32N
UTR
IEN
TS
Nutrients from atmospheric and urban sources, fertilization, andlivestock wastes can contribute to excessive algal growth in streams
Jeffrey D. Martin
Dennis A. Wentz
Kevin D. Richards
Ian R. Waite
Anonymous
33
NU
TRIE
NTSNutrients
Human activities—including agricultural and urban uses of fertilizer, agricultural use of manure, and
combustion of fossil fuels—have caused widespread increases of nitrate in shallow ground water and total
nitrogen and total phosphorus in streams across the Nation.
Nitrate did not pose a national health risk for residents whose drinking water came from streams or from
major aquifers buried relatively deep beneath the land surface. Some concerns were evident in 4 of the 33
major aquifers sampled, where nitrate concentrations in more than 15 percent of each aquifer exceeded the
USEPA drinking-water standard. The most prevalent nitrate contamination of ground water, however, was
found in relatively shallow ground water in rural areas where the water commonly is used for domestic supply.
In more than one-half of sampled streams, concentrations of total nitrogen and total phosphorus were
above national background concentrations. Elevated phosphorus levels, in particular, can lead to excessive
plant growth (eutrophication) in freshwater environments; in more than one-half of sampled streams and in
three-fourths of agricultural and urban streams, average annual concentrations of total phosphorus exceeded
the USEPA desired goal for prevention of nuisance plant growth. The highest total nitrogen and total
phosphorus concentrations were found in small streams draining watersheds with large proportions of
agricultural or urban land. Long-term monitoring of streams indicates that programs to control point-source
discharges of phosphorus and ammonia have been effective, despite population increases in most
metropolitan areas. Phosphorus concentrations have decreased as a result of reductions in the use of
phosphate detergents and in the amount of phosphorus discharged from upgraded wastewater treatment
plants. Improved wastewater treatment, which converts ammonia to nitrate, generally has resulted in a
decrease in ammonia concentrations and an increase in nitrate concentrations in streams. Thus, concentrations
of total nitrogen downstream from metropolitan areas have changed little during the past 20 years, although
toxicity to fish has decreased with decreasing ammonia levels.
Results from NAWQA studies have shown regional and seasonal differences in nutrient concentrations
that can be explained largely by the amounts and timing of fertilizer and manure applications and by the
variety of soils, geology, climate, and land- and water-management practices across the Nation. Recognition
of these differences is important for efficient protection of ground water needed for drinking and for curbing
eutrophication of surface water. ◗
34N
UTR
IEN
TS
Background concentrations ofnutrients are low in streams andground waterBackground concentrations of nutrients were estimated on the basis of samplescollected from undeveloped areas considered to be minimally affected byagriculture, urbanization, and associated land uses. Background concentrationsin undeveloped areas are controlled primarily by naturally occurring mineralsand by biological activity in soil and streambed sediment. Chemical propertiesof the atmosphere and rainwater, which can reflect human-related fuelcombustion and other activities both within and external to a watershed, canincrease background concentrations.
In this report, national background nutrient concentrations includeatmospheric contributions and are summarized in the following table. Waterswith concentrations of nutrients greater than the national backgroundconcentrations are considered to have been affected by human activities ina variety of land-use settings.
WHAT WAS MEASURED…
Nitrate is the primary form of nitrogendissolved in streams and ground water. Inthis report, nitrate refers to the sum ofnitrate plus nitrite, as reported by theUSGS laboratory. Nitrite concentrationscommonly were less than the laboratorydetection level of 0.01 milligrams per liter(mg/L), making its contribution to nitrateplus nitrite negligible.
Ammonia is a dissolved form of nitrogenthat is less common than nitrate. Asmeasured by the USGS laboratory, totalammonia includes ammonium ion andun-ionized ammonia. The latter is usuallya minor component of ammonia at pHscommonly observed in streams andground water.
Total nitrogen includes nitrate, nitrite,ammonia, and organic nitrogen. Nitrite isgenerally unstable in surface water andcontributes little to the total nitrogen.Organic nitrogen (mostly from plantmaterial or organic contaminants) canexist in considerable proportions andcontribute substantially to total nitrogen instreams.
Phosphates are the most common formsof phosphorus found in natural waters.Compared to nitrate, phosphates dissolveless readily. They are not mobile in soilwater and ground water because theytend to attach to soil and aquifer particles.They can have a significant impact,however, because eroded soil cantransport considerable amounts ofattached phosphates to streams andlakes. Orthophosphate typicallyconstitutes the majority of dissolvedphosphates, which can be readilyassimilated by aquatic plants and promoteeutrophication.
Total phosphorus includes phosphates,as well as all other phosphorus forms.Dissolved phosphates and particulateorganic phosphorus (mostly from plantmaterial) are the main components oftotal phosphorus.
For brevity, all forms of nutrientsdiscussed in this report representconcentrations as either nitrogen orphosphorus. For example, a nitrateconcentration expressed as 10 mg/Lrefers to a nitrate concentration of10 mg/L as nitrogen.
