Pesticides in Stream Sediment and Aquatic Biota Distribution, Trends, And Governing Factors - Chapter 7 ppsx

CHAPTER 7

Summary and Conclusions

This book reviews more than 400 studies of pesticides in bed sediment and aquatic biota of
United States rivers conducted over the last three decades. Considered together, the existing
literature provides a basis for at least partial assessment of the extent of pesticide contamination
of rivers and estuaries in the United States, although there are areas where our current
understanding is incomplete.
The majority of studies reviewed were monitoring studies, ranging in scale from local to
national. There was little consistency among the studies reviewed in terms of site selection
strategy, sample collection methods, species of organisms sampled, tissue type analyzed, and
analytical detection limits. In contrast, there was great consistency in the types of pesticides that
were target analytes in bed sediment and aquatic biota studies—the vast majority of studies
focused on organochlorine insecticides. Since the 1960s, there has been a decrease in process-
type studies (or field experiments) designed to assess the general environmental fate and
persistence of individual pesticides, and an increase in monitoring-type studies that focused on
analytes known to be present in bed sediment and aquatic biota (as the environmental fate of
hydrophobic pesticides became more generally understood).
Results of the hundreds of state and local monitoring studies are difficult to compare
because of the large variability in study design (which includes factors such as site selection,
sampling methods, and analytical detection limits). However, several national programs have
been conducted that constitute individual data sets large enough to provide a nationwide
perspective on pesticide occurrence and distribution. Together, these national monitoring studies
have provided good geographic coverage of pesticides in bed sediment and aquatic biota of
United States rivers and estuaries. Temporally, the national studies have provided the fewest data
for freshwater sediment, which was sampled only during the 1970s. The focus of state and local
monitoring studies was fairly evenly divided between bed sediment and aquatic biota.
Geographically, state and local monitoring efforts have been heaviest in the Great Lakes region,
along the Mississippi River, and in western states, especially California. In aquatic biota

monitoring studies, pesticide residues were more often measured in fish than in other aquatic
organisms, with other invertebrates and mollusks placing a distant second and third. In studies
that sampled fish, there was little consistency in the types of tissue analyzed (such as whole fish,
fillets, liver, or other organs). No national studies measured pesticide residues in plants, algae, or
amphibians, whereas a number of both local monitoring studies and process-type studies did.
© 1999 by CRC Press LLC

Analytical detection limits varied among studies and were not always reported. In general,
detection limits tended to be lower and to be reported more frequently in more recently published
studies. Changes in analytical methods over time (especially the change from packed-column to
capillary column gas chromatography) tended to lower detection limits.
The monitoring studies reviewed show that a large number of pesticides have been detected
in bed sediment and aquatic biota at some time over the last 30 years. Forty-one of 93 target
analytes (44 percent) were detected in bed sediment in one or more studies, and 68 of 106
analytes (64 percent) were detected in aquatic biota in one or more studies. Most of the target
analytes detected were organochlorine insecticides, by-products, or transformation products,
despite the fact that most organochlorine insecticides were banned or severely restricted by the
mid-1970s. This reflects both the environmental persistence of these compounds and the bias in
the target analyte list (which typically was limited to organochlorine compounds). DDT and
metabolites, chlordane compounds, and dieldrin were the most commonly detected pesticide
analytes in both bed sediment and aquatic biota. Other organochlorine insecticides that
sometimes were detected included endosulfan compounds, endrin and metabolites, heptachlor
and heptachlor epoxide, mirex, lindane,

α

- and

β

-HCH, kepone, methoxychlor, and toxaphene.
Besides the organochlorine insecticides, a few compounds in other pesticide classes were
detected in some studies. Most of these pesticides contained chlorine or fluorine substituents and
were intermediate in hydrophobicity. Examples in bed sediment included the herbicides ametryn,
dacthal, 2,4-DB, dicamba, and diuron; the organophosphate insecticide zytron; and the
fungicide/wood preservative pentachlorophenol. In aquatic biota, examples included the herbi-
cides dacthal, oxadiazon, and trifluralin; the organophosphate insecticides zytron and chlor-
pyrifos; and the fungicide/wood preservative pentachlorophenol and (its metabolite) pentachloro-
anisole. Of pesticides from other chemical classes that were analyzed in bed sediment or aquatic
biota, most were targeted at relatively few sites nationwide, and those sites generally came from
one or a few studies. Aggregate detection frequencies (calculated by combining data from the
studies reviewed) are not necessarily representative of freshwater resources in the United States.
If a given pesticides is absent from the list of pesticides detected in bed sediment or aquatic
biota, it does not necessarily mean that the individual pesticide is not present in these sampling
media, but that it may not have been looked for in the studies reviewed. Even for those
compounds with zero or few detections, the results may have been biased by sampling location
or study design.
The monitoring studies reviewed suggest that pesticides were detected more often in
aquatic biota than in bed sediment. More pesticides (both the total number and the percentage of
those targeted) were detected in aquatic biota than in bed sediment and the cumulative detection
frequencies for a given pesticide generally were higher. Moreover, in direct comparison between
benthic fish and associated surficial sediment samples collected by the National Oceanic and
Atmospheric Administration’s (NOAA) National Status and Trends (NS&T) Program (Benthic
Surveillance Project), detection frequencies and concentrations (dry weight) of most hydro-
phobic organic contaminants tended to be higher in estuarine fish liver than in associated
sediment. The exceptions were aldrin and heptachlor, both of which are metabolized rapidly in
aquatic biota. Their transformation products (dieldrin and heptachlor epoxide, respectively)
showed the expected pattern, that is, they were found at higher levels in fish liver samples than in
associated sediment.
© 1999 by CRC Press LLC

