year. The effect of human activity on the global cycling of nitrogen is equally immense, and furthermore, the rate of change in the pattern of use is much greater (Galloway et al. 1995). The single largest global change in the nitrogen cycle comes from increased reliance on synthetic inorganic fertilizers, which accounts for more than half of the human alteration of the nitrogen cycle (Vitousek et al. 1997). The process for making inorganic nitrogen fertilizer was invented during World War I, but was not widely used until the 1950s. The rate of use increased steadily until the late 1980s, when the collapse of the former Soviet Union led to great disruptions in agriculture and fertilizer use in Russia and much of eastern Europe. These disruptions resulted in a slight decline in global nitrogen fertilizer use for a few years (Matson et al. 1997). By 1995, the global use of inorganic nitrogen fertilizer was again growing rapidly, with much of the growth driven by increased use in China (Figure 5-1). Use as of 1996 was approximately 83 Tg N yr−1. Approximately half of the inorganic nitrogen fertilizer that was ever used on Earth has been applied during the last 15 years. Production of nitrogen fertilizer is the largest process whereby human activity mobilizes nitrogen globally (Box 5-1). However, other human-controlled processes, such as combustion of fossil fuels and production of nitrogen-fixing crops in agriculture, convert atmospheric nitrogen into biologically available forms of nitrogen. Overall, human fixation of nitrogen (including production of fertilizer, combustion of fossil fuel, and production of nitrogen-fixing agricultural crops) increased globally some two- to three-fold between 1960 to 1990 and continues to grow (Galloway et al. 1995). By the mid 1990s, human activities made new nitrogen avail-

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution FIGURE 5-1 Annual global nitrogen fertilizer consumption for 1960-1995 (1 Tg = 1012 g; data from FAO 1999). The rate of increase was relatively steady until the late 1980s, when collapse of the former Soviet Union reduced fertilizer use in Russia. Fertilizer use is growing again, driven in large part by use in China (modified from Matson et al. 1997). BOX 5-1 The Fate of Nitrogen Fertilizer in North America When nitrogen fertilizer is applied to a field, it can move through a variety of flow paths to downstream aquatic ecosystems (Figure 5-2). Some of the fertilizer leaches directly to groundwater and surface waters, with the range varying from 3 percent to 80 percent of the fertilizer applied, depending upon soil characteristics, climate, and crop type (Howarth et al. 1996). On average for North America, some 20 percent is leached directly to surface waters (NRC 1993a; Howarth et al. 1996). Some fertilizer is volatilized directly to the atmosphere; in the United States, this averages 2 percent of the application, but the value is higher in tropical countries and also in countries that use more ammonium-based fertilizers, such as China (Bouwman et al. 1997). Much of the nitrogen from fertilizer is incorporated into crops and is removed from the field in the crops when they are harvested, which is of course the objective of the farmer. A recent National Research Council report

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution (NRC 1993a) suggests that on average 65 percent of the nitrogen applied to croplands in the United States is harvested, although other estimates are somewhat lower (Howarth et al. 1996). By difference, on average approximately 13 percent of the nitrogen applied must be building up in soils or denitrified to nitrogen gas. Since much of the nitrogen is harvested in crops, it is important to trace its eventual fate. The majority of the nitrogen is fed to animals (an amount equivalent to 45 percent of the amount of fertilizer originally applied, if 65 percent of the nitrogen is actually harvested in crops; Bouwman and Booij 1998). Some of the FIGURE 5-2 The average fate of nitrogen fertilizer applied to agricultural fields for North America. The numbers in parentheses are calculated by difference, and the other numbers are direct estimates (unpublished figure by R. Howarth).

