w.h. mullins © - u.s. geological survey publications … ·  · 2001-07-09most pesticides are...

24
56 PESTICIDES Pesticides used to control weeds, insects, and other pests on farms and in urban areas can be harmful to humans and the environment if they contaminate our water resources W.H. Mullins © 1974 Dennis A. Wentz Gregg Patterson Kevin J. Breen W.H. Mullins ©

Upload: lythu

Post on 28-Apr-2018

217 views

Category:

Documents


3 download

TRANSCRIPT

56PE

STIC

IDES

Pesticides used to control weeds, insects, and otherpests on farms and in urban areas can be harmful tohumans and the environment if they contaminate ourwater resources

W.H. Mullins © 1974

Dennis A. Wentz

Gregg Patterson

Kevin J. Breen

W.H. Mullins ©

57

PEST

ICID

ES

PesticidesResults of NAWQA studies show that pesticides are widespread in streams and ground water sampled within

agricultural and urban areas of the Nation. As expected, the most heavily used compounds are found most

often, occurring in geographic and seasonal patterns that mainly correspond to distributions of land use and

associated pesticide use.

The frequency of pesticide contamination, however, is greater than expected. At least one pesticide was

found in almost every water and fish sample collected from streams and in about one-half of all wells

sampled. Moreover, individual pesticides seldom were found alone— almost every water and fish sample

from streams and about one-half of samples from wells with a detected pesticide contained two or more

pesticides.

For individual pesticides in drinking water, NAWQA results are generally good news relative to current

water-quality standards and guidelines. Average concentrations in streams and wells rarely exceeded

standards and guidelines established to protect human health. For aquatic life and wildlife, however, NAWQA

results indicate a high potential for problems in many streams, particularly in urban areas, where

concentrations of more than one pesticide often approached or exceeded established water-quality

guidelines.

Important questions remain unanswered about potential risks of pesticide contamination to humans and

the environment. Currently, standards and guidelines are available only for a limited number of individual

pesticides, do not account for mixtures of pesticides or for pesticide breakdown products, and are based on

tests that have assessed a limited range of potential health and ecological effects. Long-term exposure to

low-level mixtures of pesticide compounds, punctuated with seasonal pulses of higher concentrations, is the

most common pattern of exposure, but the effects of this pattern are not yet well understood.

The uncertainty about whether present-day levels of pesticide contamination are a threat to human health or

the environment makes it imperative that we document and understand the nature of pesticide exposure, the

causes of contamination, and the actions we can take to reduce pesticide levels in streams and ground water. ◗

58PE

STIC

IDES

WHAT WAS MEASURED...

Many of the Nation’s most heavily used agricultural and urbanpesticides were measured in the NAWQA Program. The 83 targetcompounds analyzed in water include 76 pesticides and 7 selectedbreakdown products and account for about 75 percent of theNation’s agricultural use of synthetic pesticides. They include 17 ofthe top 20 herbicides and 15 of the top 20 insecticides.

Historically used organochlorine insecticides, like DDT, weremeasured in bed sediment and fish, where they accumulate andpersist for decades. The 32 organochlorine compounds analyzed inbed sediment or fish consist of 8 individual parent compounds,1 individual breakdown product, and 7 groups of parent compoundsplus related breakdown products or chemical impurities in themanufactured product. These compounds account for more than90 percent of the Nation’s historical use of organochlorineinsecticides in agriculture.

WHAT WAS NOT...

Many important pesticide compounds were not measured becauseof analytical and budget constraints. The top 20 herbicides notmeasured were glyphosate (ranked 10), MSMA (14), and propazine(17). The top 20 insecticides not measured were cryolite (12),acephate (13), dimethoate (14), methomyl (15), and thiodicarb (18).Other pesticides not measured include inorganic pesticides, suchas sulfur and copper, oil, and biological pesticides. Importantomissions also include numerous pesticide breakdown productsand carrier agents that may affect water quality.

Although NAWQA studies are targeting the broadest and mostcomplete range of pesticides ever measured in a singleassessment, these omissions are important to keep in mind andmust temper conclusions.

Further information on pesticides measured is available via theWorld Wide Web at <http://water.usgs.gov/lookup/get?nawqapest>.

More than 90 percent of water and fish samples from all streams contained oneor, more often, several pesticides. Pesticides found in water were primarilythose that are currently used, whereas those found in fish and sediment areorganochlorine insecticides, such as DDT, that were heavily used decades ago.Most of the pesticides in use today are more water soluble and break downfaster in the natural environment than the long-lived organochlorineinsecticides of the past.(31)

About 50 percent of the wells sampled contained one or more pesticides,with the highest detection frequencies in shallow ground water beneathagricultural and urban areas and the lowest frequencies in major aquifers,which generally are deeper. Ground water has a lower incidence of pesticidecontamination than streams because water infiltrating the land surface moves

slowly through soil and rock formations on its way toground water and through the aquifer. This contact withsoil and rock and the slow rate of flow allow greateropportunity for sorption and degradation of pesticides,and varied flow pathways mean that some wells do not tapground water that originated from places or times affectedby pesticide use.

Although streams and rivers are more vulnerable thanground water to rapid and widespread contamination,ground-water contamination is extremely difficult toreverse because of the slow rate of ground-water flow.Management practices that reduce the transport ofpesticides to streams can yield rapid improvements inwater quality. Ground water, on the other hand, willrespond slowly to changing practices—sometimestaking many years or even decades to recover.

Decades of pesticide use have resulted in their wide-spread occurrence in streams and ground water

59%

100%

92%

99%

49%

33%

Streams

Streams

Major Rivers

Shallow Ground Water

Shallow Ground Water

Major Aquifers

Agricultural areas

Urban areas

Mixed land use

96%Fish

Fish 100%

Fish 85%

SAMPLES WITH ONE OR MORE PESTICIDES

59

PEST

ICID

ES

Pesticides are a potential concern forhuman health and aquatic lifeMost pesticides are manufactured compounds that are designed to kill specificpests, such as weeds and insects. Many pesticides have the potential to harmnontarget organisms, especially if the organisms are exposed to high levels orfor a long period of time. In the early 1960s, Rachel Carson’s widelypublicized book “Silent Spring”(32) described the ecological impacts of DDTand other pesticides. Concerns about the unintended effects of pesticidescontinue to this day, and evaluation of the risk to humans and the environmentfrom present-day levels of pesticide exposure remains highly controversial.

A difficult aspect of evaluating potential effects of pesticides is determiningwhat may occur as a result of varying types and durations of exposure.Exposure is complicated by pesticide mixtures, breakdown products, strongseasonal concentration pulses, and high concentrations during stormflows. Incontrast, most toxicity assessments are based on controlled experiments with asingle contaminant over a limited range of concentrations.

Although uncertainties remain, water-quality standards and guidelines havebeen developed for many pesticides in order to protect human health andaquatic life, and they are used in this report to signal potential problem areas.Concentrations that exceed a standard or guideline, however, may not be aproblem at some sites. Conversely, the absence of an exceedance does notensure that there is no problem.

