&EPA
United States
Environmental Protection
Agency
Results of the Lake Michigan Mass
Balance Study: Atrazine Data Report
December 2001
-------
U.S. Environmental Protection Agency
Great Lakes National Program Office (G-17J)
77 West Jackson Boulevard
Chicago, IL 60604
EPA 905R-01-010
-------
Results of the Lake Michigan Mass Balance Study:
Atrazine Data Report
Prepared for:
USEPA Great Lakes National Program Office
77 West Jackson Boulevard
Chicago, IL 60604
Prepared by:
Robert N. Brent, Ph.D.
Judy Schofield,
and Ken Miller
DynCorp Science and Engineering Group
6101 Stevenson Avenue
Alexandria, VA 22304
December 2001
-------
Acknowledgements
This report was prepared under the direction of Glenn Warren, Project Officer, USEPA Great Lakes
National Program Office; and Louis Blume, Work Assignment Manager and Quality Assurance Officer,
USEPA Great Lakes National Program Office. The report was prepared by Robert Brent, Judy Schofield,
and Ken Miller of DynCorp's Science and Engineering Group, with assistance from Chip McCarty,
Heather Wolfe, and Kim Malloy of DynCorp, and with significant contributions from the LMMB
Principal Investigators for atrazine. GLNPO thanks these investigators and their associates for their
technical support in project development and implementation. GLNPO specifically thanks Clyde Sweet
of the Illinois State Water Survey for his written summary of results for atrazine in atmospheric
components. GLNPO thanks Bill Sukloff of Environment Canada and Syd Allan for use and
implementation of the Research Data Management Quality System (RDMQ).
LMMB Principal Investigators for Atrazine
Clyde Sweet (atmosphere) Steve Eisenreich (tributary and open lake)
Illinois State Water Survey University of Minnesota
2204 Griffith Drive Gray Freshwater Biological Institute and
Champaign, Illinois 61820 Rutgers University
New Brunswick, New Jersey 08901
Ron Hites and Ilora Basu (atmosphere)
Indiana University
1005 East Tenth Street
Bloomington, Indiana 47405
-------
Portions of this document were developed using information presented in the following publications and
internal reports:
Miller, S.M. 1999. Spatial and Temporal Variability of Organic and Nutrient Compounds in Atmospheric
Media Collected During the Lake Michigan Mass Balance Study. M.S. thesis. University of New York at
Buffalo, Buffalo, New York. 181 pp.
Miller, S.M.; Sweet, C.W.; DePinto, J.V.; Hornbuckle, K.C. 2000. Atrazine and Nutrients in
Precipitation: Results from the Lake Michigan Mass Balance Study. Environ. Sci. Technol. 34(1): 55-
61.
Rygwelski, K.; Richardson, W.; Endicott, D. 1999. A Screening-Level Model Evaluation ofAtrazine in
the Lake Michigan Basin. J. Great Lakes Res. 25(1): 94-106.
Schottler, S.P; Eisenreich, S. 1997. Mass Balance Model to Quantify Atrazine Sources, Transformation
Rates, and Trends in the Great Lakes. Environ. Sci. Technol. 31(9): 2616-2625.
Sweet, C. 2000. Measurement ofAtrazine in Atmospheric Samples in the Lake Michigan Mass Balance
Study (LMMB), 1994-1995. Internal document. Great Lakes National Program Office.