ESTIMATES OF NATIONAL BACKGROUND NUTRIENT CONCENTRATIONS
Nutrient Background concentration
(mg/L)
Total nitrogen in streams[Data from 28 watersheds in first 20 Study Units]
1.0
Nitrate in streams (26) 0.6
Ammonia in streams (26) 0.1
Nitrate in shallow ground water (27) 2.0
Total phosphorus in streams (26) 0.1
Orthophosphate in shallow ground water[Data from 47 wells in first 20 Study Units]
0.02
Background nutrient concentrations can vary considerably from region toregion, or even within watersheds, because of differences in hydrology and innaturally occurring nutrient levels in soils, rocks, and the atmosphere. The dataanalyzed for this report are insufficient to define background nutrientconcentrations on a regional basis. Thus, all available data from undevelopedareas were combined to derive national background concentrations. Thenational background concentrations are higher than most concentrationsmeasured in relatively undeveloped areas across the Nation and may not beapplicable for use in regional or local analyses.
35
NU
TRIE
NTS
Human activities have increasednutrients above backgroundconcentrationsEffects of human activities on nutrients were assessed by comparingconcentrations in streams and ground water to national backgroundconcentrations. Waters with nutrient concentrations above background arereferred to as “enriched” in this report. Fifty-seven percent of sampled streamswere enriched with total phosphorus on the basis of average annual totalphosphorus concentrations exceeding national background concentrations.Similarly, 61 percent of sampled streams were enriched with total nitrogen andnitrate, but only 23 percent of sampled streams were enriched with ammonia.Only 1 of 28 relatively undisturbed forested or rangeland streams had averageannual concentrations of total phosphorus or total nitrogen above nationalbackground concentrations. Most of the streams that were enriched withnutrients drained areas of agricultural and (or) urban land.
In 53 percent of shallow ground-water studies in agricultural and urbanareas, median nitrate concentrations were above the national backgroundconcentration. Median nitrate concentrations were above background in only3 of 33 major aquifers studied. Those three aquifers were beneath agriculturalareas in three different Study Units.
In most cases, enrichment of streams with nutrients occurred in smallwatersheds and (or) regions dominated by agricultural or urban land use.Effects of human activities were found in shallow ground beneathagricultural and urban areas throughout the Nation, but not in many ofthe major aquifers sampled.
The Otter Tail River, Minnesota (above), which supports a healthy growthof wild rice, and Fir Creek, Oregon (right), which contributes to Portland’sdrinking-water supply, are examples of streams with low nutrientconcentrations.
STANDARDS AND GUIDELINES FORPROTECTING WATER QUALITY
The USEPA has established a Federaldrinking-water standard or MaximumContaminant Level (MCL) of 10 mg/L fornitrate.(28) An MCL is a concentrationabove which adverse human healtheffects may occur.
The USEPA(29) has established criteria forun-ionized ammonia in surface waterbecause of its toxicity to fish. The chroniccriteria vary from 0.07 to 2.1 mg/L of totalammonia for pHs of 6.5–9.0 and watertemperatures of 0–30 °C.
National criteria have not been establishedfor concentrations of dissolvedphosphates in streams or ground water.
National criteria have not beenestablished for total phosphorus or totalnitrogen in streams. The USEPA hasestablished a desired goal of 0.1 mg/Ltotal phosphorus for the prevention ofnuisance plant growth in streams andother flowing waters not dischargingdirectly to lakes or impoundments.(29)
Jeffr
ey D
. Sto
ner
Denn
is A
. Wen
tz
36N
UTR
IEN
TS
Nutrients are a potential concern forhuman healthFrom a national perspective, nitrate contamination did not pose a healthrisk for residents who drank water from major aquifers buried relativelydeep below land surface. Some concerns were evident, however, in 4 of the33 major aquifers sampled. In each of these four aquifers, nitrate concentra-tions in more than 15 percent of samples exceeded the USEPA drinking-waterstandard. All four aquifers are relatively shallow, in agricultural areas, andcomposed of sand and gravel that is vulnerable to contamination by landapplication of fertilizers. In nearly one-half of the major aquifers sampled,water from at least one well, out of 20 to 30 wells, exceeded the drinking-water standard. Many of the major aquifers exhibiting high nitrateconcentrations were used for rural domestic water supply.
About 15 percent of all shallow ground water sampled beneath agriculturaland urban areas exceeded the drinking-water standard for nitrate. Thepresence of elevated nitrate concentrations in shallow ground water raisesconcerns, particularly where these vulnerable aquifers contain deeperwells used for rural domestic water supply. Contamination of shallowground water may be a warning to alert populations to potential futurerisks from consumption of water from deeper wells in these aquifers.
LowerLowerSusquehannaSusquehanna
River BasinRiver Basin
See p. 31 for more information about this map
A national ranking of NITRATE concentrations in major aquifers
Bold outline indicates median values greater than background concentration (2 milligrams per liter)
Percentage of samples exceeding drinking-water standard for nitrate (10 milligrams per liter)—Each circle represents a major aquifer
Background concentration
0 samples exceed standardLess than 15 (but at least 1 sample)Greater than 15
Red Riverof the North Basin
Central Nebraska Basins
LowerSusquehanna
River Basin
SanJoaquin-TulareBasins
37
NU
TRIE
NTS
Because of its proximity to the land surface, shallowground water is younger and more vulnerable tocontamination from human activities than deep groundwater. Major aquifers generally are buried deep beneaththe land surface, where they are protected by layers ofclay or rock that are relatively impermeable and thatimpede downward movement of water and nitrate.Ground water in major aquifers sampled by the NAWQAProgram can be tens to hundreds of years old and,therefore, minimally affected by recent land-use practices.