Some people believe that any presence of pesticides in their drinking wateris too much, whereas others feel that the standards and guidelines establishedfor many of the major pesticides provide adequate protection. Which of theseperspectives is closest to the truth remains unclear, but certainly the effects ofcommon patterns of pesticide exposure found in NAWQA studies have not yetbeen fully evaluated.

The uncertainty in whether or not present-day levels of pesticidecontamination are a threat to human health or aquatic life makes it imperativethat we understand the nature of exposure, the causes of contamination, andthe actions we can take to reduce pesticide levels in streams and ground water.Only by accurately characterizing the nature and causes of environmentalexposure can we develop effective strategies to minimize exposure and reliablyevaluate relations between exposure and effects.

HORMONE LEVELS IN FISH SHOW SIGNS OF POSSIBLEENDOCRINE DISRUPTION

A reconnaissance study of sex hormones in carp collected at 11NAWQA stream sites indicates that pesticides may be affectingthe ratio of estrogen to testosterone in both males and females.(33)

The hormone ratio, which is sometimes used as an indicator ofpotential abnormalities in the endocrine system, was significantlylower at sites with the highest pesticide concentrations. Althoughthe lower hormone ratios may not be associated with measurableeffects on fish populations, they are a signal that furtherinvestigation is needed.

STANDARDS AND GUIDELINES FORPROTECTING WATER QUALITY

Water-quality standards and guidelinesgenerally are maximum acceptableconcentrations of pesticides for protectinghumans, aquatic life, or wildlife. They areestablished by the United States andother nations, international organizations,and some States and tribes. For thisreport, precedence was given tostandards and guidelines established bythe USEPA and then to those establishedby Canada or the International JointCommission for the Great Lakes, althoughsome states may have different standardsand guidelines that take priority forparticular water bodies.(25)

Drinking-water standards or guide-lines have been established for 43 of the76 pesticides analyzed, and aquatic-lifeguidelines have been established for 28 ofthe 76 pesticides. Aquatic-life or wildlifeguidelines are available for 8 of the 16pesticides (compounds or groups)analyzed in bed sediment or fish.

Current standards and guidelines donot completely eliminate risks because:(1) values are not established for manypesticides, (2) mixtures and breakdownproducts are not considered, (3) theeffects of seasonal exposure to highconcentrations have not been evaluated,and (4) some types of potential effects,such as endocrine disruption and uniqueresponses of sensitive individuals, havenot yet been assessed.

0

1

2

3

4

5

0 1 2 3

RA

TIO

OF

ES

TR

OG

EN

:TE

STO

ST

ER

ON

E

TOTAL CONCENTRATION OF PESTICIDES IN WATER, in micrograms per liter

FEMALE CARP

MALE CARP

60PE

STIC

IDES

The most frequently detected pesticides are those mostheavily used…now or in the pastNot surprisingly, the top 15 pesticide compounds found inwater are among those with the highest current use. Theyinclude five of the most heavily used agriculturalherbicides and one degradation product, five herbicidesthat are extensively used in urban areas, and four of themost commonly used insecticides.

The pesticide compounds found most often in fish andbed sediment are related to three major groups ofinsecticides that were heavily used in the 1960s.Organochlorine compounds related to DDT and dieldrinwere widely used in both agricultural and urban areas, andchlordane was mainly used in urban areas.

0

20

40

60

80

100

0

20

40

60

80

100

FR

EQ

UE

NC

Y O

F D

ET

EC

TIO

N, a

s p

erce

nta

ge

of

sam

ple

s

p,p'

-DD

E

p,p'

-DD

D

p,p'

-DD

T

o,p'

-DD

T

o,p'

-DD

D

o,p'

-DD

E

tran

s-no

nach

lor

cis-

chlo

rdan

e

cis-

nona

chlo

r

tran

s-ch

lord

ane

oxyc

hlor

dane

DIE

LD

RIN

0

20

40

60

80

100

Agricultural land

Urban land

Major rivers and streams with mixed land use

DDT CHLORDANE

Fish

Bed sediment

Fish and Bed SedimentWater

0

20

40

60

80

100

0

20

40

60

80

100

FR

EQ

UE

NC

Y O

F D

ET

EC

TIO

N, a

s p

erce

nta

ge

of

sam

ple

s

Atr

azin

e

Dee

thyl

atra

zine

Met

olac

hlor

Cya

nazi

ne

Ala

chlo

r

EP

TC

Sim

azin

e

Pro

met

on

2,4-

D

Diu

ron

Teb

uthi

uron

Dia

zino

n

Car

bary

l

Mal

athi

on

Chl

orpy

rifos

0

20

40

60

80

100

Agricultural land

Urban land

Major rivers and aquifers with mixed land use

AGRICULTURAL HERBICIDES URBAN HERBICIDES INSECTICIDES

Ground Water

Streams

PESTICIDE DETECTIONS

Char

les

G. C

raw

ford

61

PEST

ICID

ES

Different pesticides dominate in different land-use areasThe occurrence of pesticides instreams and ground water followsbroad patterns in land use andassociated pesticide use. The patternsare complex, however, and differbetween streams and ground waterbecause of the wide range of usepractices and processes that governthe movement of pesticides in thehydrologic environment.

AGRICULTURAL AREASHerbicides are the most common typeof pesticide found in streams andground water within agriculturalareas. The most common herbicidesin agricultural streams were atrazineand its breakdown product

deethylatrazine (DEA), metolachlor,cyanazine, alachlor, and EPTC. All 5of the parent compounds rank in thetop 10 in national use. Atrazine wasfound in about two-thirds of allsamples from agricultural streams,often occurring year-round.

Similar to streams, the mostcommon compounds found inshallow ground water were atrazineand DEA, but only about one-third ofthe samples had detectable levels.The lower rates of atrazine and DEAdetection in ground water comparedto streams result from longer traveltimes, greater opportunity for

sorption or breakdown, and greatervariability of source water in wells.

One of the most striking results forshallow ground water in agriculturalareas, compared with streams, is thelow rate of detection for several high-use herbicides other than atrazine.This is probably because theseherbicides break down faster in thenatural environment compared toatrazine. Studies show that break-down products of metolachlor,alachlor, and cyanazine are muchmore commonly found in groundwater than are the parentcompounds.(34)

Compared to herbicides, currentlyused insecticides were less frequentlyfound in most agricultural streams.But some streams in agriculturalareas with particularly high use ofspecific insecticides, such as diazinonin the San Joaquin-Tulare Basins, hadamong the highest concentrationsmeasured. Insecticides were rarelydetected in ground water inagricultural areas. The less frequentoccurrence of currently usedinsecticides in streams comparedwith herbicides, and theirinfrequent occurrence in groundwater, result from their relativelylow application rates and rapidbreakdown in the environment.

In contrast to currently usedinsecticides, the organochlorineinsecticides of the past still persist inagricultural streams because of theirextreme resistance to breakdown inthe environment. DDT was the most

commonly detected organochlorinegroup—found in almost every fishsample— followed by dieldrin andchlordane. DDT and aldrin (whichbreaks down rapidly to dieldrin in theenvironment) were two of the topthree insecticides used for agriculturein the 1960s.