About 1 billion pounds of pesticides currently are used each year in the United States in a
wide variety of agricultural and nonagricultural applications. Total pesticide use and the number
of different chemicals applied have grown steadily since the early 1960s, when the first reliable
records were kept. More than 75 percent of total pesticide use is applied in agriculture. However,
pesticides also are applied on lawns and gardens, to control nuisance insects (both indoors and
outdoors), in subterranean termite control, in landscape maintenance, to control public-health
pests, in industrial settings, in forestry, along roadways and rights-of-way, and in direct
application to aquatic systems. Although quantitative estimates of pesticide use are available for
agriculture, there is relatively little quantitative data available on nonagricultural uses of
individual pesticides.
Agricultural use of insecticides declined during the 1980s. Also, there was a pronounced
change in the types of insecticides used from the 1960s to the 1980s as organochlorine
insecticides were replaced by organophosphate, carbamate, and other insecticides. The only four
organochlorine insecticides used in agriculture in 1988 (endosulfan, dicofol, methoxychlor, and
lindane) are less hydrophobic and less persistent than DDT and the other restricted compounds
in the organochlorine family. In contrast, agricultural use of herbicides has increased markedly
since the 1960s. Most herbicide classes contributed to this rise, except the chlorophenoxy acid
herbicides. Over 60 percent of the herbicides used in agriculture in 1988 were triazine, amide, or
carbamate herbicides.
The principal pesticides used in and around the home and garden today tend to differ from
those used in agriculture. However, almost all of the restricted organochlorine insecticides
(except endrin) once had extensive urban use. Several organochlorine insecticides were used in
homes and gardens in 1990, but they accounted for fewer than 4 percent of products or outdoor
applications. Home and garden insecticide use was dominated by pyrethroid insecticides,
pyrethroid synergists, and the insect repellent diethyltoluamide.
As a first approximation, the geographic distribution of pesticides detected in bed sediment
and aquatic biota may be expected to relate directly to the quantities of pesticides applied. To
this end, aggregate detection frequencies (which represent a crude composite detection
frequency from the 1960s to 1993) were calculated for individual pesticides on the basis of

combined data from all of the monitoring studies reviewed. These aggregate detection
frequencies were compared with estimates of national agricultural use in either 1966 (for
organochlorine insecticides) or 1988 (for currently used pesticides). There was no clear relation
in either case. Factors that may obscure this relation include pesticide use in nonagricultural
settings, sources other than pesticide application (such as hexachlorobenzene from industrial
sources), the influence of physical and chemical properties, and bias introduced by combining
data from studies that have different study designs and methods.
To increase the chances of detecting a relation between pesticide occurrence and pesticide
use, the variability in both data sets was reduced: Pesticide occurrence data were taken from a
single national program at a time, and pesticide use estimates by agricultural production region
or by county were employed. For several organochlorine insecticides, there was a moderately
strong relation between detection in both bed sediment and whole fish from major United States
rivers during the 1970s and agricultural use by region in 1966. This was observed for total DDT
and total dieldrin in bed sediment sampled by the U.S. Geological Survey (USGS) and the U.S.
Environmental Protection Agency (USEPA) during 1972–1982, and for total DDT, total dieldrin,
chlordane, and total heptachlor in whole fish sampled by the Fish and Wildlife Service (FWS)
© 1999 by CRC Press LLC