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution nitrogen is directly consumed by humans eating vegetable crops—in North America perhaps 10 percent of the amount of nitrogen originally applied to the fields (Bouwman and Booij 1998). By difference, perhaps 10 percent of the amount of nitrogen originally applied to fields is lost during food processing, being placed in landfills or released to surface waters from food-processing plants. Of the nitrogen that is consumed by animals, much is volatilized from animal wastes to the atmosphere as ammonia. In North America, this volatilization is roughly one-third of the nitrogen fed to animals (Bouwman et al. 1997), or 15 percent of the amount of nitrogen originally placed on the fields. This ammonia is deposited back onto the landscape, often near the source of volatilization, although some of it first travels for long distances through the atmosphere (Holland et al. 1999). Some of the nitrogen in animals is consumed by humans, an amount roughly equivalent to 10 percent of the amount of nitrogen fed to the animals, or 4 percent of the nitrogen originally applied to fields. By difference, the remainder of the nitrogen—over 25 percent of the amount of nitrogen originally applied to the fields—is contained in animals wastes that are building up somewhere in the environment. Most of this may be leached to surface waters. Of the nitrogen consumed by humans, either through vegetable crops or meat, some is released through wastewater treatment plants and from septic tanks. In North America, this is an amount equivalent to approximately 5 percent of the amount of nitrogen originally applied to fields (Howarth et al. 1996). By difference, the rest of the nitrogen is placed as food wastes in landfills or is denitrified to nitrogen in wastewater treatment plants and septic tanks. The conclusion is that fertilizer leaching from fields is only a portion of the nitrogen that potentially reaches estuaries and coastal waters. Probably of greater importance for North America as a whole is the nitrogen that is volatilized to the atmosphere or released to surface waters from animal wastes and landfills. Since food is often shipped over long distances in the United States, the environmental effect of the nitrogen can occur well away from the original site of fertilizer application. able at a rate of some 140 Tg N yr-1 (Vitousek et al. 1997), matching the natural rate of biological nitrogen fixation on all the land surfaces of the world (Vitousek et al. 1997; Cleveland et al. 1999). Thus, the rate at which humans have altered nitrogen availability globally far exceeds the rate at which humans have altered the global carbon cycle (Figure 5-3). The human alteration of nutrient cycles is not uniform over the earth, and the greatest changes are concentrated in the areas of greatest popula-

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution FIGURE 5-3 The relative change in nitrogen fixation caused by human activities globally compared to the increase in carbon dioxide in the atmosphere since 1900. Note that humans are having a much greater influence on nitrogen availability than they are on the production of carbon dioxide, an important greenhouse gas (modified from Vitousek et al. 1997). tion density and greatest agricultural production. Some regions of the world have seen very little change in the flux of either nitrogen or phosphorus to the coast (Howarth et al. 1995, 1996), while in other places the change has been tremendous. Human activity is estimated to have increased nitrogen inputs to the coastal waters of the northeastern United States generally, and to Chesapeake Bay specifically, by some six- to eight-fold (Boynton et al. 1995; Howarth et al. 1996; Howarth 1998). Atmospheric deposition of nitrogen has increased even more than this in the northeast (Holland et al. 1999). The time trends in human perturbation of

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution nutrient cycles can also vary among regions. For example, while the global use of inorganic nitrogen fertilizer continues to increase, the use of nitrogen fertilizer in the United States has increased relatively little since 1985 (Figure 5-4; Evans et al. 1996). Note, however, that the use of nitrogen fertilizer in the United States in the next century may again increase to support greater exports of food to developing countries. Countries such as China have been largely self sufficient in food production for the past two decades, in part because of increased use of nitrogen fertilizer. The use of fertilizer in China is now very high—almost 10-fold greater than in the United States—and further increases in fertilizer use are less likely to lead to huge increases in food production as they have in the past. Therefore, if China’s population continues to grow it may once again be forced to import food from the United States and other developed countries, leading to more use of nitrogen fertilizer here. WASTEWATER AND NONPOINT SOURCE INPUTS Traditionally, most water quality management emphasizes control of discharges from wastewater treatment plants and other point sources. FIGURE 5-4 U.S. commercial fertilizer use (modified from Evans et al. 1996).

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution However, generally of greater concern for nutrients and coastal eutrophication are “nonpoint sources” of nutrients (NRC 1993a). A regional-scale analysis of fluxes of nitrogen from the landscape to the coast of the North Atlantic Ocean demonstrated that nonpoint sources of nitrogen exceeded sewage inputs for all regions in both Europe and North America (Howarth et al. 1996). Overall, sewage contributed only 12 percent of the flux of nitrogen from the North American landscape to the North Atlantic Ocean (Howarth et al. 1996). Nonpoint sources also dominate for phosphorus inputs to surface waters in the United States (Sharpley and Rekolainen 1997; Carpenter et al. 1998), and because of an effort to control phosphorus point source pollution, nonpoint sources of phosphorus have grown in relative importance since 1980 (Jaworski 1990; Sharpley et al. 1994; Litke 1999). Wastewater inputs can sometimes be a major source of nitrogen to an estuary when the watershed is heavily populated and small relative to the surface area of the estuary itself (Nixon and Pilson 1983). Even in some estuaries fed by larger watersheds, wastewater can be the largest source of nitrogen if the watershed is heavily populated. For example, wastewater contributes an estimated 60 percent of the nitrogen inputs to Long Island Sound, largely due to sewage from New York City (CDEP and NYSDEC 1998). However, nitrogen and phosphorus inputs from nonpoint sources in most estuaries are greater than are inputs from wastewater, particularly in estuaries that have relatively large watersheds (NRC 1993a). For example, only one-quarter of the nitrogen and phosphorus inputs to Chesapeake Bay come from wastewater treatment plants and other such point sources (Boynton et al. 1995; Nixon et al. 1996). For the Mississippi River, sewage and industrial point sources contribute an estimated 10 percent (Howarth et al. 1996) to 20 percent (Goolsby et al. 1999) of the total nitrogen flux (organic and inorganic nitrogen) and 40 percent of the total phosphorus flux (Goolsby et al. 1999). As discussed in more detail in Chapter 9, many technologies exist for reducing nutrient discharges from wastewater treatment plants. The relatively standard approaches of using primary and secondary sewage treatment lower phosphorus and nitrogen discharges on average by approximately 20 percent to 25 percent, although there is a significant variation among plants (Viessman and Hammer 1998; NRC 1993a). Additional tertiary treatment for nutrient removal can lower nitrogen discharges by 80 percent to 88 percent and phosphorus discharges by 95 percent to 99 percent (NRC 1993a). However, most wastewater treatment plants in the United States do not have adequate nitrogen removal capabilities. In Tampa Bay, wastewater treatment plants were a major source of nitrogen prior to the institution of tertiary nitrogen removal, and this treatment has