Jana

S. S

tew

art

Denn

is A

. Wen

tz

Denn

is A

. Wen

tz

62PE

STIC

IDES

URBAN AREASThe most distinct differences betweenpesticides found in urban andagricultural areas are the greaterprevalence of insecticides in urbanstreams and the relatively frequentoccurrence of urban herbicides inboth streams and shallow groundwater. Insecticides were found moreoften, and usually at higherconcentrations, in urban streams thanin agricultural streams. Diazinon,carbaryl, chlorpyrifos, and malathion,which nationally rank 1, 8, 4, and 13among insecticides used for homesand gardens, accounted for mostdetections in water. Historically usedinsecticides also were found morefrequently in urban streams. Urbanstreams had the highest detectionfrequencies of DDT, chlordane, and

dieldrin in fish and bed sediment, andthe highest concentrations ofchlordane and dieldrin. Chlordaneand aldrin were widely used fortermite control until the mid-1980s,although their agricultural uses wererestricted during the 1970s.(35, 36)

Much more chlordane was used fortermite control than for agriculture.

Insecticides in urban streams area concern for aquatic life, fordownstream water supplies, andpossibly for recreational users.Effective management will likelyrequire a combination of reducingcurrent home, garden, andcommercial use and controllingsediment sources to streams.

Similar to agriculturalareas, insecticides wereseldom detected in groundwater in urban areas.Interestingly, however, themost commonly detectedinsecticide in shallow groundwater was dieldrin, whichwas found in about 3 percentof the wells sampled.Although dieldrin is not verymobile in water, its

environmental persistence and theheavy historical use of dieldrin andaldrin have combined to yieldcontamination of some wells.

The herbicides most commonlyfound in urban streams, in addition toatrazine and metolachlor,are simazine, prometon,2,4-D, diuron, andtebuthiuron, all of whichare commonly used innonagricultural settings formaintenance of roadsides,commercial areas, lawns,and gardens. Prometon and2,4-D have among the

highest frequencies of urban use. Ofthe urban herbicides, 2,4-D, simazine,and diuron also have substantialagricultural use, ranking in the top 25nationally. Diuron and 2,4-D were notdetected as frequently as othercompounds with similar use,probably because the analyticalmethod for these two compounds isless sensitive and resulted in fewerdetections than for other compounds,even when concentrations weresimilar. As in streams, the mostfrequently found herbicides inshallow ground water in urban areaswere atrazine, DEA, simazine, andprometon. Unlike streams, however,metolachlor was seldom detected,probably because of its lower urbanuse and lower persistence in theenvironment compared to the otherherbicides.

Pete

r E. H

ughe

s

Barb

ara

J. D

awso

n

Kevi

n J.

Bre

en

63

PEST

ICID

ES

Pesticides found in major rivers and aquifers reflectcontributions from both agricultural and urban areasMajor rivers and streams drainingareas of mixed land use containpesticides from both agricultural andurban sources and from both past andpresent use. In water that comesmainly from agricultural areas, themost commonly found pesticides arethe major herbicides atrazine (andDEA), metolachlor, cyanazine, andalachlor. In water that comes mainlyfrom urban areas, the most commonpesticides are the herbicides simazineand prometon and the insecticidesdiazinon and carbaryl.

Like water, the fish and bedsediment of major rivers and streamswith mixed land-use influencescontain mixtures of organochlorineinsecticides from agricultural andurban areas. Detection frequenciesand concentrations of DDT, dieldrin,and chlordane were generallyintermediate between those ofagricultural and urban streams.

Many large rivers with mixed land-use influences tend to have lowerconcentrations of pesticides comparedwith agricultural and urban streamsbecause of a larger influence ofundeveloped land. Some rivers inintensive agricultural regions,however, have concentrations that aresimilar to those in agriculturalstreams, although they are lessvariable over time. Rivers with mixedland uses almost always containdetectable pesticides that reflect thediversity of sources present.

In contrast, ground water in majoraquifers has a substantially lowerfrequency of pesticide occurrencethan shallow ground water inagricultural and urban areas. Thisdifference results from the generallydeeper wells sampled in majoraquifers and the greater influence ofundeveloped areas. Additionally,owing to the slow rate of ground-

water flow, much of the watersampled in the major aquifers mayhave infiltrated into the ground beforepesticides were applied. The twomost frequently detected compoundsin major aquifers were atrazine andDEA, resulting from the high andextensive use of atrazine, the greaterextent of agricultural land comparedto urban land affecting most majoraquifers sampled, and the highmobility and long-lived nature ofatrazine and DEA.

Because the pesticides found inmajor rivers and aquifers reflectcontributions from both agricul-tural and urban land uses, effortsto improve the quality of thesewater resources will requiremanagement of nonpoint sources inboth agricultural and urban areas.

Pete

r A. S

teev

es

64PE

STIC

IDES

This satellite image of theCentral Columbia Plateau, takenin 1992, shows irrigated fields ingreen and fallow fields andrangeland in red. Agriculturalrunoff, tile drainage, and returnflows from the irrigatedfarmland drains into theColumbia River, which forms thewestern border of the areabefore the Snake River joins itfrom the east.

Data courtesy of the U.S. GeologicalSurvey, National Mapping Division,EROS Data Center

65

PEST

ICID

ES

Geographic distributions of pesticidesfollow patterns in land use andpesticide useAn essential step toward understanding and managing the effects ofpesticides on water quality is to examine the geographic distribution ofpesticide levels in relation to land use and pesticide use and to determineareas of the Nation and environmental settings that merit the greatestconcern and attention.

The geographic distribution of pesticide levels is summarized in a series ofmaps that show results for herbicides and insecticides in streams and groundwater for agricultural, urban, and mixed land uses, the latter including majorrivers and aquifers. To identify potential water-quality problems, pesticideconcentrations in water and bed sediment from streams are compared toaquatic-life guidelines because most streams sampled are not directly used asdrinking-water sources. Pesticide concentrations in shallow ground water andwater from major aquifers are compared to drinking-water standards andguidelines for human health. Most of the major aquifers, and shallow groundwater in about one-half of the study areas, are sources of drinking water.Methods used to construct the maps are explained on page 31.

The national maps show national and regional patterns, or in some cases theapparent lack of pattern, in pesticide levels. They cannot, however, showimportant aspects of local variability in pesticide levels—for this, the reader isreferred to the individual reports available for each NAWQA Study Unit (seepage 80).

A.C. Haralson, Arkansas Department of Parks and Tourism Phil Schofield ©Alan R. Wycheck, Harrisburg Hershey Carlisle Tourism and Convention Bureau

66PE

STIC

IDES

Herbicides in streams and major riverswere highest in the most intensivelyfarmed agricultural regionsTotal herbicide concentrations consistently ranked highest in agriculturalstreams and major rivers of the White River Basin and Central NebraskaBasins, which are on the eastern and western margins of the Corn Belt,respectively. The Corn Belt has the highest herbicide use in the Nation. Thehigh concentrations measured in the White River Basin and Central NebraskaBasins are consistent with other studies in the Mississippi River Basin, whichshow broad-scale herbicide contamination of streams and rivers, including theMississippi River.(37)

All seven agricultural streams and the two major rivers sampled in theWhite River Basin and Central Nebraska Basins frequently had concentrationsof one or more herbicides that exceeded a Canadian aquatic-life guideline.Atrazine exceeded its guideline of 2 µg/L at all sites, and cyanazine exceededits guideline of 2 µg/L at four sites. At this time, there are no national aquatic-life guidelines for these compounds in the United States, and individual Stateshave varying guidelines.