during 1976–1979. In all cases, however, higher than expected concentrations were observed in
some parts of the United States, possibly due to intensive nonagricultural use, heavy local use (in
excess of average regional use), incidental industrial release, and atmospheric deposition. The
geographic distributions of some organochlorine insecticides in whole fish measured by the FWS
were associated with specific uses such as corn production (dieldrin, chlordane, heptachlor) or
red fire ant control (such as mirex) or with manufacturing inputs (such as kepone and mirex).
In estuarine bed sediment and mollusks sampled by NOAA, the geographic distributions of
most organochlorine insecticides (during 1984–1989) were related to human population levels.
Contaminated coastal or estuarine sites sampled by NS&T Program in 1986 were sometimes, but
not always, associated with contaminated inland sites sampled by the FWS’s National
Contaminant Biomonitoring Program (NCBP) also in 1986.
Relatively few national studies monitored currently used pesticides in bed sediment or

aquatic biota. Although the USGS and USEPA analyzed bed sediment from major United States
rivers (during 1975–1980) for a few organophosphate insecticides, and chlorophenoxy acid and
triazine herbicides, there were too few detections of any pesticides other than the organochlorine
insecticides to permit association with pesticide use. In fish sampled by the U.S. Environmental
Protection Agency (1986–1987), three organochlorine insecticides that were still used in
agriculture in 1988 (dicofol, methoxychlor, and lindane) had higher detection frequencies than
might be expected on the basis of agricultural use alone. These organochlorine insecticides,
although less persistent than DDT and other canceled organochlorine insecticides, still have
longer environmental lifetimes than most other chemical classes. They also can be applied in
residential or urban settings. Several currently used pesticides were detected in association with
agricultural drainages in three large-scale studies that analyzed for them in fish tissues. The
studies were the USEPA’s National Study of Chemical Residues in Fish (NSCRF), the NCBP,
and California’s Toxic Substances Monitoring Program (TSMP). The currently used pesticides
that were associated with agricultural drainages were chlorpyrifos (NSCRF and TSMP); dacthal
(NCBP and TSMP); diazinon, endosulfan, and parathion (TSMP); and dicofol and trifluralin
(NSCRF).
Insufficient data are available at this time to assess trends for organochlorine insecticides in
either river or estuarine bed sediment. Some national studies were of sufficient duration,
however, to provide information on organochlorine insecticide trends in fish and shellfish
contamination. In whole freshwater fish sampled by the FWS, hexachlorobenzene residues
appeared to decline in frequency and magnitude nationwide from 1976 to 1986. Concentrations
of dieldrin, total DDT, endrin, and

α

-HCH appeared to decline during the 1970s, then level off in
the early 1980s. Toxaphene residues in whole fish appeared to peak nationally around 1980,
declining afterwards through 1986. For most organochlorine insecticides, high concentrations
appear to be persisting in some areas in the United States. In coastal and estuarine biota, DDT
residues have declined nationally in the last two decades, especially during the 1970s. There is

evidence for declining dieldrin residues at some marine sites, but it remains to be determined
whether there is a nationwide trend for coastal and estuarine areas, especially for dieldrin
residues in fish. Toxaphene residues in coastal and estuarine biota appear to have declined
nationally, although local “hot spots” remain. For estuarine biota, insufficient data exist to
evaluate national trends in residues of chlordane, endrin, total heptachlor, and
hexachlorobenzene. Ongoing collections of coastal and estuarine bivalve mollusks and benthic
© 1999 by CRC Press LLC

fish (as well as bed sediment) under the NS&T Program are expected to provide additional data
for chlordane, dieldrin, and lindane.
Of currently used pesticides, long-term national data are available only for lindane, one of
the four moderately persistent organochlorine insecticides still used in agriculture (in 1988). In
whole freshwater fish sampled by the FWS, lindane concentrations appeared to decline in the
late 1970s, then level off in the 1980s. Several other currently used pesticides (1988) were
analyzed for a limited period of time in one or more national programs, but not long enough to
permit assessment of trends.
Sources of pesticides to bed sediment and aquatic biota include agricultural and
nonagricultural applications and atmospheric contamination. The frequent detection of
organochlorine insecticides in bed sediment and aquatic biota in remote areas of the United
States and the world has been attributed to atmospheric transport from high-use or contaminated
areas, and subsequent deposition. Agriculture is undoubtedly a major contributor to pesticide
contamination of bed sediment and aquatic biota. There is considerable evidence that the source
of DDT in bed sediment and fish in many rivers in the United States is from past insecticide use
in agriculture. Many agricultural areas still contain soil residues of hydrophobic, persistent
pesticides that were applied in the 1970s or earlier. Field monitoring studies indicate that DDT
half-lives in soil are on the order of 15 years or longer, that soil contaminated with total DDT
may enter drains as a result of normal field and irrigation practices (such as land leveling), and
that the rate of DDT degradation appears to increase slightly once it enters a river or other
waterway. DDD and DDE, the most common transformation products of DDT, also are resistant
to degradation. High residues of total DDT (the sum of DDD, DDE, and DDT) and a high ratio