Geology and hydrology control the movement ofcontaminated water from shallow to deep systems, andunderstanding their effects allows an anticipation ofpossible areas of concern in major aquifers. For example,elevated concentrations of nitrate were detected in a majoraquifer in the Lower Susquehanna River Basin (medianconcentration about 7 mg/L) because karst (weatheredcarbonate rock) contains open conduits that allow rapiddownward movement of water and chemicals. Medianconcentrations of nitrate also were high in the alluvialaquifer of the Central Nebraska Basins (about 6 mg/L)and in the alluvial fans of the San Joaquin-Tulare Basins(about 5 mg/L). Extensive pumping causes verticalmixing of ground water in these relatively permeable sandand gravel aquifers. In contrast, the median concentrationof nitrate was low (0.4 mg/L) in a surficial sand andgravel aquifer in the Red River of the North Basin;however, 15 percent of the samples exceeded thedrinking-water standard. This last example demonstratesthe complex effects that local geology and nitrate sourcescan have on nitrate contamination.
For large rivers and most of the smaller streamssampled, nitrate is not a drinking-water issue. Thisconclusion is based on comparisons of average annual andaverage monthly concentrations to the drinking-waterstandard. In only two of all sampled rivers and streamsdid the average annual concentration of nitrate exceed thedrinking-water standard. These two streams drained smallagricultural watersheds in the Lower Susquehanna Riverand Willamette Basins, and neither was used to supplydrinking water.
The percentage of ground-water samples with concentrationsof nitrate exceeding the drinking-water standard of 10 mg/Ldecreases as depth to water increases. Mixing of shallow groundwater with deeper, uncontaminated water and increased thicknessof protective, impermeable geologic materials with depth may helpexplain this relation.
From a national perspective, nitrate did notpose a health risk for residents who drankwater from major aquifers buried relativelydeep below the land surface. People drinkingground water from shallow wells in vulnerablegeologic settings (sand, gravel, or karst) inrural agricultural areas, however, are at riskof exposure to nitrate contamination.
0-100 100-200 Greater than 200
20
15
10
5
0
DEPTH TO GROUND WATER, in feet
PE
RC
EN
TA
GE
OF
SA
MP
LE
S E
XC
EE
DIN
G T
HE
NIT
RA
TE
DR
INK
ING
-WA
TE
R S
TA
ND
AR
D
1,345 80 33
Number of samples
0 percent
Kevi
n F.
Denn
ehy
38N
UTR
IEN
TS
Nutrients are a potential concern foraquatic lifeAverage annual and average monthly concentrations of un-ionizedammonia did not exceed USEPA aquatic-life criteria for most streamssampled. Exceptions include an agricultural stream affected by upstreamwastewater treatment plant effluent in the San Joaquin-Tulare Basins and twourban streams, one in the South Platte River Basin and another in the NevadaBasin and Range. The urban streams, which are in relatively arid climates andexceeded the criteria year-round, also received effluent from wastewatertreatment plants.
Eutrophic conditions were noted in some streams across the Nationbecause of elevated concentrations of nutrients. For example, the averageannual concentration of total phosphorus in 57 percent of all streams sampledwas greater than the USEPA desired goal of 0.1 mg/L for preventing nuisanceplant growth in streams. In addition, about 75 percent of agricultural and urbanstreams exceeded this goal. It is difficult and premature, however, toattempt a national summary of eutrophication effects because of limitedavailable methodologies for deriving criteria based only on nutrientconcentrations. Moreover, the uncertainty regarding how nutrientcontamination of streams harms aquatic life and affects nuisance plant growthdoes not lessen the value of accurate information for management of ourNation’s streams. A strategy, spearheaded by the USEPA in collaboration withother Federal and State agencies, is underway to evaluate excessive aquaticplant growth, such as algae, in surface water. This strategy includes anunderstanding of stream nutrient dynamics, stream habitat (including shadingand temperature), turbidity, and algal-growth processes.
Un-ionized ammonia concentrations exceededUSEPA criteria for protection of aquatic life in LasVegas Wash, Nevada, downstream from wastewatertreatment plant discharges. Concentrations in allsamples collected from April 1993 to April 1995exceeded the criteria. The median ammoniaconcentration downstream from wastewatertreatment plant discharges was more than 100 timesthe median value upstream from the discharges.Downstream ammonia concentrations decreasedfivefold during 1996–97, following full implementationof tertiary treatment of wastewater, and USEPAcriteria probably are no longer exceeded at this site.
Nitrogen and phosphorus have differenteffects on aquatic plant growth infreshwater and saltwater. Eutrophicationof freshwater streams generally resultsfrom high phosphorus concentrations. Incontrast, excess nitrogen, and nitrate inparticular, can lead to algal blooms incoastal waters. The USEPA suggests adesired goal of 0.1 mg/L total phosphorusfor freshwater streams, but there are nonational criteria established for nitrogenconcentrations to control excessiveaquatic plant growth in coastal bays andestuaries.
115°30'
115°
36°30'
36°
Downstream
Upstream
0.01
0.1
1
10
UN
-IO
NIZ
ED
AM
MO
NIA
CO
NC
EN
TR
AT
ION
, in
mill
igra
ms
per
lite
r
LAS VEGASVALLEYAREA
0 10 20 MILES
0 10 20 KILOMETERS
Range, barren, or other
Forest
Irrigated agriculture
Urban
Land use
Open water
Study-basin boundary
39
NU
TRIE
NTS
In April 1994, total phosphorus concentrations in theSouth Platte River exceeded the USEPA desired goal forpreventing plant nuisances (0.1 mg/L) in a 150-mile reachdownstream from Denver, Colorado.