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution successfully reversed the trend in eutrophication there (Johansson and Greening 2000). Reduction in the eutrophication of most estuaries requires the management of nutrient inputs from nonpoint sources in addition to those of wastewater treatment plants and industrial sources (NRC 1993a). The nature of these sources is described in the remainder of this chapter. DISTURBANCE, NONPOINT NUTRIENT FLUXES, AND BASELINES FOR NUTRIENT EXPORTS FROM PRISTINE SYSTEMS In a landscape that is completely undisturbed by humans, export of nitrogen and phosphorus to downstream aquatic ecosystems tends to be small, particularly in the temperate zone (Hobbie and Likens 1973; Omernik 1977; Rast and Lee 1978; Howarth et al. 1996). Assuming that the landscape is in an ecological steady state, the export of nutrients cannot exceed the inputs. For nitrogen, these inputs are biological nitrogen fixation and deposition of nitrogen compounds from the atmosphere; in the temperate zone both tend to be small in the absence of human disturbance (Howarth et al. 1996; Cleveland et al. 1999; Holland et al. 1999). Thus, the export of nitrogen from undisturbed temperate landscapes must also be low, in fact lower than the input because there is some accumulation of nitrogen in the system and some loss of nitrogen through denitrification (the bacterial conversion of reactive nitrate into nonreactive molecular nitrogen). For tropical regions, rates of biological nitrogen fixation and natural deposition of nitrogen from the atmosphere are far higher, and so nitrogen export to downstream ecosystems from undisturbed ecosystems may also be greater (Howarth et al. 1996; Cleveland et al. 1999; Holland et al. 1999; Lewis et al. 1999). Unfortunately, it is difficult to determine with any precision the magnitude of the natural flux of nitrogen from a temperate landscape like the United States. Atmospheric pollution and the resulting elevated nitrogen deposition are widespread, providing some level of disturbance virtually everywhere in the country, and in fact in most of the world’s temperate ecosystems (Holland et al. 1999). There are a few remaining temperate forests that do not receive elevated nitrogen deposition from pollution sources, such as some remote forests in Chile (Hedin et al. 1995). However, these are poor models for most of the temperate systems of the United States as the Chilean forests receive high precipitation and runoff, and have vastly different ecological histories. An expert panel under the auspices of the International SCOPE (Scientific Committee on Problems of the Environment) Nitrogen Project estimated that pristine temperate-zone ecosystems, such as those that had

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution characterized much of North America and Europe prior to human disturbance, would export between 75 and 230 kg N km−2 yr−1 to downstream aquatic ecosystems, with the median estimate being 133 kg N km−2 yr−1 (Howarth et al. 1996; Howarth 1998). This provides the best estimate available for the natural, background load of nitrogen from the landscapes in the continental United States. Valigura et al. (2000) estimate that for estuaries with small watersheds, the nitrogen flux off the pristine landscape prior to European colonization was at the low end of this range, perhaps 78 to 108 kg N km−2 yr−1. Assuming a baseline flux of 133 kg N km−2 yr−1 for an undisturbed temperate landscape, human activity has increased the nitrogen flux in the Mississippi River by more than 4-fold, in the rivers of the northeastern United States by 8-fold, and in the rivers draining to the North Sea by 11-fold (Howarth 1998). In an independent analysis for Chesapeake Bay, Boynton et al. (1995) estimated that nitrogen fluxes have increased some 6- to 8-fold since pre-colonial times, a value consistent with the conclusion from the International SCOPE Nitrogen Project. In an undisturbed landscape, the major source of phosphorus to a terrestrial ecosystem is the weathering of the soil and parent-rock material, which tends to be relatively slow and therefore sets a low limit on the export of phosphorus. As a global average, the export of phosphorus from the terrestrial landscape prior to human disturbance can be estimated from the oceanic sedimentary record and was somewhat greater than 50 kg P km−2 yr−1, expressed per area of land surface (Howarth et al. 1995). However, this clearly depends on the phosphorus content of the parent-rock material, the rate of weathering, and other environmental conditions, including the rate of erosion. The current flux of phosphorus from the landscape is in fact less than 50 kg P km−2 yr−1 for more than half of the area in the Mississippi River basin (Goolsby et al. 1999), and is only 5 kg P km−2 yr−1 for the watersheds of Hudson’s Bay, Canada (Howarth et al. 1996). On the other hand, the rather large export of phosphorus from the Amazon River basin of over 230 kg P km−2 yr−1 appears to be a largely natural phenomenon (Howarth et al. 1996). Given the site-specific nature of phosphorus export and the paucity of information on background phosphorus losses from a given location prior to cultivation, no baseline for the natural rate of phosphorus export exists. Disturbance of the landscape increases the export of both nitrogen and phosphorus, although there are some major differences in the responses of these two nutrients. As a general rule, most export of phosphorus from disturbed systems occurs as phosphorus bound to particles, so factors regulating erosion and sedimentation are critical in controlling phosphorus fluxes. An important exception can occur in sandy soils with