of DDT/total DDT in soil, bed sediment, or biota samples do not necessarily indicate recent use
of DDT. A high ratio of DDT/total DDT in river sediment or aquatic biota probably indicates
that the residues recently entered the hydrologic system, such as by erosion of DDT-
contaminated soil.
Existing monitoring data suggest that various nonagricultural uses of pesticides may
contribute to residues of some pesticides in bed sediment and aquatic biota, at least in some
areas. Pesticides used in forestry and in urban areas, like those used in agriculture, have changed
over time. Of the pesticides currently used in forestry, only a few have been targeted in any
studies that monitored bed sediment and aquatic biota (2,4-D, picloram, carbaryl, and
glyphosate), and these were detected at very few (less than 10 percent) sites. Some process-
oriented studies were conducted in which pesticides and their transformation products were
measured in forest streams following known applications of those pesticides. Results indicate
that the behavior of pesticides in forest streams is in agreement with their behavior in
agricultural streams. DDT and metabolites appear to have a long residual lifetime in bed
sediment of forest streams. Concentrations of less hydrophobic pesticides may be detected in
bed sediment and aquatic biota for some time following application, but would be expected to
decline in these media more rapidly than the organochlorine insecticides.
In urban areas, organochlorine insecticides have been largely replaced over time by
insecticides in other classes. However, several organochlorine insecticides were used around
homes and gardens in 1990, including dicofol, chlordane, endosulfan, heptachlor, lindane, and
methoxychlor. Many monitoring studies have reported the frequent detection of organochlorine
pesticides in bed sediment, aquatic biota, and water in rivers in urban areas. The actual sources
of these pesticide residues are not completely known, however, because the organochlorine
© 1999 by CRC Press LLC

insecticides had both urban and agricultural uses, and many urban sites are also located near
agricultural areas. Most pesticides found in bed sediment and aquatic biota in such urban areas
probably derive from both agricultural and urban sources, with their relative importance varying
by location and chemical compound.
Taken together, the studies reviewed in this book provide a good understanding of

processes affecting the fate, transport, and transformation of pesticides within the hydrologic
system and how these relate to the physical and chemical properties of a given pesticide.
Pesticides may enter surface-water bodies via surface runoff and drainage, ground-water
discharge (for hydrophilic pesticides), atmospheric deposition (for volatile pesticides), or direct
application to the water body. Once a pesticide enters a hydrologic system, its environmental fate
is controlled by the physical and chemical properties of the pesticide and by environmental
conditions in the hydrologic system. Environmental processes that occur may be categorized as
phase-transfer, transport, or transformation processes. Pesticides with low water solubility and
high

n

-octanol-water partition coefficients (

K

ow

, a measure of lipophilicity) tend to associate with
organic matter in the hydrologic system, including dissolved organic matter, particulates, and
biota. Pesticides associated with particulate matter in the water column may be deposited to bed
sediment. Once there, pesticides may be remobilized by chemical diffusion or biological and
physical mixing or both. Remobilized pesticides can reenter the water column and be carried
downstream and(or) redeposited again. Aquatic biota may take up pesticides by partitioning
(diffusion through surface membranes) from water, by ingestion of contaminated food or
particles, or by direct contact with sediment. Once the pesticide is taken up by living biota tissue,
it can be stored, metabolized, or depurated. Final removal of the pesticide from bed sediment or
biota in a hydrologic system can occur either by physical removal (such as burial of bed
sediment in long-term depositional zones, sediment discharge to the oceans, or consumption of
contaminated fish by humans or other terrestrial animals) or by chemical transformation (such as

biodegradation or hydrolysis).
Bioaccumulation in an organism is the net effect of competing processes of uptake and
elimination. Contaminant accumulation by aquatic biota is affected by biological factors (such as
species, sex, body size, age, reproductive state, lipid content, metabolic capability, growth rate,
blood flow, and gill ventilation volume), chemical characteristics (such as molecular size and
shape, solubility in lipid and water, and chemical stability), and environmental conditions (such
as temperature, pH, salinity, concentrations of dissolved organic matter and particulates, and
degree of water oxygenation). There has been some disagreement in the literature as to whether
biomagnification (the process whereby the tissue concentrations of a chemical increase as it
passes up the food chain through two or more trophic levels) occurs in hydrologic systems. An
opposing view has held that contaminant accumulation by aquatic organisms can be described
using equilibrium partitioning theory, in which contaminant concentrations in water, blood, and
tissue lipids approach equilibrium; regardless of the mechanism of uptake, the contaminant
partitions into and out of these phases according to its relative solubility (bioconcentration). The
relative importance of contaminant uptake from the diet, and from water via partitioning, has
also been debated in the literature. The relative importance of dietary uptake and uptake from
water by partitioning, and of biomagnification and bioconcentration, in hydrologic systems
appears to depend on the chemical, as well as on the type of organisms involved. From a review
of field, laboratory, and model ecosystem studies that assessed bioconcentration, dietary uptake,
and potential biomagnification of organic chemicals, it appears that biomagnification may occur
© 1999 by CRC Press LLC