Given the regional extent of intensive herbicide use and elevated levelsof herbicides in streams within the Corn Belt, management strategies thatare successful in reducing use and runoff of herbicides that are applied forcorn and soybean production will likely lead to regional-scale improvementsin water quality.

Other streams ranking high in herbicide concentrations were agriculturalstreams that drain intensively farmed areas in the Willamette Basin, SanJoaquin-Tulare Basins, South Platte River Basin, and Trinity River Basin. Adiverse group of herbicides, including trifluralin, metolachlor, and 2,4-D, inaddition to atrazine and cyanazine, exceeded aquatic-life guidelines in one ormore of these streams.

Most streams with low herbicide concentrations were agricultural streams inareas with low to moderate herbicide use in their drainage basins. Exceptionsto this are low concentrations of herbicides in agricultural streams of the RedRiver of the North Basin and in the Southeast, even though use is moderate tohigh. One possible reason for the low concentrations in the Red River of theNorth Basin is a higher retention of herbicides in the soil because ofparticularly high levels of organic matter.

Among urban sites, only Las Vegas Wash in Las Vegas had relatively highherbicide concentrations compared to other streams. Only Little Buck Creek inthe Indianapolis area had concentrations that exceeded a Canadian aquatic-lifeguideline, and that was in a small percentage of samples because of atrazineuse on agricultural land in its watershed.

The heavy use of herbicides on corn inthe Central Nebraska Basins is reflectedin high atrazine concentrations in thePlatte River during runoff from rainfallfollowing spring herbicide applications.Low-level atrazine concentrations werefound throughout much of the year,punctuated by seasonal pulses of highconcentrations that exceeded thedrinking-water standard (MCL) and theCanadian aquatic-life guideline. Theannual average concentration, however,did not exceed the drinking-waterstandard.

Lincoln, Omaha, and smaller citiesalong the Platte River withdraw drinkingwater from an aquifer adjacent to theriver. Much of the ground water that ispumped from the sand and gravelportions of this aquifer is vulnerable tocontamination from atrazine in the PlatteRiver. This is a concern to water providers

because studies have shown thatconventional water treatment isineffective in removing herbicides likeatrazine from the treated water suppliedto households.(38)

S O N D J F M A M J J A1991 1992

0

30

5

10

15

20

25

AT

RA

ZIN

E C

ON

CE

NT

RA

TIO

N, i

n m

icro

gra

ms

per

lite

r

Canadian aquatic-life guideline

Drinking-water standard

ground waterAtrazine to

Atrazine in runoff

Water-treatmentplant Ground-

water wellpump

HERBICIDES EXCEEDED WATER-QUALITYSTANDARDS OR GUIDELINES IN SOMESTREAMS IN THE CORN BELT

67

PEST

ICID

ES

Urban areas

Central Nebraska Basins

WhiteRiverBasin

WillametteBasin

Red Riverof the North Basin

SanJoaquin-

TulareBasins

TrinityRiverBasin

South PlatteRiver Basin

Dallas-Ft Worth •

• Atlanta

• Tallahassee

• Las Vegas

• Norwalk

• Indianapolis

• Portland

• Denver

• Harrisburg

• Washington D.C.

Albany •

2713

33

64

5

6

14

15

21 23

943

23

11

4

10

21

26

6

Agricultural streamsConcentrations were highest and most often exceeded aquatic-life guidelines in streams in the White River Basin and Central Nebraska Basins in the Corn Belt, where herbicide use is among the highest reported nationwide.

Urban streamsMost urban streams had moderate or low herbicide concentrations compared to streams in agricultural and mixed land-use settings.

Rivers and streams with mixed land useAquatic-life guidelines were exceeded in about one fourth of the samples from the two major rivers sampled in the Corn Belt, but most major rivers had moderate herbicide concentrations.

See p. 31 for more information about these maps

Herbicide use—in pounds per acre of agricultural land

Highest (greater than 0.461)Medium (0.162 to 0.461)Lowest (less than 0.162)No reported use

A national ranking of HERBICIDES in streams

Bold outline indicates exceedance by one or more herbicides. Number ispercentage of samples that exceededa guideline

Sum of herbicide concentrations

Aquatic-life guidelines

Lowest 25 percentMiddle 50 percentHighest 25 percent of streams

9

68PE

STIC

IDES

Most urban streams sampled, plus two major rivers dominated by urbaninfluences— the South Platte River downstream from Denver and the TrinityRiver downstream from Dallas-Fort Worth—had among the highest insecticideconcentrations of all streams and rivers sampled. Nine of 11 urban streams andboth rivers had concentrations that exceeded aquatic-life guidelines, usually inmore than 20 percent of the samples. The most common insecticides to exceedguidelines were diazinon, chlorpyrifos, and malathion. Chlorpyrifos andmalathion have USEPA aquatic-life criteria of 0.041 µg/L and 0.100 µg/L,respectively, and diazinon has a guideline of 0.080 µg/L established by theInternational Joint Commission for the Great Lakes.

Insecticides in urban streams, largely from use around homes and ingardens, parks, and commercial areas, frequently occur at levels ofconcern for aquatic life and may be a significant obstacle for restoringurban streams.

Most agricultural streams had moderate or low concentrations ofinsecticides but, as for herbicides, several streams that drain intensively farmedareas that are irrigated had among the highest insecticide levels. Although

concentrations of insecticides inagricultural streams tended to be lowcompared to urban streams,concentrations above aquatic-lifeguidelines were common. For aboutone-half of the agricultural streams,samples exceeded a guideline for oneor more insecticides. In addition todiazinon and chlorpyrifos, aninsecticide that frequently exceeded itsguideline in agricultural streams wasmethyl azinphos, which has a USEPAaquatic-life criterion of 0.010 µg/L.

Insecticide concentrations in mostmajor rivers usually were lower thanthose measured in urban streams andexceeded aquatic-life guidelines inrelatively few samples. Exceptions arethe San Joaquin River, which drainsfarmlands with some of the heaviestinsecticide use in the Nation, and theSouth Platte and Trinity Rivers, whichare affected by both point andnonpoint sources from urban areas.

J F M A M J J A S O N D

Carbofuran

Chlorpyrifos

Diazinon

Malathion

Methyl azinphos

Detected

Exceededaquatic-lifeguidelines

Approachedaquatic-lifeguidelines

No

t a

na

lyz

ed

INSECTICIDES DETECTED IN WATER IN ZOLLNER CREEK, OREGON, 1993 AND 94

Zollner Creek in the Willamette Basin receives agricultural runofffrom intensively irrigated crops, including row crops, grass, wheat,hops, nurseries, and orchards. A wide variety of insecticides wasapplied to these crops; the insecticides were transported to thecreek by irrigation and stormwater runoff. One or more insecticideswere found in most water samples collected during the 2-yearperiod, and several approached or exceeded concentrations thatmay be harmful to aquatic life, sometimes accurring as mixtures.