“As part of the Clean Water ActionPlan, the Vice President calledupon USEPA to acceleratedevelopment of nutrient water-quality criteria for beneficialecological uses in every geo-graphic region in the country.We will work with States and tribesto develop a methodology forderiving criteria, as well asdeveloping criteria where dataare available, for nitrogen andphosphorus runoff for lakes, rivers,and estuaries by the year 2000.We intend to develop such criteriaon a regional basis usingscientifically defensible data andanalysis of nutrients, such asthose available from the USGS.We will assist States and tribes inadopting numerical nutrientcriteria as water-quality standardsby the end of 2003.”
Robert Cantilli, Nutrients CriteriaCoordinator, USEPA
Large amounts of nitrate enter theChesapeake Bay from the SusquehannaRiver. Nitrate concentrations in theSusquehanna River at Harrisburg,Pennsylvania, generally were less than 2mg/L. However, these concentrations,when multiplied by the large flows of theSusquehanna River, contribute largeamounts of nitrate to Chesapeake Bay(especially compared with other riversentering the bay) and provide enoughnitrate to stimulate algal growth andaffect the bay ecosystem.
4000 100 200 300
3
0
1
2
TO
TA
L P
HO
SP
HO
RU
S,
in m
illig
ram
s p
er li
ter
Den
ver
Hen
ders
on
Ker
sey
Bal
zac
Jule
sbur
g
Nor
th P
latte
RIVER MILES
Larry
J. P
ucke
tt
40N
UTR
IEN
TS
Nutrient conditions differ byland useThe highest nitrogen and phosphorus concentrationsgenerally were found in agricultural and urban streams.Nutrient concentrations in areas of mixed land use werelower than in agricultural or urban areas but were higherthan in undeveloped areas. Orthophosphate concentrationsin ground water typically were so low that relations toland use are not definitive.
Except for nitrogen in agricultural areas, nutrientconcentrations in streams generally were higher thanthose in shallow ground water, regardless of land use. Inagricultural areas, nitrate concentrations in shallowground water typically were higher than total nitrogenconcentrations in streams, although exceptions occurred inareas where soil characteristics restrict downwardmovement of water.
Regional patterns in nutrient concentrations can beuseful for determining areas of the Nation whereenvironmental settings deserve the greatest concern andattention. The discussion and maps on pages 41–45 focuson geographic patterns in relation to land use and nutrientinputs. Methods used to construct the maps are explainedon page 31.
The highest median concentrations of total nitrogen instreams and nitrate in shallow ground water occurred inagricultural basins (left). The highest median total phosphorusconcentrations were found in urban streams, butorthophosphate concentrations in ground water were loweverywhere (right).
Total phosphorus in streams and orthophosphate in ground water
CO
NC
EN
TR
AT
ION
, in
mill
igra
ms
per
lite
r
0
2.5
2.0
1.5
1.0
.5
Undeveloped Agricultural Urban Mixed
DOMINANT LAND USE
Total nitrogen in streams and nitrate in ground water
CO
NC
EN
TR
AT
ION
, in
mill
igra
ms
per
lite
r
0
15
12
9
6
3
Undeveloped Agricultural Urban Mixed
DOMINANT LAND USE
Median
Outlier
Includes 90 percent of sites
18Number of sites
28 4 75 36 22 13 87 33Number of sites
28 4 75 36 22 13 87 33
Streams Shallow ground water
Streams and rivers Major aquifers
41
NU
TRIE
NTS
NITRATE IN SHALLOW GROUND WATERHigh concentrations of nitrate in shallow ground water were widespreadand strongly related to agricultural land use, but there were no apparentregional patterns. Based on comparisons with background concentrations,human activities have increased nitrate concentrations in ground water for abouttwo-thirds of agricultural areas studied, compared to about one-third of urbanareas. Median nitrate concentrations for 13 of 36 agricultural areas were greaterthan 5 mg/L and ranked among the highest of all shallow ground-water studies.Only 1 of 13 urban areas fell into this high-concentration group. It is likely thatnitrogen sources in urban areas are relatively localized when compared with thegenerally more intensive and widespread use of fertilizers on cropland. Also, theimpervious surfaces typically found in urban areas generally result in surfacerunoff of nutrient-laden water, rather than seepage to ground water.
Tampa •
Dallas-Ft Worth •
• Atlanta • Albuquerque
• Las Vegas
Virginia
Beach
• Indianapolis
• Reno-Sparks• Denver
• Harrisburg
• SpringfieldAlbany •
• Portland
Bold outline indicates median values greater than background concentration (2 milligrams per liter)
Median concentration of nitrate—in milligrams per liter.Each circle represents a ground-water study
Background concentration
Lowest (less than 0.5)Medium (0.5 to 5.0)Highest (greater than 5.0)
A national ranking of NITRATE concentrations in shallow ground water
See p. 31 for more information about these maps
Average annual total nitrogen input—in pounds per acre, by county, for 1991–94. Inputs are from fertilizer, manure, and the atmosphere
Lowest (less than 6)Medium (6 to 25)Highest (greater than 25)
Urban areasNitrate concentrations in urban areas generally were lower than in agricultural areas, but 40 percent of urban areas had median values above the national background concentration.
Agricultural areasNitrate concentrations in agricultural areas were among the highest measured, but not all agricultural areas had median values above the national background concentration.
42N
UTR
IEN
TS
NITROGEN IN STREAMSAverage annual concentrations of total nitrogen in about 50 percent ofagricultural streams ranked among the highest of all streams sampled in thefirst 20 Study Units, and concentrations in about 36 percent of urban streamswere among the highest measured. In contrast, total nitrogen levels in streamsdraining relatively undeveloped, forested watersheds (not shown on thenational maps) ranked among the lowest of all streams sampled.