under conditions of low water concentration for compounds of high lipophilicity, high
persistence, and low water solubility. Biomagnification is most likely to occur for chemicals with
log

K

ow

values greater than 5 or 6. The predominant route of uptake also appears to depend on
the organism involved. At one extreme are air-breathing vertebrates such as sea birds, seals, and
whales. These organisms have no external surface for rapid exchange, such as gills in fish, so
that contaminant accumulation is by biomagnification. At the other extreme are autotrophic
organisms that draw their food from dissolved components in water; for such organisms,
bioconcentration is the uptake mechanism. For intermediate organisms, both mechanisms
probably occur, with their relative importance depending on the organism, the chemical, and
various other factors.
The process of biomagnification initially was believed to result from the loss of food
substances due to respiration, whereas resistant contaminants were retained by the organism.
This mechanism is now considered unlikely. It has been proposed that food digestion and
absorption from the gastrointestinal tract, accompanied by inflow of more contaminated food,
may increase the concentration of the chemical in the gastrointestinal tract relative to that in the
original food. This also increases the fugacity (a thermodynamic measure of the escaping
tendency of a chemical from a phase, equivalent to chemical activity) of the chemical in the
gastrointestinal tract relative to that in the original food. This creates a fugacity gradient (or
fugacity pump) that drives the passive diffusion of chemical from the gastrointestinal tract into
the organism, raising the fugacity of the predator over that in the prey (the consumed food). This
modified fugacity (partitioning) theory subscribes to and attempts to explain observations of
biomagnification. As such, it represents a convergence of the previously competing theories of
equilibrium partitioning and biomagnification.
Seasonal events that strongly affect the movement of pesticides to and within surface
waters also may influence the accumulation of pesticide residues in bed sediment and aquatic
biota. These include agricultural management practices (such as seasonal application of
pesticides, irrigation practices, land leveling, and tillage practices), seasonal or episodic water-
management practices (such as reservoir release and impoundment), weather-driven events (such
as precipitation, snowmelt, and strong winds), seasonal environmental conditions (such as
temperature and salinity), and, in the case of residues in aquatic biota, seasonal biological or
physiological factors (such as lipid content, reproductive cycle, and enzyme activity).
Although few studies attempted to measure seasonal changes in contaminant residues in

bed sediment, the studies that were conducted (together with current knowledge of seasonal and
episodic events that may affect contaminants in sediment) suggest that seasonal perturbations of
bed sediment can cause changes in the concentrations of hydrophobic organic chemicals.
Irrigation and rains will increase the water discharge and velocity of a river, which are likely to
resuspend some of the bed sediment. At this time, surficial bed sediment will have its lowest
total DDT concentration and the water column its highest concentration. As the irrigation or
rains subside, the water discharge and velocity will decrease, and the suspended particles will
settle out of the water to the depositional zones of bed sediment. The cycle may repeat itself,
depending on the timing of each irrigation or rain event. Reservoir release, which tends to occur
on an episodic rather than seasonal basis, would have similar effects on soil erosion, streamflow
velocity, and bed sediment disturbance. Deeper bed sediment that is generally protected from
resuspension is not affected by this seasonal cycle, but it may be moved by infrequent
catastrophic events, such as flooding.
© 1999 by CRC Press LLC

In aquatic biota, seasonal variations in pesticide residues have been observed in both
saltwater and freshwater systems. Most of the factors that affect seasonality in aquatic biota
appear to act by influencing organochlorine availability (determined by a combination of
pesticide application, industrial discharge rates, streamflow, runoff, etc.) or physiological
changes (especially those related to lipid content and the reproductive cycle), or both.
Contaminant availability and physiological condition (lipid content and reproductive stage) do
not necessarily correspond, so that multiple peaks may be observed in seasonal profiles.
Differences in the timing of seasonal maxima and minima may occur among different species at
the same location and among organisms of the same species from different locations. Given the
number of factors involved and the interaction between them, it may be difficult to attribute
seasonal profiles observed in the field to individual factors or combinations of factors. The
occurrence of seasonal fluctuations, whatever the cause, points out the importance of considering
seasonality in study design and data interpretation.
The physical and chemical properties of a pesticide influence its tendency to accumulate in
bed sediment and aquatic biota. Two types of properties are particularly important. First are those

characteristics of the pesticide that promote its association with sediment or biota (such as low
water solubility and high