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution low phosphorus adsorption capacities (Sharpley et al. 1998; Sims et al. 1998); such soils can be important to consider when managing eutrophication in portions of the Atlantic coastal plain. Some phosphorus moves through the atmosphere as dust particles, and this can contribute greatly to the phosphorus economies of some remote oceanic waters and forests. In general, such inputs of phosphorus to estuaries and coastal waters are not as important as inputs in surface waters. For nitrogen, some export also occurs in particle-bound forms, but nitrogen tends to be much more mobile through soils in dissolved form than phosphorus, so significant exports can occur in groundwater (Paerl 1997) or as dissolved nitrogen in surface waters. In addition (and also unlike phosphorus), reactive nitrogen compounds can be quite mobile in the atmosphere. For example, significant amounts of ammonia gas from agricultural sources (particularly urea- and ammonia-based fertilizers, manures, and animal feedlot wastes) volatilize to the atmosphere and are deposited elsewhere in the landscape (Bouwman et al. 1997; Holland et al. 1999). Globally, of the 60 to 80 Tg N yr−1 applied as inorganic nitrogen fertilizer, 21 to 52 Tg N yr−1 are estimated to be volatilized to the atmosphere as ammonia, either directly from the fertilizer or from animal waste (Holland et al. 1999). That is, on average some 40 percent of the inorganic nitrogen fertilizer that is applied cycles through the atmosphere and is redeposited. In the United States, the value is somewhat lower, but still 25 percent of the inorganic nitrogen fertilizer that is used is volatilized to the atmosphere (Holland et al. 1999). For phosphorus, agriculture is the largest disturbance controlling nonpoint fluxes of phosphorus in the landscape (Carpenter et al. 1998). For nitrogen, both agriculture and fossil-fuel combustion contribute significantly to nonpoint source flows to estuaries and coastal waters (Howarth et al. 1996). Some of this nitrogen export comes directly from agricultural fields, but because of both substantial nitrogen transport in the atmosphere and nitrogen mobility in dissolved forms, the nitrogen export from other types of ecosystems, including forests, can be substantial. Since agriculture dominates the nonpoint source flux of phosphorus and contributes significantly to nonpoint sources of nitrogen (often dominating it as well), changes in agricultural practices over the last few decades contribute to these nutrient fluxes. Industrial and fossil fuel sources of nitrogen and the mechanisms that control both nitrogen and phosphorus fluxes in the landscape will be discussed later in this chapter.