High concentrations of nitrogen in agricultural streams correlated withnitrogen inputs from fertilizers and manure used for crops and from livestockwastes. The Upper Midwest is a notable region of high nitrogen levels inagricultural streams; however, there are also many such examples in the Westand East. High nitrogen levels in urban streams probably were related tonitrogen introduced from fertilizers applied to suburban lawns and golfcourses, emissions from automobiles and electric powerplants, and effluentfrom sewage treatment facilities.
Streams and large rivers that drain areas of mixed land use had averageannual concentrations of total nitrogen at various levels across the Nation, withno apparent regional pattern. The highest average annual concentrationsoccurred in watersheds with the highest nitrogen inputs and in riversdownstream from major metropolitan areas. The lowest concentrations instreams draining areas of mixed land use were for watersheds havingconsiderable proportions of forest.
For all streams sampled, the highest concentrations of total nitrogencorresponded to broad patterns of nitrogen inputs from agricultural andurban areas. Coastal waters near such areas of high nitrogen input are atgreatest risk of eutrophication.
During high spring streamflows following fertilizerapplication in the northern Willamette Basin, nitrateconcentrations increased in proportion to thepercentage of drainage area in agriculture. Nutrientconcentrations also were found to increase with thepercentage of drainage area in agriculture forwatersheds in the Ozark Plateaus, Potomac River Basin,and Trinity River Basin.
0 1000 50
PERCENTAGE OF DRAINAGE AREAIN AGRICULTURE
0.01
100
0.02
0.050.10.2
0.512
51020
50
NIT
RA
TE
CO
NC
EN
TR
AT
ION
,in
mill
igra
ms
per
lite
r
Minimum Reporting Level (MRL)
Open symbol representsvalue less than the MRL
43
NU
TRIE
NTS
OzarkPlateaus
Potomac River Basin
TrinityRiverBasin
WillametteBasin
Dallas-Ft Worth •
• Atlanta
• Tallahassee
• Las Vegas
• Norwalk
• Indianapolis
• Portland
• Denver
• Harrisburg
• Washington D.C.
Albany •
• Albuquerque
Milwaukee •
Average annual concentration of total nitrogen—in milligrams per liter
Lowest (less than 0.6)Medium (0.6 to 2.9)Highest (greater than 2.9)
Average annual total nitrogen input—in pounds per acre, by county, for 1991–94. Inputs are from fertilizer, manure, and the atmosphere
A national ranking of TOTAL NITROGEN concentrations in streams
Lowest (less than 6)Medium (6 to 25)Highest (greater than 25)
See p. 31 for more information about these maps
Agricultural streamsTotal nitrogen concentrations in agricultural streams were among the highest measured and generally correlated with nonpoint nitrogen inputs across the Nation.
Urban streamsThe highest total nitrogen concentrations in urban streams typically were in densely populated areas in relatively arid Western basins and in the Northeast.
Rivers and streams with mixed land useTotal nitrogen concentrations generally correlated with nonpoint nitrogen inputs, but levels in large rivers downstream from major metropolitan areas were among the highest measured.
44N
UTR
IEN
TS
PHOSPHORUS IN STREAMSIn most streams draining agricultural, urban, or mixed land use,concentrations of total phosphorus were greater than backgroundconcentrations and the USEPA desired goal for preventing nuisance plantgrowth in streams. About one-half of urban streams had average annualconcentrations of total phosphorus that ranked among the highest measured inthe first 20 Study Units. The highest average annual concentrations of totalphosphorus were in streams near metropolitan areas in the semiarid westernand southwestern regions of the Nation. Examples include the Santa Fe Riverdownstream from Santa Fe, New Mexico; Las Vegas Wash downstream fromLas Vegas, Nevada; and the South Platte River downstream from Denver,Colorado. In these areas, discharges from wastewater treatment plants can be asignificant proportion of the streamflow.
The broad geographical pattern observed for concentrations of totalnitrogen in streams also holds true for concentrations of total phosphorus.This comparability is not surprising because proportions of nitrogen andphosphorus fertilizer used across the Nation are similar. Elevatedphosphorus concentrations in agricultural streams can also come fromlivestock waste, such as in Prairie Creek in the Central Nebraska Basins, orfrom poultry wastes, such as in streams of the Apalachicola-Chattahoochee-Flint River Basin and the Ozark Plateaus. In addition, agricultural sites can beaffected by effluent from upstream wastewater treatment plants, such as inTurlock Irrigation District Lateral 5 in the San Joaquin-Tulare Basins. Somehigh concentrations of phosphorus can occur naturally. For example,concentrations of phosphorus in the Pembina River of the Red River of theNorth Basin were high because most agricultural land in this area includesphosphorus-rich soils in relatively steep, easily eroded terrain; highconcentrations of phosphorus in some streams in the Albemarle-PamlicoDrainage were derived from ground water in contact with phosphate minerals.
Concentrations of total nitrogen and total phosphorus typically were low inlarge rivers that drain areas of mixed land use. Examples include the AltamahaRiver, Georgia; Connecticut River, Connecticut; Menominee River, Wisconsin;and Upper Snake River, Idaho. In these large watersheds, streams drainingforested and other relatively undeveloped land dilute nutrient-rich runoff fromagricultural and urban areas. A few large rivers (such as the South Fork of thePalouse River, Washington, and the Trinity River, Texas) had extremely highaverage annual concentrations of total nitrogen and (or) total phosphorus thatcan be attributed primarily to the effect of upstream discharges of wastewatereffluent. Large watersheds dominated by agricultural land, such as in the GreatPlains and Upper Midwest, also exhibited concentrations of total phosphorusthat ranked among the highest measured.