K

ow

). Second, a pesticide must be environmentally persistent for it to
accumulate to substantial levels over time in bed sediment or aquatic biota. Pesticides with low
water solubility, high

K

ow

, and long soil half-life (which is the best measure of persistence that is
available for a large number of pesticides) generally were observed in bed sediment and aquatic
biota in the monitoring studies reviewed, whereas those pesticides with high water solubility, low

K

ow

, or short soil half-life generally were not. However, a number of moderately hydrophobic,
moderately persistent pesticides (such as endosulfan, lindane, chlorpyrifos, permethrin,
trifluralin, dacthal, oxadiazon, and pentachlorophenol) also were detected in bed sediment and
aquatic biota, although at lower frequencies than the more persistent organochlorine insecticides.
These intermediate compounds generally have water solubilities of 0.01–1 mg/L or log

K

ow

values of 3–5, and soil half-lives of 30–150 days. Some of these compounds (such as
chlorpyrifos and trifluralin) are very high use pesticides in agriculture today. Their presence in
bed sediment and aquatic biota in the studies reviewed suggests that other currently used
pesticides of intermediate hydrophobicity and persistence also might be found, if they were
target analytes in these media.
Most pesticides in high use during the 1990s are more water soluble and have a shorter
environmental residence time than the organochlorine insecticides. Several have a water
solubility less than 1 mg/L, a log

K

ow

value greater than 3, and a soil half-life greater than 30
days, but have not been analyzed in bed sediment or aquatic biota at many sites nationwide.
Examples are the insecticides esfenvalerate, fenthion, fenvalerate, and propargite; the herbicides
benfluralin, bensulide, ethalfluralin, pendimethalin, and triallate; and the fungicide dichlone.
These pesticides have the potential to be found in bed sediment and aquatic biota, if analyzed in
these media. However, such moderately hydrophobic, moderately persistent compounds would
probably be found at much lower frequencies than the more persistent organochlorine
insecticides (such as DDT or dieldrin). Because these intermediate pesticides are not as resistant
to degradation as DDT, concentrations of these intermediate pesticides would be expected to
decline following application or introduction to the hydrologic system. Currently used pesticides
that are intermediate in hydrophobicity and persistence, however, may reach fairly high
© 1999 by CRC Press LLC

concentrations and detection frequencies in bed sediment and aquatic biota in areas of high or

repeated use. The probability of detection would be likely to increase if study designs considered
location and time of application.
Detection of hydrophobic pesticides in bed sediment and aquatic biota serves as an
indicator that these compounds are present as contaminants in the hydrologic system. Their
biological significance is more difficult to assess. Organochlorine pesticides such as DDT have
been shown to adversely affect the survival of various organisms in laboratory tests (including
aquatic invertebrates, fish, birds, and mammals); to disrupt the reproduction of fish and birds in
the field; and to accumulate to high levels in fish-eating mammals. Pesticide contamination in
the field also has been associated with fish kills and fish diseases.
Potential effects of pesticide residues measured in the monitoring studies reviewed were
assessed by comparing the reported pesticide residue levels with appropriate standards and
guidelines designed for protection of aquatic life, wildlife, and human health. Despite the
limitations of this kind of analysis, some important points emerge. For both bed sediment and
aquatic biota, a substantial number of recently published studies (1984–1994) reported
maximum concentrations of some pesticides that exceeded applicable guidelines for protection
of aquatic organisms or wildlife. Residue levels for some pesticides at the most severely
contaminated site in each study reviewed appear to be sufficient to adversely affect benthic
organisms (in 25–50 percent of studies, depending on the pesticide), fish (10–20 percent of
studies), and fish-eating wildlife (25–75 percent of studies). The pesticides most frequently
present at levels that may cause toxicity were DDT and metabolites, dieldrin, and chlordane.
Because this analysis was based on the maximum concentrations in each study, these results
indicate potential adverse effects only at the most contaminated site in each study. They give no
information on how often, or at how many sites within these studies, these guidelines were
exceeded. Also, the concentrations observed in the monitoring studies reviewed are not
necessarily representative of the water resources in the United States. It is important to consider
site selection and other aspects of study design in interpretation of study findings.
Human exposure to organochlorine contaminants is demonstrated by their frequent, almost
universal, detection in samples of human breast milk, blood, and various tissues, including
adipose and reproductive organ tissues. These organochlorine contaminants include pesticides,
as well as polychlorinated biphenyls (PCB), polychlorodibenzodioxins (PCDD), and