-------
Table of Contents
Table of Contents
Executive Summary ix
Chapter 1 Project Overview 1-1
1.1 Background
1.2 Description .
1.3 Scope
1.3.1 Modeled Pollutants
1.3.1.1 Poly chlorinated Biphenyls
1.3.1.2 Trans-Nonachlor
1.3.1.3 Atrazine
1.3.1.4 Mercury
1.3.2 Other Measured Parameters
1.3.3 Measured Compartments ..
1.4 Objectives
1.5 Design . .
1.5.1 Organization
1.5.2 Study Participants
-1
-1
-2
-2
-2
-3
-4
-4
-6
-7
-8
-9
-9
-9
1.5.3 Workgroups 1-10
1.5.4 Information Management 1-10
1.5.4.1 Data Reporting 1-12
1.5.4.2 Great Lakes Environmental Monitoring Database 1-12
1.5.4.3 Public Access to LMMB Data 1-12
1.5.5 Quality Assurance Program 1-13
1.6 Project Documents and Products 1-15
Chapter 2 Atrazine Study Overview 2-1
2.1 Atrazine Introduction 2-1
2.1.1 Physical/Chemical Properties 2-1
2.1.2 Atrazine Use 2-1
2.1.3 Regulatory Background 2-2
2.1.4 Fate and Effects 2-2
2.2 Study Design 2-3
2.2.1 Description 2-3
2.2.2 Scope 2-3
2.2.3 Organization/Management 2-4
2.3 Sampling Locations 2-5
2.3.1 Atmospheric Components 2-5
2.3.2 Tributaries 2-7
2.3.3 Open Lake 2-9
2.4 Sampling Methods 2-9
2.4.1 Atmospheric Components 2-9
2.4.2 Tributaries 2-10
2.4.3 Open Lake 2-11
2.5 Analytical Methods 2-11
2.5.1 Atmospheric Components 2-11
2.5.2 Tributaries and Open Lake 2-11
-------
Results of the LMMB Study: Atrazine Data Report
2.6 Quality Implementation and Assessment 2-12
Chapter 3 Atrazine in Atmospheric Components 3-1
3.1 Results 3-1
3.1.1 Vapor Fraction 3-2
3.1.1.1 Seasonal Variation 3-4
3.1.1.2 Geographical Variation 3-4
3.1.1.3 Analysis of Breakdown Products 3-4
3.1.2 Particulate Fraction 3-4
3.1.2.1 Seasonal Variation 3-5
3.1.2.2 Geographical Variation 3-6
3.1.2.3 Analysis of Breakdown Products 3-9
3.1.3 Precipitation Fraction 3-9
3.1.3.1 Seasonal Variation 3-10
3.1.3.2 Geographical Variation 3-14
3.1.3.3 Analysis of Breakdown Products 3-16
3.2 Data Interpretation 3-17
3.2.1 Atmospheric Sources and Concentrations 3-17
3.2.2 Seasonality 3-18
3.2.3 Regional Considerations 3-18
3.2.4 Atrazine Breakdown Products 3-19
3.3 Quality Implementation and Assessment 3-20
Chapter 4 Atrazine in Tributaries 4-1
4.1 Results 4-1
4.1.1 Seasonal Variation 4-1
4.1.2 Geographical Variation 4-5
4.1.3 Analysis of Breakdown Products 4-8
4.2 Data Interpretation 4-10
4.2.1 Atrazine Levels in Lake Michigan Tributaries 4-10
4.2.2 Comparison to Regulatory Limits 4-11
4.2.3 Seasonality 4-11
4.2.4 Regional Considerations 4-11
4.2.5 Atrazine Breakdown Products 4-13
4.3 Quality Implementation and Assessment 4-13
Chapter 5 Atrazine in the Open-Lake Water Column 5-1
5.1 Results 5-1
5.1.1 Geographical Variation 5-1
5.1.2 Seasonal Variation 5-2
5.1.3 Vertical Variation 5-5
5.1.4 Analysis of Breakdown Products 5-5
5.2 Data Interpretation 5-6
5.2.1 Atrazine Levels in Lake Michigan 5-6
5.2.2 Comparison to Regulatory Limits 5-7
5.2.3 Lateral Variation 5-7
5.2.4 Temporal Trends 5-7
5.2.5 Vertical Trends 5-8
5.3 Quality Implementation and Assessment 5-8
-------
Table of Contents
References R-l
List of Figures
Figure 1-1. Lake Michigan Mass Balance Study Sampling Locations 1-8
Figure 1-2. Flow of Information in the Lake Michigan Mass Balance Study 1-11
Figure 2-1. Chemical Structure of Atrazine 2-1
Figure 2-2. Estimated Annual Usage Rates of Atrazine in the United States for 1991-1995 (from U.S.