Insecticides in streams were highest in urban areas

Den

nis

A. W

entz

69

PEST

ICID

ES

Urban areas

WillametteBasin

South PlatteRiver Basin

SanJoaquin-

TulareBasins

Trinity River Basin

Dallas-Ft Worth •

• Atlanta

• Tallahassee

• Las Vegas

• Norwalk

• Indianapolis

• Portland

• Denver

• Harrisburg

• Washington D.C.

Albany •

14

13

3

21

39

77

30

13

28

43

7

32128

6

43

137 8

9

1018

48

16

9

39

38

45

6

6

9

Agricultural streamsMost streams had moderate or low concentrations, but several in irrigated areas of the West had among the highest concentrations. About one-half of the agricultural streams had concentrations that exeeded an aquatic-life guideline.

Urban streamsMost streams had among the highest concentrations. Typically, 10 to 40 percent of samples had concentrations that exceeded one or more aquatic-life guidelines.

Rivers and streams with mixed land useConcentrations were low to moderate except for the urban-affected South Platte and Trinity Rivers, and the San Joaquin River, which drains farmlands with some of the most intensive insecticide use in the Nation.

See p. 31 for more information about these maps

Insecticide use—in pounds per acre of agricultural land

Highest (greater than 0.086)Medium (0.033 to 0.086)Lowest (less than 0.033)No reported use

A national ranking of INSECTICIDES in streams

Bold outline indicates exceedance by one or more insecticides. Number ispercentage of samples that exceeded a guideline

Sum of insecticide concentrations

Aquatic-life guidelines

Lowest 25 percentMiddle 50 percentHighest 25 percent

9

70PE

STIC

IDES

Organochlorine insecticides werehighest in urban streams and wherehistorical agricultural use wasgreatestConcentrations of organochlorine insecticides in bed sediment and fishcorrespond to land use and past application rates. Although most uses oforganochlorine insecticides ended 10 to 25 years ago, they remain a significantwater-quality issue for many streams. Overall, 14 percent of bed-sedimentsamples had concentrations that exceeded sediment-quality guidelines forprotection of aquatic life,(39) and 19 percent of sites had concentrations in fishthat exceeded New York guidelines for protection of fish-eating wildlife.(40)

Compounds that most often exceeded guidelines were DDT and chlordane inbed sediment and DDT and dieldrin in fish.

Almost all urban streams had high or medium concentrations of theorganochlorine insecticides compared with other sites. Sediment-qualityguidelines were exceeded at 37 percent of urban sites, with several sites eachin urbanized areas of the Connecticut, Housatonic, and Thames River Basins,Hudson River Basin, Trinity River Basin, and Georgia-Florida Coastal Plain.Concentrations in whole fish exceeded guidelines for the protection of fish-eating wildlife at 21 percent of urban sites.

In agricultural streams, concentrations of organochlorine insecticides werehighest in areas of high past use. High concentrations were most common forstreams in the Central Columbia Plateau, Georgia-Florida Coastal Plain, andTrinity River Basin. One or more sediment-quality guidelines were exceededat 15 percent of agricultural sites, and concentrations in whole fish exceededwildlife guidelines at 20 percent of sites.

Many streams and rivers with mixed land-use influences also had highconcentrations in bed sediment, particularly in basins with extensiveagricultural areas where past use was high, such as in the Southeast and theirrigated West, and in basins with high population density, such as in theNortheast. Sediment-quality guidelines were exceeded at 11 percent of thesesites, and wildlife guidelines were exceeded in whole fish at 24 percent ofthese sites. In undeveloped areas, organochlorine concentrations generallywere low and did not exceed sediment-quality guidelines.

A significant health concern in some regions is consumption of fish withhigh levels of organochlorine insecticides in their flesh. Human-healthguidelines for edible fish tissue(41) are not directly applicable to NAWQAresults, which are based on whole-fish analysis of mostly carp and suckers.Nevertheless, the NAWQA fish data provide a relative indication of potentialconcern. At about 30 percent of NAWQA sites, insecticide concentrationsin whole fish exceeded human-health guidelines for edible fish tissue.(41)

For any of these streams that are active fisheries, additional assessment offillets of edible species is advisable if this has not already been done.

CONTROL OF SOIL EROSION IS KEY TOREDUCING ORGANOCHLORINEINSECTICIDES

Organochlorine insecticides bind stronglyto soils and are carried with eroded soilsto streams by runoff from irrigation andrainfall. In streams, the soil-boundinsecticides may dissolve in water, remainsuspended, or settle to the streambed.They also accumulate in fish. Under-standing and managing soil erosion is akey to reducing organochlorinecontamination.

For example, furrow irrigation causesmore erosion than sprinkler or dripirrigation. In the Central Columbia Plateau,DDT concentrations in streambedsediment and fish increased as thepercentage of furrow irrigation in thebasin increased.

In the San Joaquin-Tulare Basins,the amount of DDT transported withsuspended sediment in the San JoaquinRiver and tributaries generally wasgreater during winter runoff than duringthe irrigation season. Controllingirrigation-induced soil erosion wouldreduce but not eliminate DDT in thestreams because large quantities aretransported during infrequent storms.

1

10

100

1,000

10,000

100,000

0 10 20 30 40 50 60 70 80 90 100

Streambed sedimentFish tissue

PERCENTAGE OF IRRIGATEDCROPLAND IN FURROW IRRIGATION

TOTA

L D

DT,

in m

icro

gra

ms

per

kilo

gra

m

0.01 10,0000.1 1 10 100 1,000

TOTAL DDT TRANSPORTED,in grams per day

SpanishGrant Drain

San JoaquinRiver

NewmanWasteway

OrestimbaCreek

Olive AvenueDrain

Del PuertoCreek

IngramCreek

HospitalCreek

Irrigation seasonWinter runoff

71

PEST

ICID

ES

Urban areas

SanJoaquin-

TulareBasins

CentralColumbia Plateau

SanJoaquin-

TulareBasins

Georgia-Florida Coastal Plain

TrinityRiverBasin

TrinityRiverBasin

Georgia-Florida Coastal Plain

Connecticut, Housatonic,and Thames River Basins

Hudson RiverBasin

Agricultural streamsHighest concentrations occurred where historical use was highest on crops such as cotton, peanuts, orchards, and vegetables.

Urban streamsMost urban streams had higher concentrations than the majority of agricultural streams, and concentrations exceeded sediment-quality guidelines at almost 40 percent of the sites.

Rivers and streams withmixed land useConcentrations followed the patterns in contributing agricultural and urban areas, with the highest concentrations in areas of high population densities or intensive historical use in agriculture.