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution analysis) provide a potential framework based on quantifying inputs to the watershed, but these analyses are relatively recent and have not yet been applied to the management of most estuaries. In an effort to determine the validity of using SPARROW-derived estimates for a given estuary, Valigura et al. (2000) conducted a preliminary comparison of SPARROW-derived estimates with independently derived estimates of nitrogen loading to 27 estuaries on the Atlantic and Gulf of Mexico coasts of the United States. Based on that comparison, Valigura et al. (2000) concluded that while SPARROW accurately predicted the mean loading to the estuaries as a group, it did a poor job of predicting the load to any one particular estuary (i.e., a linear regression of the SPARROW estimates and the locally derived estimates had a slope of 1 and an R2 of 0.49). However, as with many such analyses involving locally derived information, the observed data from each estuary varies in quality and quantity and the methods used to calculate estimates varied as well. Thus, the locally derived estimates were not obtained from directly comparable data sets and most were not verified. Thus the poor match between SPARROW predictions and local estimates may lie with the quality of the individual estimates for the 27 estuaries. (Chapters 7 and 8 expand on the limitations imposed on understanding individual estuarine behavior by inconsistent observations.) Perhaps the greatest uncertainty with estuary nitrogen budgets concerns the contribution of atmospheric deposition. In most classical estuarine studies, nitrogen inputs from the atmosphere were completely ignored. This has changed since Fisher and Oppenheimer (1991) pointed out the potential importance of atmospheric deposition as a source of nitrogen to Chesapeake Bay, and since Paerl (1985) showed the importance of atmospheric deposition as a nitrogen source to the coastal waters of North Carolina. However, even many nutrient budgets constructed during the last decade have no estimate for the input of nitrogen from atmospheric deposition. In many other estuaries, budgets estimate the importance only of direct deposition onto the surface waters of the estuary itself (and generally only wet deposition, not dry deposition), and do not estimate deposition onto the landscape with subsequent export to the estuary. Available evidence (although constrained by limited monitoring) indicates that direct deposition onto the water surface alone (not including the contribution of nitrogen which falls on the landscape and is then exported to estuaries) contributes between 1 percent and 40 percent of the total nitrogen input to an estuary—depending in large part on the relative area of the estuary and its watershed (Nixon et al. 1996; Valigura et al. 2000). In estuaries where the ratio of the area of the estuary to the area of its watershed is greater than 0.2, direct atmospheric depositions usually

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution make up 20 percent or more of the total nitrogen loading (Valigura et al. 2000). Where the ratio of the estuarine area to the area of its watershed is less than 0.1, atmospheric deposition directly onto the water surface generally makes up less than 10 percent of the total nitrogen input (Valigura et al. 2000). For estuaries that have relatively large watersheds, the deposition of nitrogen from the atmosphere onto the landscape with subsequent runoff into the estuary is probably greater than the deposition of nitrogen directly onto the water surface. Unfortunately, the magnitude of this flux is poorly characterized for most estuaries. The deposition onto the landscape can be estimated for most watersheds, although the error associated with these estimates can be considerable due to inadequate monitoring and the difficulty with measuring dry deposition. The larger problem, however, is with determining what portion of the nitrogen deposition is retained in the landscape and what portion is exported to rivers and the coast. The two major approaches for making this determination are to use statistical models or to use process-based models on nitrogen retention in the watershed. In their application to estuaries, both approaches are quite recent and are relatively untested. There is an urgent need for further development and evaluation of these techniques; however, it appears that the statistical approaches have led to more reliable estimates, for reasons discussed below. Both the SPARROW model and regressions comparing nitrogen flux in rivers to sources of nitrogen across landscapes (used by the International SCOPE Nitrogen Project) represent examples of statistical approaches that appear to provide reliable estimates of the portion of the nitrogen deposition retained in the landscape versus what is exported to rivers and coastal areas. Jaworski et al. (1997) used a similar approach in the northeastern United States, comparing atmospheric deposition and riverine flux for 17 watersheds with relatively little agricultural activity or sewage inputs. This led to the conclusion that approximately 40 percent of the nitrogen deposition is exported from the landscape (correcting their analysis by assuming that dry deposition is equal to wet deposition), a value remarkably similar to the results from applying the SPARROW model at the national scale. By applying this result to other watersheds in the northeast, including those with agricultural activity, Jaworski et al. (1997) estimated that between 36 percent and 80 percent of the total nitrogen flux in rivers was originally derived from atmospheric deposition onto the landscape. Note that the riverine nitrogen fluxes were estimated at U.S. Geological Survey (USGS) gauging stations above the tidal portions of these rivers, and generally excluded the large urban influences at the river mouths. In another recent effort, a National Oceanic and Atmospheric Admin-