Geological Survey, Pesticide National Synthesis Project, 1998) 2-1
Figure 2-3. Chemical Structure for Atrazine Breakdown Products, Deethyl-atrazine (DBA) and
Deisopropyl-atrazine (DIA) 2-3
Figure 2-4. Sources and Stores of Atrazine Measured in the LMMB Study 2-4
Figure 2-5. Principal Investigators Responsible for Sampling and Analysis of Atrazine in Atmospheric,
Tributary, and Open-lake Water Column Components 2-5
Figure 2-6. Atmospheric Sampling Stations 2-6
Figure 2-7. Tributary Sampling Stations 2-7
Figure 2-8. Open-lake Water Column Sampling Stations 2-9
Figure 3-1. Detection Frequency of Atrazine in Vapor Phase Samples 3-3
Figure 3-2. Detection Frequency of Atrazine in Particulate Phase Samples 3-5
Figure 3-3. Frequency Distribution of Atrazine Detection in Particulate Phase Atmospheric Samples
3-5
Figure 3-4. Seasonal Trend of Atrazine Concentrations in the Particulate Phase 3-7
Figure 3-5. Mean Spring/Summer Atrazine Concentrations Measured in the Particulate Phase 3-9
Figure 3-6. Detection Frequency of Atrazine in Precipitation Samples 3-10
Figure 3-7. Frequency Distribution of Atrazine Detection in Precipitation Samples 3-10
Figure 3-8. Seasonal Trend of Atrazine Concentrations in Precipitation 3-11
Figure 3-9. Mean Spring/Summer Atrazine Concentrations Measured in Precipitation 3-15
Figure 3-10. Detection Frequency of DBA in Precipitation Samples 3-16
Figure 3-11. Detection Frequency of DIA in Precipitation Samples 3-16
Figure 3-12. Correlation of DBA and DIA Concentrations with Measured Atrazine Concentrations in
Precipitation Samples 3-17
Figure 4-1. Seasonal Trend of Atrazine Concentrations in Lake Michigan Tributaries 4-2
Figure 4-2. Seasonal Flow Patterns and Atrazine Concentrations in Selected Lake Michigan Tributaries
4-4
Figure 4-3. Atrazine Concentrations Measured in Lake Michigan Tributaries 4-5
Figure 4-4. Mean Atrazine Concentrations Measured in Lake Michigan Tributaries 4-7
Figure 4-5. DBA and DIA Concentrations Measured in Lake Michigan Tributaries 4-9
Figure 4-6. Correlation of DBA and DIA Concentrations with Measured Atrazine Concentrations in
Tributary Samples 4-10
Figure 4-7. Mean Tributary Atrazine Concentrations (in parenthesis) and Land Use Patterns in the Lake
Michigan Watershed 4-12
Figure 4-8. Mean Tributary Atrazine Concentrations (in parenthesis) and Atrazine Use in the Lake
Michigan Watershed Estimated for 1987 to 1989 (graphic modified from Battaglin, U.S.
Geological Survey, 1994) 4-13
Figure 5-1. Atrazine Concentrations Measured in Lake Michigan During Four Sampling Cruises (Cruise
1= April/May 1994, Cruise 2= June 1994, Cruise 3 = August 1994, Cruise 6 = March/April
1995) 5-3
Figure 5-2. DBA and DIA Concentrations Measured in Lake Michigan During Four Sampling Cruises
(Cruise 1 = April/May 1994, Cruise 2 = June 1994, Cruise 3 = August 1994, Cruise 6 =
vii
-------
Results of the LMMB Study: Atrazine Data Report
March/April 1995) 5-4
Figure 5-3. Depth Profile of Atrazine, DBA, and DIA Concentrations at Station 41 in Lake Michigan on
August 12, 1994 5-5
Figure 5-4. Correlation of DBA and DIA Concentrations with Measured Atrazine Concentrations in
Open-lake Samples 5-6
List of Tables
Table 1-1. Characteristics of Lake Michigan Mass Balance Modeled Pollutants 1-5
Table 1-2. Lake Michigan Mass Balance Study Parameters 1-6
Table 2-1. Watershed Characteristics for Tributaries Monitored in the LMMB Study 2-8
Table 3-1. Number of Atmospheric Samples Analyzed for Atrazine, DBA, and DIA 3-1
Table 3-2. Atrazine Concentrations in the Vapor Fraction Measured Above Sample-specific Detection
Limits 3-3
Table 3-3. Mean Spring/Summer Atrazine Concentrations Measured in the Particulate Phase 3-8
Table 3-4. Volume-weighted Mean Spring/Summer Atrazine Concentrations Measured in
Precipitation 3-13
Table 3-5. Volume-weighted Mean Spring/Summer DBA and DIA Concentrations Measured in
Precipitation 3-14
Table 3-6. Summary of Routine Field Sample Flags for the Analysis of Atrazine in Atmospheric
Samples 3-22
Table 3-7. Data Quality Assessment for the Analysis of Atrazine in Atmospheric Samples 3-24
Table 4-1. Tributary Samples Collected and Analyzed for Atrazine, DBA, and DIA 4-1
Table 4-2. Correlation of Tributary Atrazine Levels with Tributary Flow 4-3
Table 4-3. Mean Atrazine Concentrations Measured in Lake Michigan Tributaries 4-6
Table 4-4. Mean DBA and DIA Concentrations Measured in Lake Michigan Tributaries 4-8
Table 4-5. Summary of Routine Field Sample Flags for the Analysis of Atrazine in Tributary Samples
4-14
Table 4-6. Data Quality Assessment for the Analysis of Atrazine in Tributary Samples 4-15
Table 5-1. Summary of Open-lake Samples Collected 5-1
Table 5-2. Summary of Routine Field Sample Flags for the Analysis of Atrazine in Open-lake Samples
5-9
Table 5-3. Data Quality Assessment for the Analysis of Atrazine in Open-lake Samples 5-10
VIII
-------
Executive Summary
The U.S. Environmental Protection Agency's Great Lakes National Program Office (GLNPO) and
partners instituted the Lake Michigan Mass Balance (LMMB) Study to measure and model the
concentrations of representative pollutants within important compartments of the Lake Michigan
ecosystem. The goal of the LMMB Study was to develop a sound, scientific base of information to guide
future toxic load reduction efforts at the Federal, State, Tribal, and local levels. Objectives of the study
were to: 1) estimate pollutant loading rates, 2) establish a baseline to gauge future progress, 3) predict the
benefits associated with load reductions, and 4) further understand ecosystem dynamics. The LMMB
Study measured the concentrations of poly chlorinated biphenyls (PCBs), fra«s-nonachlor, atrazine, and
mercury in the atmosphere, tributaries, lake water, sediments, and food webs of Lake Michigan. This
document summarizes the atrazine data collected as part of the LMMB Study, and is one in a series of
data reports that documents the project. Future documents will present the results of mass balance
modeling.