See p. 31 for more information about these maps

Historical organochlorine use—in pounds per acre of agricultural land

Highest (greater than 0.278)Medium (0.095 to 0.278)Lowest (less than 0.095)No reported use

A national ranking of ORGANOCHLORINES in bed sediment

Bold outline indicates exceedance by one or more organochlorine insecticides

Sum of organochlorine insecticide concentrations

Aquatic-life guidelines

Lowest 50 percent (all with no detections)Second highest 25 percentHighest 25 percent

72PE

STIC

IDES The highest frequencies of detection

for pesticides in ground water werefor herbicides in shallow groundwater beneath agricultural areas. Inthese areas where herbicide use wasmoderate to high, soil and geologicconditions favored rapid movement ofherbicides to the ground water. Moststudies of shallow ground water inagricultural areas detected herbicidesin more than 50 percent of wellssampled.

Compared to streams, ground-water detections were dominated byfewer compounds—mainly those thathave the combination of relativelyhigh mobility and chemical stabilitythat allows them to move and persistin the flow system long enough toreach a well. Only atrazine, itsbreakdown product DEA, metolachlor,prometon, and simazine were foundin more than 5 percent of all wells.

Of the 36 studies of shallowground water in agricultural areas,which included more than 1,000wells, only one well in an unusedshallow ground-water area in theConnecticut, Housatonic, and Thames

PREDICTING ATRAZINE CONTAMINATION IN GROUND WATER

As part of Idaho’s State Pesticide Management plans forherbicides, maps have been developed to portray the potential foratrazine contamination in ground water in southeastern Idaho.(42)

Atrazine data from the NAWQA Program in the Upper Snake RiverBasin were used to calibrate and verify predictive models.Significant factors used to successfully predict atrazineconcentrations in ground water were atrazine use, land use,precipitation, soil type, and depth to ground water. Continueddevelopment of these types of modeling tools will aid in designingcost-effective programs for monitoring and protecting ground-water resources across the Nation.

River Basins had an atrazineconcentration that exceeded thedrinking-water standard of 3 µg/L.

Herbicides were moderatelycommon in shallow ground waterbeneath urban areas. In an urban areaof the Albemarle-Pamlico Drainage,a shallow aquifer used for drinking-water supply had one monitoring wellwhere an atrazine concentrationexceeded the drinking-water standard.

Major aquifers, all of which aredrinking-water sources, are generallydeeper than the shallow ground waterstudied and had distinctly lowerdetection frequencies of herbicides.Only 3 of 33 aquifers sampled hadamong the highest ranked detectionfrequencies, and none of the wellssampled in major aquifers hadherbicide concentrations that exceededdrinking-water standards or guidelines.

Ground-water contamination,compared to stream contamination, ismore strongly governed by soil andgeologic conditions, and each well isuniquely affected by sources ofpesticides and flow conditions in itsimmediate vicinity. Local variability

in these conditions can result indegradation of water quality in one ora few wells, even if most wells arenot affected. The greatest frequenciesof herbicide detection in majoraquifers occurred in vulnerablesettings. The three aquifers with thehighest frequencies of detection were(1) the Platte River Alluvial aquifer inthe Central Nebraska Basins, which isshallow and overlain by permeablesandy soils, (2) the Upper Floridanaquifer in the Appalachicola-Chattahoochee-Flint River Basin,which is a limestone formation whereflow rates are high, and (3) a shallowlimestone aquifer in the LowerSusquehanna River Basin.

Herbicides in shallow ground water were most commonbeneath agricultural areas

73

PEST

ICID

ES

Urban areas

Albemarle-PamlicoDrainage

Connecticut, Housatonic,and Thames River Basins

Upper SnakeRiver Basin

Central NebraskaBasins

Apalachicola-Chattahoochee-Flint River Basin

LowerSusquehanna

River Basin

Tampa •

Dallas-Ft Worth •

• Atlanta

• Albuquerque

• Las Vegas

• Virginia

Beach

• Indianapolis

• Reno-Sparks • Denver

• Harrisburg

• SpringfieldAlbany •

• Portland

37

7

Shallow ground water inagricultural areasThe highest detection frequencies occurred where use is moderate to high and where soil and geologic conditions promote rapid infiltration.

Shallow ground water inurban areasOnly two urban areas had detection frequencies in the highest 25 percent of all ground-water studies.

Major aquifersDetections were infrequent, except for a few aquifers in vulnerable settings—shallow aquifers with permeable sandy soils or limestone formations.

See p. 31 for more information about these maps

Herbicide use—in pounds per acre of agricultural land

Highest (greater than 0.461)Medium (0.162 to 0.461)Lowest (less than 0.162)No reported use

A national ranking of HERBICIDES in ground water

Bold outline indicates exceedance byone or more herbicides. Number ispercentage of wells that exceeded a a standard or guideline

Herbicide detection frequency—Each circle represents a ground-water study

Drinking-water standards or guidelines

Lowest 25 percentMiddle 50 percentHighest 25 percent

3

74PE

STIC

IDES

Insecticides were seldom found inground water but may be a concern insome areasInsecticides, in contrast to herbicides, were not detected in a number ofground-water studies and, where detected, were usually found in less than 10percent of wells. The most frequently detected insecticides in ground waterwere dieldrin and diazinon, although each was found in only 1 to 2 percent ofall wells. The relative abundance of dieldrin was unexpected because of its lowmobility in water compared with many currently used pesticides. Dieldrin,however, is one of the more mobile compounds within the historically usedorganochlorine group. Moreover, it is long-lived in the environment, whichresults in its great persistence in the ground-water flow system.

Although insecticides were much less common than herbicides inground water, they exceeded drinking-water standards or guidelines moreoften. In all but one well where exceedances occurred, dieldrin was theinsecticide that exceeded the guideline. The guideline used for dieldrin is aUSEPA Risk Specific Dose of 0.02 µg/L, which corresponds to a cancer risklevel of 1 in 100,000. The wells that exceeded the Risk Specific Dose fordieldrin were mainly wells tapping shallow ground water that is not used forhuman consumption.

The infrequent but potentially important occurrences of dieldrin in somewells may be the result of local contamination of individual wells. Thecombination of relatively shallow ground water and pesticide use in thevicinity of wells increases the likelihood that some wells will have flowpathways that allow pesticides to move from the land surface to the well,sometimes down the borehole itself.

DIELDRIN PERSISTS IN SHALLOW URBAN GROUND WATER

In the Apalachicola-Chattahoochee-Flint River Basin, insecticide concentrations in ground-water samples generally were less than current drinking-water standards or guidelines.However, dieldrin concentrations in water samples collected during 1994–95 from 5 of 37shallow wells and springs in Metropolitan Atlanta exceeded the USEPA Risk Specific Doseof 0.02 µg/L, which corresponds to a cancer risk level of 1 in 100,000. Dieldrin and aldrin,which breaks down to dieldrin in the environment, had been used on agricultural land priorto 1975 and for structural termite control until 1987.(36) Although this ground water is not usedas a source of drinking water, the presence of dieldrin in ground-water samples collectedseveral years after being banned is indicative of the compound’s persistence in soils andground water and its potential to be a problem in some wells.

Daniel J. Hippe

75

PEST

ICID

ES

Urban areasApalachicola-

Chattahoochee-Flint River Basin

Tampa •

Dallas-Ft Worth •

• Atlanta

• Albuquerque

• Las Vegas

• Virginia

Beach

• Indianapolis

• Reno-Sparks • Denver

• Harrisburg

• SpringfieldAlbany •

• Portland

5

14

7

4

8

57

5Shallow ground water inagricultural areasDetection frequencies ranked low to moderate in most studies.