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution istration (NOAA)-sponsored project brought together researchers from around the United States to examine atmospheric deposition to coastal waters (Valigura et al. 2000). Valigura et al. (2000) summarized and compared the four different approaches included in the NOAA project, including a process-based model and an application of the statistical approach used by SPARROW. They report that, for 42 estuaries in the United States, atmospheric deposition onto the landscape contributed between 6 percent and 50 percent of the total nitrogen load to the receiving body. Jaworski et al (1997) and Valigura et al. (2000) give estimates in common for only one river/estuary—the Hudson-Raritan—and for this system, their estimates are similar to the statistical model results, but quite different from the process-based model estimates. Jaworski et al. (1997) estimate that 34 percent of the nitrogen flux in the Hudson comes from atmospheric deposition onto the landscape, after correction for the point source inputs from New York City (Hetling et al. 1996). In contrast, estimates from the process-based model indicated 9 percent of the nitrogen flux of the Hudson-Raritan total nitrogen load comes from nitrogen deposition onto the landscape. The statistical SPARROW model approach estimated the flux to the estuary from atmospheric deposition onto the watershed as 26 percent for this system. Great uncertainty about the importance of atmospheric deposition as a nitrogen source to specific estuaries may exist. However, there is little doubt that the relative importance of fossil-fuel combustion versus agricultural activity in controlling atmospheric deposition of nitrogen to estuaries depends both on the nature and extent of farming activities in the watershed and on the nature and extent of fossil-fuel combustion in the airsheds upwind of the watershed. In estuaries fed by watersheds with little agricultural activity but significant loads of atmospheric pollution (such as the Connecticut and Merrimack rivers and most of the northeastern United States), atmospheric deposition of nitrogen from fossil-fuel combustion can account for up to 90 percent or more of the nitrogen contributed by nonpoint sources. On the other hand, for watersheds such as the Mississippi Basin where agricultural activity is high and atmospheric pollution from fossil-fuel combustion is relatively low (Figure 5-18), agricultural sources dominate the fluxes of nitrogen. Interestingly, the major hot-spots of agricultural activity that dominate the nitrogen fluxes for the Mississippi and Gulf of Mexico appear to be far from the Gulf in Iowa, Illinois, Indiana, Minnesota, and Ohio (Goolsby et al. 1999). For many estuaries, both atmospheric deposition of nitrogen derived from fossil-fuel combustion (NOy) and nitrogen from agricultural sources are likely to be major contributors. For example, the model used by managers to estimate nitrogen inputs to Chesapeake Bay predicts that

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution FIGURE 5-18 A comparison of human-controlled inputs of nitrogen and nitrogen losses (kg N km−2 yr−1) as food exports and in riverine exports between the northeastern United States and the Mississippi River basin. Note that, on average, nitrogen is exported in foods and feedstocks from the Mississippi basin and imported to the northeastern United States.

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution agriculture contributes 59 percent of the nonpoint source inputs, NOy deposition onto the landscape is slightly less important (Magnien et al. 1995). The comparative analysis of Jaworski et al. (1997), on the other hand, suggests that atmospheric deposition is the dominant source of nitrogen from nonpoint sources in the major tributaries of Chesapeake Bay. Further study and analysis is necessary to determine whether Jaworski et al. (1997) have overestimated the importance of atmospheric deposition or whether Magnien et al. (1995) have underestimated it. However, the process-based model of nitrogen retention used by Magnien et al. (1995) has not been independently verified and is subject to large uncertainties (Boesch et al. 2000). Small changes in the assumed ability of forests to retain or export nitrogen from atmospheric deposition can lead to large changes in the relative importance of NOy deposition to the bay. As discussed above, there is great variation among forests in their ability to retain nitrogen from atmospheric deposition, and regional and large-scale analysis of nitrogen fluxes for the United States indicate a greater mobility of nitrogen from deposition (less retention) than is often found in small-scale watershed studies. Further, the model of Magnien et al. (1995) does not include some of the latest findings on nitrogen export from land, such as the large export of nitrogen in dissolved organic forms that was noted above (Campbell et al. 2000). A recent report from the Environmental Protection Agency (EPA) estimates that between 10 percent and 40 percent of the total nitrogen input to estuaries comes from atmospheric deposition, including deposition directly onto the water surface and onto the watershed (EPA 1999c). However, it must be stressed that very few of the individual studies upon which this conclusion is based had adequate methodologies for determining the input of nitrogen from atmospheric deposition, particularly the indirect input through atmospheric deposition onto the landscape with subsequent runoff into the estuary. Many of these studies have probably underestimated the importance of this pathway, and it seems likely that atmospheric deposition is a greater input to estuaries than suggested by the 1999 EPA report. OCEANIC WATERS AS A NUTRIENT SOURCE TO ESTUARIES AND COASTAL WATERS In addition to receiving nutrient inputs from land and from atmospheric deposition, estuaries can receive nutrients across their boundary with the ocean. This term is often ignored, but can be substantial. For example, Nixon et al. (1995) estimate that for total nutrient inputs to Narragansett Bay, 15 percent of the nitrogen and 40 percent of the phosphorus inputs come from offshore, oceanic sources; despite this, the net