Effluent can make up a substantial partof the streamflow in some areas. Forexample, wastewater treatment plantsannually contribute about 69 percent (andat times 100 percent) of the flow in theSouth Platte River downstream fromDenver, Colorado. About 1,200 tons ofphosphorus enter the South Platte RiverBasin every year from wastewatertreatment plants.
Phosphorus transport in watersheds iscomplex. Phosphorus commonly attachesto soil particles and either remains closeto application areas or moves to streamsprimarily by soil erosion. As phosphoruslevels increase in some soils, a largeramount is available for transport in thedissolved form. This amount variesaccording to the phosphorus adsorptioncapacity of the soil. Areas with high levelsof soil phosphorus relative to their soiladsorption capacity export relatively largeamounts of phosphorus to streams.
Kevin F. Dennehy
45
NU
TRIE
NTS
Red Riverof the North Basin
OzarkPlateaus
Central Nebraska Basins
Apalachicola-Chattahoochee-Flint River Basin
Albemarle-PamlicoDrainage
SanJoaquin-
TulareBasins
Dallas-Ft Worth •
• Atlanta
• Tallahassee
• Las Vegas
• Norwalk
• Indianapolis
• Portland
• Denver
• Harrisburg
• Washington D.C.
Albany •
Milwaukee •
• Albuquerque
See p. 31 for more information about these mapsSee p. 31 for more information about these maps
Average annual total phosphorus input—in pounds per acre, by county, for 1991–94. Inputs are from fertilizer and manure
Lowest (less than 2)Medium (2 to 5)Highest (greater than 5)
See p. 31 for more information about these maps
A national ranking of TOTAL PHOSPHORUS concentrations in streams
Bold outline indicates median values greater than background concentration (0.1 milligram per liter)
Average annual concentration of total phosphorus—in milligrams per liter
Background concentration
Lowest (less than 0.045)Medium (0.045 to 0.25)Highest (greater than 0.25)
Urban streamsThe highest total phosphorus levels in urban streams typically were in densely populated areas in relatively arid Western basins and in the East.
Rivers and streams with mixed land useTotal phosphorus concentrations generally correlated with nonpoint phosphorus inputs; however, in contrast to total nitrogen, levels in large rivers were highest in the Midwest, Great Plains, and West, where high concentrations of suspended sediment from erosion are common.
Agricultural streamsTotal phosphorus concentrations in agricultural streams were among the highest measured and generally correlated with nonpoint phosphorus inputs across the Nation.
46N
UTR
IEN
TS NUTRIENT INPUTS AND ENVIRONMENTAL FACTORSCONTROL NUTRIENT LOSSES FROM WATERSHEDS TOSTREAMSEnrichment of streams with nutrients is not simply explained bydifferences in land use. Land-management practices, nitrogen inputs tothe land surface, local and regional environmental characteristics, andseasonal effects also control the degree of enrichment. Such an integrationof factors explains why nutrient concentrations can be so different in differentregions of the Nation, despite seemingly similar land-use settings.
Nutrient inputs to the land are key to explaining variations in theamount of nutrients lost from watersheds to streams (nutrient yields). The
amounts of nutrients reaching streams generally increase as the totalnonpoint nutrient inputs increase. For streams, nitrogen yields wereless than or equal to about one-half of the total nonpoint inputs ofnitrogen from the atmosphere, commercial fertilizer, and manure.This is consistent with the tendency of nitrate to dissolve in water andbe transported with surface and subsurface runoff. Phosphorus yields,on the other hand, were less than or equal to about one-sixth of thetotal phosphorus inputs from commercial fertilizer and manure.Again, this is consistent with the general tendency of phosphorus toreadily attach to soil particles rather than to dissolve in water that runsoff to streams or seeps to ground water. In watersheds where cropsand other plants cannot use all nutrients applied during a growingseason, excess nutrients may be available for runoff to streams.
Local watershed characteristics and environmental settingsalso play key roles in determining nutrient yields. Watersheds withhigh nitrogen yields compared to total nitrogen inputs includeBachman Run, East Mahantango Creek, Kishacoquillas Creek, andMuddy Creek in the Lower Susquehanna River Basin; Broad Brookin the Connecticut, Housatonic, and Thames River Basins; andZollner Creek in the Willamette Basin. These agricultural streamsgenerally are in areas of high precipitation, which enhances runoff ofsurface water and flushing of shallow ground water, along withnitrogen, to streams. These watersheds also have a long history offarming, and they are located where soils and underlying geologicformations allow rapid movement of nitrogen-rich water throughshallow aquifers and into streams.
Characteristics of soil drainage can accentuate or mitigatenutrient yields to streams. For example, the Lost River in the WhiteRiver Basin exhibited high nitrate and phosphorus yields, particularlyduring high-flow conditions. Despite relatively low nutrient inputs,high nutrient yields probably result from sloping, clayey soils andshallow depth to permeable karst bedrock, which allow rapidtransport of nutrients to the Lost River. Watersheds with high nitrogeninputs and low nitrogen yields, such as Prairie and Shell Creeks in theCentral Nebraska Basins, have relatively flat-lying, sandy or silty
Average annual total nitrogen and totalphosphorus yields to agricultural streamsgenerally increase as nutrient inputsincrease; however, a greater proportion ofnitrogen than phosphorus input generallywas lost to streams. The large amount ofscatter in the data can be explained by localdifferences in agricultural practices, soils,geology, and hydrology across the Nation.