polychlorodibenzofurans (PCDF). Several studies indicate that food consumption is the principal
mode of intake of hydrophobic pesticide contaminants. Fish and shellfish consumption was
reported to be a major source of human exposure (relative to other foods) for a number of
organochlorine compounds, including DDT, mirex, kepone, dieldrin, hexachlorobenzene, HCH
isomers, PCBs, and tetrachlorodibenzo-

p

-dioxins (TCDD). Moreover, organochlorine residues in
human blood have been shown to be related to consumption of contaminated fish.
To estimate potential effects on human health due to consumption of fish and shellfish
contaminated at the levels found by the monitoring studies reviewed in this book, the maximum
pesticide concentrations reported in these monitoring studies were compared with applicable
standards and guidelines for the protection of human health. These consist of USEPA tolerances
for pesticides in food, Food and Drug Administration (FDA) action levels for unavoidable
pesticide residues in food, and USEPA screening values for pesticides in edible fish (part of
USEPA’s guidance for use in setting fish consumption advisories). USEPA tolerances exist only
© 1999 by CRC Press LLC

for currently registered pesticides, and so, do not apply to most organochlorine insecticides. Both
FDA action levels and USEPA screening values exist for most of the commonly detected
organochlorine insecticides.
The maximum concentrations of most organochlorine compounds in edible fish tissues
reported in recently published (1984–1994) studies did not exceed applicable FDA action levels.
The exceptions were total chlordane, mirex, total DDT, and toxaphene. FDA action levels for
these compounds were exceeded at the most contaminated sites in 56 percent (total chlordane),
40 percent (mirex), 13 percent (total DDT), and 6 percent (toxaphene) of recently published
(1984–1994) studies. For the affected sites, if the contaminated fish were in interstate commerce,
the pesticide residues would be high enough to warrant enforcement action by FDA to remove
them.

Generally, USEPA screening values were exceeded by maximum concentrations from a
higher percentage of recently published (1984–1994) studies than were FDA action levels,
because USEPA screening values for most organochlorine insecticides are lower than FDA
action levels. This occurs because FDA action levels are based on a different risk assessment
methodology than that used by USEPA in setting guidance for use in fish consumption
advisories. Moreover, FDA action levels are not based on health considerations alone, but also
consider factors such as the economic costs of banning a foodstuff, analytical detection limits,
and the extent to which residues cannot be avoided by good agricultural or current good
manufacturing practice. USEPA screening values for several pesticides were exceeded by
maximum concentrations in at least 40 percent of recently published (1984–1994) studies: total
DDT (in 100 percent of studies), total chlordane (75 percent), dieldrin (73 percent), heptachlor
epoxide (50 percent), and mirex (40 percent). This analysis suggests that pesticide residues at the
most severely contaminated site in each of these monitoring studies would be high enough to
cause adverse human health effects if the fish from these sites were consumed at average rates by
the general adult population over a 70-year lifetime. For these five organochlorine compounds
(total DDT, total chlordane, dieldrin, heptachlor epoxide, and mirex), the expected effects would
be cancer, and the excess cancer risk may exceed 1 in 100,000 people. Adverse health risks may
be higher for sport or subsistence fishers or for sensitive subpopulations such as children or
pregnant women.
The standards and guidelines discussed above generally were based on chronic toxicity and
(for human health) carcinogenicity. It is possible, however, that additional effects on develop-
ment and reproduction in fish, wildlife, and humans may occur at concentrations below existing
standards and guidelines. This is currently a research area of great interest and controversy. More
than 50 synthetic and naturally occurring chemicals have been shown to disrupt the endocrine
systems of humans and other animals. An environmental endocrine disruptor is an exogenous
agent that interferes with the synthesis, secretion, transport, binding, action, or elimination of
natural hormones in the body that are responsible for the maintenance of homeostasis,
reproduction, development, or behavior. The list of known endocrine disruptors include several
persistent organochlorine insecticides and their transformation products (DDT, DDE, dicofol,
dieldrin, heptachlor, hexachlorobenzene, kepone, lindane and other HCH isomers, methoxychlor,

mirex, and toxaphene), as well as synthetic pyrethroids, triazine herbicides, and ethylene bis-
dithiocarbamate fungicides. A number of field studies have shown an association between
exposure to endocrine-disrupting chemicals in the environment and adverse effects on
reproduction, development, and the immune system in birds, fish, shellfish, turtles, and
© 1999 by CRC Press LLC