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution flux of both nitrogen and phosphorus for Narragansett Bay is an export of these nutrients from the estuary to offshore waters (Nixon et al. 1996). On the other hand, Chesapeake Bay is a net importer of phosphorus from offshore ocean waters, although it too is a net exporter of nitrogen (Boynton et al. 1995; Nixon et al. 1996). The physical circulation pattern of an estuary is a major determinant in the importance of nutrient import to the estuary from offshore sources. Partially mixed estuaries (such as Chesapeake Bay) and fully mixed estuaries (such as Narragansett Bay) often import nutrients from offshore, whereas salt-wedge estuaries (such as the southwest pass of the Mississippi River and Oslo Fjiord) and hypersaline estuaries (such as portions of Shark Bay, Australia) do not (Howarth et al. 1995). Offshore waters on the continental shelf can themselves receive nutrients from several sources, including deep ocean water, river and sewage inputs from land, and direct deposition from the atmosphere (Nixon et al. 1996; Prospero et al. 1996; Howarth 1998). The relative importance of these sources varies among the coastal waters of the United States, in part because of differences in ocean circulation patterns (particularly advection of water from the deep ocean—water that is extremely high in nutrients—onto the continental shelf). For most of the continental shelf area of the United States, this advection of water is the dominant nutrient input. However, input from the Mississippi River is the dominant source for the Gulf of Mexico. Human activity has tended to greatly increase inputs of nitrogen from rivers and atmospheric deposition, but has had no impact on the advection of water from the deep ocean onto the continental shelf. Consequently, human activity has almost tripled nitrogen input to the Gulf of Mexico, but has increased nitrogen inputs to the waters on the continental shelf of the northeastern United States by only 28 percent (Table 5-1). Of course, much of this input in the northeastern United States is concentrated in the plumes of a few rivers, such as that of the Hudson River, and these waters may therefore be experiencing eutrophication (Howarth 1998). Rate of Change of Nutrient Inputs to the Coast Historical data on fluxes of total nitrogen in rivers are rare, but data for trends in nitrate concentrations are available for many rivers going back to the early 1900s. Since human activity preferentially mobilizes nitrate over other forms of nitrogen in rivers (Howarth et al. 1996), these historical nitrate data are valuable in tracking the effects of humans on nitrogen fluxes to the coast. For the Mississippi River, the nitrate flux to the Gulf of Mexico is now some three-fold larger than 30 years ago, and most of this increase occurred between 1970 and 1983 (Figure 5-19;

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution TABLE 5-1   Nitrogen Sources (Tg yr-1) North Atlantic Continental Shelves Rivers and Estuaries Direct Atmospheric Deposition Deep Ocean Increase Due to Humans (%) North Canada rivers 0.16 (0.16) 0.10 (0.03) 0.77 7 St. Lawrence basin 0.34 (0.11) 0.13 (0.01) 1.26 25 Northeast coast of the United States 0.27 (0.03) 0.21 (0.01) 1.54 28 Southeast coast of the United States 0.13 (0.03) 0.06 (0.01) 1.36 11 Gulf of Mexico 2.10 (0.50) 0.28 (0.03) 0.14 275 North Sea and Northwest Europe 0.97 (0.14) 0.64 (0.02) 1.32 98 Southwest European coast 0.11 (0.04) 0.03 (0.001) 0.20 40 TABLE 5-1 Sources of nitrogen to the continental shelves of the temperate zone portions of the North Atlantic Ocean. Flux from rivers and estuaries is the direct input of rivers that discharge onto the continental shelf, minus nitrogen consumed in estuaries. Atmospheric deposition estimates are those directly onto the waters of the continental shelf and do not include deposition onto the landscape (which is part of the flux from rivers and estuaries). The flux from the deep ocean represents the advection of nitrate-rich deep Atlantic water onto the continental shelf. Data for modern values are means reported by Nixon et al. (1996). Pristine values as outlined by Nixon et al. (1996) for their treatment of modern estimates, but with data for pristine river fluxes from Howarth et al. (1996) and for pristine values of deposition from Prospero et al. (1996). “Increase due to humans” is the percentage comparison of total modern inputs compared to pristine inputs. Fluxes from the deep ocean are assumed not to have been affected by human activities (modified from Howarth 1998). Goolsby et al. 1999). Similarly, nitrate fluxes in many rivers in the northeastern United States have increased two- to three-fold or more since 1960, with much of this increase occurring between 1965 and 1980 (Figure 5-20; Jaworski et al. 1997). Interestingly, most of the increase in nitrate in the Mississippi River was due to increased use of nitrogen fertilizer (Goolsby et al. 1999), whereas most of the increase in nitrate in the northeastern rivers was due to increased nitrogen deposition from the atmosphere onto the landscape, with the nitrogen originating from fossil-fuel combustion (Jaworski et al. 1997). The increase in nitrate flux in the

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution FIGURE 5-19 Bar chart showing the annual flux of nitrogen as nitrate (NO3) from the Mississippi River basin to the Gulf of Mexico, indicating significant increases beginning in the late 1970s (modified from Goolsby et al. 1999). FIGURE 5-20 Flux of nitrate nitrogen from five major rivers in the northeastern United States from the early 1900s to 1994 (modified from Jaworski et al. 1997).