0 5 0 100 150 200 250
75
50
25
0
NITROGEN INPUT, in pounds per acre
in p
ou
nd
s p
er a
cre
Low inputsHigh yields
0 10 40 5020 30AV
ER
AG
E A
NN
UA
L T
OT
AL
PH
OS
PH
OR
US
YIE
LD
,
PHOSPHORUS INPUT, in pounds per acre
in p
ou
nd
s p
er a
cre
High inputsLow yields
1/2 or less ofnitrogen inputreaches stream
AV
ER
AG
E A
NN
UA
L T
OT
AL
NIT
RO
GE
N Y
IEL
D,
10
4
2
0
6
8
Low inputsHigh yields 1/6 or less of
phosphorus inputreaches stream
High inputsLow yields
Differences in occurrence and behavior of nutrientscomplicate prediction of effects and management options
47
NU
TRIE
NTS
soils where water infiltrates readily and nitrate migrates to shallow groundwater instead of being transported to streams.
Environmental factors controlling phosphorus yields to streams can bedifferent from those controlling nitrogen yields. Watersheds with lowphosphorus inputs and high phosphorus yields include Bullfrog Creek in theGeorgia-Florida Coastal Plain—an unusual case reflecting contributions fromnaturally occurring phosphate minerals—and Broad Brook in the Connecticut,Housatonic, and Thames River Basins, which receives supplementalfertilization from phosphorus-rich manure. Basins with high phosphorusinputs and low yields include streams in the Lower Susquehanna River Basin,San Joaquin-Tulare Basins, and Albemarle-Pamlico Drainage. These streamsgain significant flow from ground-water discharge, a source typically low inphosphorus because phosphates tend to be retained by the soil. Some of thesesame streams receive ground water that is high in nitrate because nitrogeninputs to the basins are high and nitrate can remain in solution. This isparticularly true where denitrification, a microbial process that can transformnitrate to nitrogen gas, is not a controlling factor.
Local differences in soils, geology, and hydrology affectnitrate migration from nonpoint sources to ground waterin a more pronounced way than for nutrient yields tostreams. Inputs of nitrogen were estimated from atmospheric,commercial fertilizer, and manure sources for areas within aone-third-mile radius of each monitoring well. Study areaswith low inputs of nitrogen and high median nitrateconcentrations (greater than about 4 mg/L) generally areunderlain by karst or fractured rock or by unconsolidated sandand gravel that allow nitrate to move readily to shallow groundwater. Such areas are found in the San Joaquin-Tulare Basins,Central Columbia Plateau, Red River of the North Basin,Western Lake Michigan Drainages, Lower Susquehanna RiverBasin, Potomac River Basin, and Connecticut, Housatonic,and Thames River Basins.
Areas with high nitrogen inputs but low median nitrateconcentrations (less than about 2 mg/L) generally areunderlain by relatively impermeable rock, silt, or clay, whichimpede downward movement of water. Examples of theseareas are found in the Rio Grande Valley, White River Basin,and Western Lake Michigan Drainages. The Jerome-Goodingagricultural site in the Upper Snake River Basin also fell in thehigh-input and low-concentration group, but this was morelikely related to the deep water table (median of 153 feet) inthis area.
A plot of nitrogen inputs to agriculturalland versus median nitrate concentrationsin underlying shallow ground watershows considerable scatter. Porous soilsand bedrock, which allow rapiddownward movement of water andnitrate, underlie areas with low nitrogeninputs and high nitrate concentrations inshallow ground water. Areas with highinputs and low concentrations generallyare underlain by less permeable geologicmaterials.
0 50 100 150 2000
5
10
15
NITROGEN INPUT, in pounds per acre
ME
DIA
N N
ITR
AT
E C
ON
CE
NT
RA
TIO
N,
in m
illig
ram
s p
er li
ter
Low inputsHigh concentrations
High inputsLow concentrations
NITROGEN INPUTS AND ENVIRONMENTAL FACTORS CONTROLNITRATE CONCENTRATIONS IN SHALLOW GROUND WATER
48N
UTR
IEN
TS
Nutrient concentrations varyseasonallyNutrient concentrations vary throughout the year, largely in response tochanges in precipitation and streamflow and to differences in time sincefertilizer or manure application. Nutrient concentrations in streams typicallyare elevated during high spring and summer streamflows, or peak irrigationperiods, following fertilizer application. In two agricultural streams in theWestern Lake Michigan Drainages, for example, more phosphorus wastransported during storms in June 1993 than during the 24 months thatfollowed.
High nutrient concentrations also can be found in streams during seasonallow-flow conditions. Nitrate concentrations in agricultural streams can be highduring winter low flow because of contributions from ground-water dischargeand (or) because algal uptake is low. Nitrogen and phosphorus concentrationsin streams downstream from metropolitan areas may be highest during variousseasonal low flows, when contributions from point sources are greater relativeto streamflow, and dilution is less.