Shallow ground water inurban areasDetection frequencies ranked high in urban areas compared with other study areas but still were low compared to herbicides. Although aldrin and dieldrin have not been used for years, dieldrin was the most frequently detected insecticide.

Major aquifersMost major aquifers ranked low to moderate in detection frequency and, only one well exceeded a drinking-water standard or guideline (dieldrin).

See p. 31 for more information about these maps

Insecticide use—in pounds per acre of agricultural land

Highest (greater than 0.086)Medium (0.033 to 0.086)Lowest (less than 0.033)No reported use

A national ranking of INSECTICIDES in ground water

Bold outline indicates exceedance by one or more insecticides. Number ispercentage of wells that exceeded a a standard or guideline

Insecticide detection frequency—Each circle represents a ground-water study or major aquifer

Drinking-water standards or guidelines

Lowest 29 percent (all with no detections)Middle 46 percentHighest 25 percent

5

76PE

STIC

IDES

PESTICIDES USUALLY OCCUR AS MIXTURESPesticides usually occur in mixtures of several compounds rather thanindividually, but most of our experience and research on environmental effectsis based on exposure to individual compounds. Therefore, it is vital that weunderstand and document the occurrence and composition of commonlow-level mixtures and begin to evaluate their effects.

More than 50 percent of all stream samples contained five or morepesticides, and nearly 25 percent of ground-water samples contained two ormore pesticides. In the Central Columbia Plateau, for example, 66 percent ofground-water samples with detections contained more than one pesticide, mostcommonly in shallow monitoring wells. The most common mixtures werefound more than twice as frequently in streams than in ground water, exceptfor the atrazine-DEA combination.

Mixtures of currently used pesticides in stream water may occur incombination with mixtures of organochlorine insecticides in bed sediment andfish. Moreover, about 50 percent of bed-sediment and fish samples withpesticide detections contained compounds from two or more of the majororganochlorine groups.

COMMON PESTICIDE MIXTURES IN WATER

The composition of the most common pesticide mixtures differs between urban and agriculturalareas and between agricultural areas with different crops and pests. In urban areas, simazine andprometon were the most common pesticides found together, whereas atrazine, DEA (deethyl-atrazine), and metolachlor were the most common compounds found in mixtures from agriculturalareas. Mixtures containing both herbicides and insecticides were a common occurrence in urbanstreams. More than 10 percent of urban stream samples contained a mixture of at least 4 herbicidesplus the insecticides diazinon and chlorpyrifos.

AtrazineDEA

AtrazineDEA

Metolachlor

AtrazineDEA

PrometonMetolachlor

AtrazineDEA

PrometonMetolachlor Simazine

AtrazineDEA

PrometonMetolachlorSimazineAlachlor

0

10

20

30

40

50

60

70

FRE

QU

EN

CY

OF

DE

TEC

TIO

N A

T O

R A

BO

VE

0.01

MIC

RO

GR

AM

PE

R L

ITE

R, i

n pe

rcen

t

Agricultural land

SimazinePrometon

PrometonAtrazineSimazine

AtrazineSimazinePrometonDiazinon

AtrazineSimazinePrometon

MetolachlorDiazinon

AtrazineSimazinePrometon

MetolachlorDiazinon

Chlorpyrifos

Urban land

Surface water

Ground water

Differences in occurrence and behavior of pesticidescomplicate evaluation of potential effects

77

PEST

ICID

ES

BREAKDOWN PRODUCTSCAN BE IMPORTANTOnce released into the environment,pesticides undergo a series ofchemical and biological reactionswhereby the original pesticide breaksdown into intermediate compounds,and eventually into carbon dioxideand other harmless compounds. Somebreakdown products are short-lived,whereas others persist for years ordecades. Little is known about theoccurrence of many pesticidebreakdown products, and even less isknown about their effects on humanhealth and aquatic life.

Of the thousands of possiblebreakdown products, few havebeen looked for in streams orground water.(6,43) Some are lesstoxic than their parent compounds,whereas others have been found tohave similar or even greater toxicities.

Only seven breakdown productswere analyzed in water samples fromthe first 20 Study Units: 2,6-diethyl-aniline (parent pesticide, alachlor),3-hydroxy-carbofuran (carbofuran),aldicarb sulfone and aldicarbsulfoxide (aldicarb), DDE (DDT),alpha-HCH (lindane), and DEA(atrazine). Of the parent pesticides,atrazine is the most heavily used, andboth it and DEA were widespread instreams and ground water across theNation. The two were found together

HIGH DIAZINON CONCENTRATIONS INTHE SAN JOAQUIN RIVER WERECOMMON FOLLOWING WINTERAPPLICATION

Diazinon concentrations in the SanJoaquin River near Vernalis, California,exceeded concentrations shown to betoxic to aquatic life during January andFebruary 1993—following the winterapplication of diazinon, a dormant sprayapplied to control wood-boring insects inalmond orchards in the San Joaquin-Tulare Basins.

ATRAZINE AND ITS BREAKDOWNPRODUCTS WERE DETECTEDTHROUGHOUT THE YEAR IN THEWHITE RIVER BASIN

0M J J A

1993N D J F M A S

1.2

0.8

1.0

DIA

ZIN

ON

, in

mic

rog

ram

s p

er li

ter

0.2

0.4

0.6

O

San Joaquin River near Vernalis, California

breakdown

F M A M J J A1996

0

15

5

10

CO

NC

EN

TR

AT

ION

,

AtrazineAtrazine

Sugar Creek,White River Basin,Indiana

in m

icro

gra

ms

per

lite

r

products

in about 35 percent of stream samplesand about 25 percent of ground-watersamples from agricultural areas.

With few exceptions, most of theother breakdown products were foundin fewer than 1 percent of samples ineach of the Study Units. However,several breakdown products ofalachlor and metolachlor have beenfrequently found in other studies,often at much higher concentrationsthan the parent pesticide.(34, 44) AsNAWQA evolves, more completeanalyses of breakdown products arebeing added as analytical methodsand budget constraints allow.

CONCENTRATIONS INSTREAMS FOLLOW STRONGSEASONAL PATTERNSSeasonal patterns in concentrationsand occurrences of pesticides inagricultural streams, which tend torepeat each year, correspond topatterns in use and streamflow,including contributions from groundwater. Generally, the number andconcentrations of herbicides found inmost agricultural streams werehighest from April through July,whereas insecticides occurred morevariably throughout the summer. Thespring herbicide pulse was commonlyobserved in corn-growing areas andother agricultural areas shortly after

herbicide application, when herbicideswere transported to streams in runoffinduced by spring rain and irrigation.In some parts of the Nation, otherpatterns can occur. For example,some insecticides, such as diazinon inthe San Joaquin-Tulare Basins, havepatterns of high concentrations duringthe winter, resulting from the use ofdormant sprays on orchards.Differences in patterns also mayresult from local water-managementpractices, including the timing ofreservoir storage and water use, thetiming of runoff from agriculturalfields due to irrigation or storms, orground-water contributions duringperiods of low streamflow. Seasonalpatterns need to be characterizedand understood because theydictate the timing of highconcentrations in drinking-watersupplies and the times whenaquatic organisms may be exposedto high concentrations duringcritical stages of their life cycle. Forexample, some water suppliers reducetheir use of certain surface-watersupplies during spring runoff.