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution northeastern rivers during the 1960s and 1970s, and its stabilization since then, closely parallels the trend in human inputs of nitrogen to the landscape during that time (Jaworski et al. 1997). In contrast to nitrogen, phosphorus fluxes to estuaries have often changed little over the past several decades. For the Mississippi River, data on total phosphorus flux are only available since the early 1970s, but there has been no statistically significant change since then (Goolsby et al. 1999). Smith et al. (1987) used data from 300 river locations throughout the United States to compare water quality trends from 1974 to 1981. Many rivers showed no trend during that time; rivers that had a trend in total phosphorus flux were equally divided between those that showed an increase and those that showed a decrease. Where total phosphorus fluxes increased, it was generally attributable to increased use of phosphorus fertilizer in the watershed. Decreases in total phosphorus fluxes were generally a result of point source reductions (Smith et al. 1987). Smith et al. (1987) also analyzed the national river data for trends in nitrate flux from 1974 to 1981. For nitrate, most rivers showed a marked increase in flux during that time, particularly for rivers in the eastern United States. This increased nitrate flux was attributed both to agricultural activity and to nitrogen deposition (Smith et al. 1987). IMPLICATIONS FOR ACHIEVING SOURCE REDUCTIONS Human activity has an enormous impact on the cycling of nutrients and especially on the movement of such nutrients as nitrogen and phosphorus into estuaries and other coastal waters. Although much effort has been made in the United States to improve control of point sources of pollution, nonpoint sources as urban runoff, agricultural runoff (particularly from animal feeding operations), and atmospheric deposition are generally of greater concern in terms of impact on nutrient enrichment and eutrophication of coastal waters. While sewage inputs dominate in some estuaries, nonpoint sources dominate nationally. Insufficient effort has been expended on controlling nonpoint sources of nitrogen and phosphorus, and there are few comprehensive plans for managing nutrient enrichment of the nation’s coastal waters, particularly from nonpoint sources. Efforts to manage nonpoint and point sources of nitrogen and phosphorus are needed to reduce adverse impacts of nutrient over-enrichment in the nation’s rivers, lakes, and coastal waters. There is evidence that both atmospheric deposition of nitrogen from fossil-fuel combustion and agricultural sources of nitrogen contribute nitrogen to coastal waters. The relative importance of these varies among estuaries, but recent evidence indicates that the amount of nitrogen from

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution deposition has been historically underestimated as an input to many estuaries, particularly by the indirect pathway of nitrogen deposited onto the landscape and then exported to the estuary. Recent evidence also indicates that per unit input to the landscape, nitrogen from fossil-fuel combustion is more important than nitrogen from fertilizer and, in turn, contributes disproportionately in the input of nitrogen to coastal waters. Much uncertainly remains regarding the fluxes of nitrogen from the atmosphere to the landscape and to estuaries, and this is a critically important research priority. Although understanding some details regarding the atmospheric transport and fate of biologically available nitrogen will require additional research, the significant role atmospheric deposition of nitrogen plays in nutrient over-enrichment in some regions is clear. Addressing this component of the problem will require coordinated efforts over many states, clearly dictating a federal role in the effort. The regional nature of the atmospheric component of nitrogen loading argues that nutrient management should be a significant component of efforts to reduce air pollution and should be a key consideration during re-authorization of the Clean Air Act. In general, sources of nutrients to estuaries have been poorly characterized, and in some cases sources have been mistakenly characterized because some land-use export-coefficient models used for characterization are inadequately verified. There are currently no easy-to-use and reliable methods for the manager of an estuary to determine the sources of nutrients flowing into that estuary. As will be discussed in Chapter 8, enhanced and coordinated monitoring efforts will be a key component of any local, regional, or national effort to reduce the impacts of nutrient over-enrichment. Some critical questions related to understanding the sources of nutrients most affecting eutrophication and other impacts of nutrient over-enrichment remain unanswered. For instance, nitrogen deposition and fate in urban and suburban areas is poorly known, and wet nitrogen deposition in coastal areas is poorly understood. There is only a limited understanding of dry deposition in any environment, and understanding this in coastal areas and over water is challenging. Research efforts to expand understanding of atmospheric deposition of nitrogen should be expanded. Changes in agricultural production systems are concentrating large amounts of nutrients in localized areas, thereby increasing the risk of nutrient leakage to the environment. Most of this concentration is associated with animal feedlots and with the long-distance transport of feedstocks. Changes in farm practices are driven by economics, and this concentration and long-range transport provide economic advantages to

OCR for page 113
Clean Coastal Waters: Understanding and Reducing the Effects of Nutrient Pollution the producers; the larger costs, such as the external cost of nutrient exports to estuaries, remain unaddressed. As is discussed further in Chapter 9, a balanced and cost-effective nutrient management strategy will require an understanding of both the relative importance of various sources of nutrients, and the economic costs associated with reducing the loads attributable to each.