Str
eam
Domestic well
Shallowmonitoring
well
Tile drain
Paired shallowand deep
monitoringwells
Riv
er
Clay-richglacial deposits
Confined sand andgravel aquifers
Water table
GlacialOutwash
Nitrateshownin red
Nitrate concentrations in ground water generally were less than 0.1 mg/L in many parts of the White RiverBasin even though it receives some of the highest nitrogen fertilization rates in the Nation. The White RiverBasin is in the Upper Midwest, which contains some of the highest percentages of cropland (58 percent) andcorn cropland (30 percent) in the United States. Nitrogen fertilization rates for corn often exceed 200 poundsper acre. Despite high nitrogen input rates, nitrate concentrations (shown in red, below) typically are low inground water. This occurs because (1) poorly drained and impermeable glacial deposits, such as clay and silt,restrict the downward movement of water and nitrate to the water table, (2) nitrate is intercepted andtransported to streams by tiles and ditches in many areas, and (3) nitrate is converted by denitrification toother forms of nitrogen where dissolved-oxygen concentrations in ground water are low. However, nitrateconcentrations can be locally elevated in areas formerly traversed by glacial streams. The streams depositedcoarse-textured and well-drained sand and gravel that allow rapid infiltration of water and enable nitrate tomove below the root zone before it can be taken up by plants. Seventeen percent of shallow wells in thesedeposits had nitrate concentrations that exceeded the drinking-water standard of 10 mg/L.
49
NU
TRIE
NTS
Total phosphorus concentrations in streams of the Apalachicola-Chattahoochee-Flint RiverBasin were higher during stormflow than during low flow and correlated with suspendedsediment concentrations from March 1993 through September 1995
EXPLANATION
Detection limitMinimum value
25th percentileMedian
75th percentileMaximum value
Number of observations
Storm-flow
28
Base flow
28 18 52 30 27 19 29 15 52 35 31 16
0.006
10
0.01
0.1
1
TO
TA
L P
HO
SP
HO
RU
SC
ON
CE
NT
RA
TIO
N,
in m
illig
ram
s p
er li
ter
26 18 37 28 25 19 27 26 38 32 25 15
1
10,000
10
100
1,000
SU
SP
EN
DE
D S
ED
IME
NT
CO
NC
EN
TR
AT
ION
,in
mill
igra
ms
per
lite
r
Poultry
Subur
ban
Urban
Fores
t
Cro
pland
(Clas
tic de
posit
s)
C
ropla
nd
(Car
bona
te be
droc
k)
Tributaries with one predominant land use
DOMINANT LAND USE
PIEDMONTCOASTAL
PLAIN
Nitrate concentrations in shallow ground water beneathagricultural land can change seasonally in response toirrigation patterns.
Nitrate levels in shallow ground water can change throughout the year, buttypically the seasonal changes are noticeable only in the upper 5-10 feet of thewater table in surficial aquifers. For example, nitrate concentrations in shallowground water from less than 10 feet below the water table in parts of the RedRiver of the North Basin ranged from about 8 to 25 mg/L from March 1994through September 1995. This variation was related in part to the timing ofimportant recharge periods, which generally occurred when spring snowmeltand major summer rainstorms coincided with irrigation periods, and in part tovariations in the timing and application of fertilizers applied to crops.
W.H
. Mul
lins
© 1
980
50N
UTR
IEN
TS
Stream-aquifer interactions controlnitrate concentrations near somestream reachesIrrigation and agricultural drainage can play a major role in the timing andmagnitude of nutrient concentrations, particularly in the western part of theNation, where large fluctuations in streamflow occur because of diversions forirrigation. Return flows from agricultural land during the irrigation season canaccount for most of the flow in many western streams and rivers, andconcentrations of potential contaminants often are highest during peakirrigation periods. In addition, low nutrient concentrations in irrigation canalscan dilute concentrations in ground water in areas where direct connectionsoccur between the canals and adjacent aquifers.
Stream-aquifer interactions can affect nutrient concentrationsdifferently during different times of the year in the same river reach. Forexample, nitrate concentrations in shallow ground water adjacent to the lowerSuwannee River in the Georgia-Florida Coastal Plain vary seasonally becauseof a cycle of water exchange between the river and the adjoining aquifer.During summer low flow, ground water containing high nitrate concentrationsenters the river, increasing river nitrate concentrations. During spring highflow, river water low in nitrate enters the aquifer, resulting in a decrease inground-water nitrate concentrations adjacent to the river.
Stream-aquifer interactions also can affect nutrient concentrationsdifferently in different parts of the same river basin. For example, nitrateconcentrations in about one-half of the wells sampled near the South PlatteRiver in Colorado exceeded the USEPA drinking-water standard. Groundwater contributes a substantial amount of flow to the river in this area, butconcentrations of nitrate in the river were substantially lower than in groundwater because microbial denitrification removed nitrate as ground waterpassed through the streambed. Farther downstream in Nebraska, ground waterin the alluvial aquifer adjacent to the Platte River is used for public supply byNebraska’s largest cities, including Omaha, Lincoln, Grand Island, andKearney. Pumping water from wells in this aquifer induces flow of Platte Riverwater into the aquifer and has the potential to decrease nitrate concentrationsin the ground water.
Water from irrigation canals effectivelydecreases nitrate concentrations inground water in the Quincy-Pasco area ofthe Central Columbia Plateau. ColumbiaRiver water diverted for irrigation leaksfrom canals and decreases nitrateconcentrations by dilution in shallowground water near the canals.
0.5
0.5
16
14
6.8
1.3
202.4
2.4
0.6
13
Canal lateralWellNitrate concentration, in
milligrams per liter
0
0
.5 MILE
.5 KILOMETER