78PE

STIC

IDES

Trends in pesticide concentrationsfollow changes in usePesticides in streams and ground water change over time as the types andamounts of chemicals in use change. With the exception of organochlorineinsecticides, however, consistent data that are adequate for assessing long-termtrends have not been widely collected. Examples for the organochlorineinsecticides and recent changes in herbicide use illustrate the importance oftracking such trends.

ORGANOCHLORINE INSECTICIDES HAVE DECREASEDA striking historical trend is the reduction in concentrations oforganochlorine insecticides in sediment and fish followingrestrictions on their use, yet they continue to occur at levels ofconcern. This trend is evident in sediment cores from lakes andreservoirs and by comparison of NAWQA findings to historicalconcentrations in fish measured by the U.S. Fish and Wildlife Service(USFWS).

As sediment erodes from the land surface over time, it is deposited inlayers on the bottom of lakes and reservoirs. Age-dated sediment coresthat penetrate these layered deposits can be used to track trends insediment-associated contaminants within the drainage basin.Concentrations of total DDT (DDT plus breakdown products DDE andDDD) in sediment cores from lakes and reservoirs reflect high historicaluse of DDT followed by a ban in 1972. DDT concentrations peakedduring the 1960s, which coincides with its peak use as an insecticide.Total DDT concentrations in sediment have decreased since 1972 in allsampled lakes and reservoirs that drain urban and agricultural areaswithin the United States.(45)

Unlike DDT, aldrin and chlordane were used for termite control untilthe late 1980s, long after their agricultural uses were cancelled in theearly 1970s. Chlordane and dieldrin concentrations peaked in manyagricultural areas during the 1970s, and decreased thereafter. In someurban lakes and reservoirs, however, such as White Rock Lake in theTrinity River Basin, chlordane and dieldrin peaked much later, probablyas a result of continued urban use during the 1980s. This watershed isdominated by new (post-1960) urbanization.(46)

Concentrations of DDT, chlordane, and dieldrin in whole fish havedeclined nationally since the 1970s. To assess trends in DDTconcentrations, NAWQA data for streams and rivers with mixed landinfluences were compared with similar data from 1969 to 1986 collectedby the USFWS National Contaminant Biomonitoring Program.(47) TotalDDT concentrations in fish declined markedly from 1969 to the present.The declines were greatest during the early 1970s, with concentrationssince the mid-1980s showing a slower decline or even a plateau.

Despite the observed national decline in total DDT concentrations,the detection frequency for total DDT in whole fish from major riversremains high (94 percent in the 1990s), and locally contaminated areaspersist. This is probably caused by the presence of total DDT in thestreambed and continued inputs of total DDT to hydrologic systems ascontaminated soils erode into streams.

0

40

20

CH

LO

RD

AN

E,

in m

icro

gra

ms

per

kilo

gra

m

0

5,000

CH

LO

RD

AN

E U

SE

,in

met

ric

ton

s

0

2

1D

IEL

DR

IN,

in m

icro

gra

ms

per

kilo

gra

m

0

10,000

5,000

AL

DR

IN P

LU

S D

IEL

DR

INU

SE

, in

met

ric

ton

s

Urban use

Nondetections

0

60

20

40

TOTA

L D

DT,

in m

icro

gra

ms

per

kilo

gra

m

1900 20001925 1950 1975

0

50,000

25,000

DD

T U

SE

,in

met

ric

ton

sTotal use in United States

Agricultural use

Agricultural use

Increasing sediment depth

DATE OF LAYER IN SEDIMENT CORE

ORGANOCHLORINE PESTICIDE TRENDS INSEDIMENT CORES FROM WHITE ROCK LAKEIN DALLAS, TEXAS, 1996 (13-16, 48-50)

TRENDS IN TOTAL DDT CONCENTRATIONSIN WHOLE FISH FROM U.S. RIVERS(USFWS(47) and NAWQA data)

1969

1979

1989

0

500

1000

1500

2000

2500

3000

3500

TOTA

L D

DT

CO

NC

EN

TR

AT

ION

IN

WH

OL

E F

ISH

, in

mic

rog

ram

s p

er k

ilog

ram

wet

wei

gh

t

Median of all sitesRange for 90 percent of sites(0-90th percentiles)

79

PEST

ICID

ES

RECENT CHANGES IN HERBICIDE USE HAVE BEEN RAPIDLYREFLECTED IN STREAMSFew studies have documented long-term trends in water concentrations ofcurrently used pesticides with sufficient consistency in locations, timing, andmethods to be conclusive. Recently, however, a major change has occurred inherbicide use patterns for corn and soybeans, with a new compound, acetochlor,partially replacing alachlor beginning in 1994. The increase in acetochlorconcentrations and decrease in alachlor concentrations in the White River from1994 through 1996 illustrate the direct connection between chemical use andconcentrations in streams and in the major rivers into which they flow.

TRENDS IN HERBICIDE USE WITH TIME ARE REFLECTED IN STREAM QUALITY

Alachlor concentrations in streams in the White River steadily declined from 1992 through 1996 and corresponded with a decline inalachlor use in the basin. Application of acetochlor, a corn herbicide registered for use in 1994, has partially replaced the use ofalachlor in the basin. Acetochlor was detected at only trace concentrations during the 1994 growing season. By 1996, acetochlorwas commonly detected in the White River, where a peak concentration of about 2 µg/L was measured.

STREAMS AND GROUND WATER RESPONDDIFFERENTLY TO CHANGEGenerally, as pesticide use in a basin changes, concentrations in streamsquickly reflect these changes. In ground water, however, responses to trends inpesticide-use patterns will be highly variable depending on the nature of theflow system and variability in flow pathways, well depth, and other factors.For the most part, changes in concentrations of pesticides in ground water aremuch slower than in streams, and responses of ground water to changing usecan be delayed for years or decades in some systems.

J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D1992 1993 1994 1995 1996

0

2.5

0.5

1.0

1.5

2.0

AL

AC

HL

OR

CO

NC

EN

TR

AT

ION

, in

mic

rog

ram

s p

er li

ter

0

2.5

0.5

1.0

1.5

2.0

AC

ET

OC

HL

OR

CO

NC

EN

TR

AT

ION

, in

mic

rog

ram

s p

er li

ter

Acetochlor in White River

Alachlor in White River

1992 1993 1994 1995 19960

3

1

2

AC

TIV

E

ING

RE

DIE

NT

AP

PL

IED

,in

th

ou

san

d t

on

s Alachlor use in Indiana

Acetochlor use in Indiana

Ace

toch

lor

not

regi

ster

ed fo

r us

e

Ace

toch

lor

not

regi

ster

ed fo

r us

e

Open symbol indicates acetochlor was not detected in sample

Open symbol indicates alachlorwas not detected

in sample

No data

2.6