mammals. For the most part, these effects have been observed in areas where multiple synthetic
chemicals were present, although in some cases reproductive impairment in field populations has
been attributed to specific chemical contaminants. Studies with laboratory animals have
reproduced some of the abnormalities observed in field populations. The evidence in humans is
much more sketchy. Exposure to endocrine-disrupting xenobiotics has been suggested as
contributing to the increased incidence of breast, testicular, and prostate cancers, ectopic
pregnancies, and cryptorchidism, and to a decrease in sperm count. However, the role of
environmental endocrine disruptors in these human health trends remains uncertain. With a few
exceptions (such as diethylstilbesterol, TCDD, and DDT and its metabolite DDE),
epidemiological studies have not demonstrated causal relations between exposure to a specific
environmental chemical and an adverse effect on human health operating via an endocrine
disruption mechanism. In general, however, it is difficult to show cause and effect with
epidemiological studies that correlate known levels of chemical exposure with health effects in
humans. This is especially true for endocrine disruptors because these chemicals may have very
different effects on the human embryo and fetus than on the adult; effects are usually seen in the
offspring and not the exposed parent; the timing of exposure is critical in determining the effects;
and effects of exposure in utero may not be observed until maturity.
Animals in the environment generally are exposed to a mixture of contaminants. The
potential for interactions among mixtures of chemical contaminants exists. Possible interactions
are generally categorized as antagonistic (where the mixture is less toxic than the chemicals
considered individually), synergistic (where the mixture is more toxic than the chemicals
considered individually), and additive (where the mixture has toxicity equivalent to the sum of
the toxicities of the individual chemicals in the mixture). This is another important area of
ongoing research.

Clearly, there is a wealth of information in the existing literature on pesticides in bed
sediment and aquatic biota in United States rivers and estuaries. These are the sampling media
best suited to investigate the presence of hydrophobic contaminants in hydrologic systems. The
hundreds of national, multistate, state, and local studies that have been conducted during the last
30 years have indeed given us a national perspective on hydrophobic pesticides in United States
rivers. Still, there are gaps in our understanding of the distribution and trends in pesticide
contamination in bed sediment and aquatic biota. First, it would be worthwhile to determine
whether residues of DDT and other organochlorine contaminants have continued to decline or
have reached a plateau since last measured nationwide in bed sediment (late 1970s) or aquatic
biota (mid-1980s) of United States rivers. Although residues of most organochlorine insecticides
have declined nationally since they were banned or severely restricted, the largest declines took
place during the 1970s and early 1980s. Legitimate questions remain regarding the biological
significance of very low residues of organochlorine compounds to fish, wildlife, and humans.
Also, it will be important to continue to monitor these compounds in the contaminated areas that
still persist. Second, there is little known on the extent of contamination of currently used
pesticides, including the few organochlorine insecticides still permitted for use in agriculture.
The presence of currently used pesticides, such as chlorpyrifos, dacthal, trifluralin, and
oxadiazon, in bed sediment and aquatic biota suggests that other pesticides that are moderately
hydrophobic (water solubility less than 1 mg/L or log

K

ow

greater than 3) and moderately
persistent (soil half-life greater than 30 days) may be detected in these sampling media.
Although nationwide monitoring of these compounds may not be warranted, it may be
© 1999 by CRC Press LLC

reasonable to monitor such pesticides in areas of known high or repeated use, or in association

with specific land uses or crops. Third, the detection of pesticides at urban and suburban sites,
when targeted in the studies reviewed, combined with the little available data on urban and
residential uses of pesticides, strongly suggests a need for a systematic, large-scale study of
pesticides in the urban environment. Fourth, there is a real need for more quantitative
information on pesticide use, particularly in nonagricultural applications. Finally, there is much
to be learned about the biological significance of pesticide residues in the environment. Some
research areas are particularly compelling; for example, the effects of chemical mixtures and
potential adverse effects on aquatic life, wildlife, and human health due to endocrine disruption.
Over time, analytical methods have improved and detection limits have been lowered. There has
also been considerable progress in toxicity test method development, including use of bio-
markers to screen for potential sublethal effects (such as on reproduction). There is a better
understanding of hydrologic systems, the environmental behavior of chemicals in the
environment, and the effects of pesticides on the ecosystem and on human health. These are
formidable tools to be used by the next generation of multidisciplinary research efforts that
combine chemistry, hydrology, and ecotoxicology. A more complete national perspective of the
occurrence, distribution, and significance of pesticides in bed sediment and biota of United States
rivers is within our collective grasp.
© 1999 by CRC Press LLC