Atrazine is a triazine herbicide that is widely used in the U.S. to control broadleaf weeds in the production
of corn and sorghum. Approximately 64 to 75 million pounds of atrazine are applied per year in the U.S.,
much of which is used in the "Corn Belt" region that includes the upper midwest surrounding Lake
Michigan. Atrazine is generally applied to soil pre-planting or pre-emergence, but is sometimes also
applied to the foliage post-emergence. Atrazine can enter surface waters, including Lake Michigan,
through runoff, spray drift, discharge of contaminated groundwater to surface water, wet deposition
(dissolution of atrazine vapor in rainfall and washout of particulate bound atrazine), dry deposition (dry
settling of particulate bound atrazine), and sorption from the vapor phase. For human health protection,
EPA has set a maximum contaminant level of 3 |ig/L in drinking water. EPA also has set draft ambient
aquatic life criteria at 350 i-ig/L for protection from acute effects and 12 i-ig/L for protection from chronic
effects.
In the LMMB Study, atrazine and atrazine metabolites (deethyl-atrazine [DEA] and deisopropyl-atrazine
[DIA]) were measured in atmospheric, tributary water column, and open-lake water column samples.
From March 1994 through October 1995, over 1000 samples were collected and analyzed by gas
chromatography/mass spectrometry. Atmospheric vapor, particulate, and precipitation samples were
collected from eight stations surrounding Lake Michigan and three background stations outside the Lake
Michigan basin. Tributary water column samples were collected from 11 tributary rivers that flow into
Lake Michigan. Open-lake water column samples were collected from 35 sampling stations in Lake
Michigan, 2 stations in Green Bay, and 1 station in Lake Huron. While sediment and biological tissue
were sampled for mercury, PCBs, and trans-nonachlor, these compartments were not sampled for
atrazine, because atrazine is relatively water soluble, degradable, and does not accumulate in these
compartments.
Atrazine in Atmospheric Components
The predominant atmospheric source of atrazine, DEA, and DIA measured in this study was precipitation.
In atmospheric samples, atrazine was seldom detected in the vapor phase. Only 3.7% of vapor phase
samples were above sample-specific detection limits that averaged 32 pg/m3 and 20.7 pg/m3 for samples
analyzed at the Illinois Water Survey and Indiana University, respectively. Atrazine was more frequently
detected in the particulate phase and in precipitation, with 23% and 50% of sample concentrations
reported above sample-specific detection limits, respectively. The presence and concentration of atrazine
in both the particulate phase and in precipitation was highly seasonal. Atrazine was generally not
detectable in atmospheric samples during the fall and winter, but atmospheric concentrations peaked
during the spring in connection with the agricultural application of the herbicide. Maximum monthly
ix
-------
Results of the LMMB Study: Atrazine Data Report
atrazine concentrations in the participate phase ranged from 160 pg/m3 to 1400 pg/m3 among atmospheric
sampling stations, and mean atrazine concentrations during the spring and summer (March 20 through
September 23) ranged from 25 pg/m3 to 370 pg/m3 among atmospheric sampling stations. In
precipitation, maximum monthly atrazine concentrations ranged from 100 ng/L to 2800 ng/L among
atmospheric sampling stations, and monthly volume-weighted mean concentrations during the spring and
summer ranged from 19 ng/L to 120 ng/L among atmospheric sampling stations. Concentrations of
atrazine metabolites (DBA and DIA) were well correlated with atrazine concentrations and generally
followed the same patterns.
In general, atrazine in the particulate phase was higher at atmospheric sampling stations surrounding the
southern Lake Michigan basin than at those stations surrounding the northern basin. This is consistent
with agricultural land use that decreases in intensity from south to north in the Lake Michigan region.
Atrazine concentrations in precipitation were less reflective of local land use conditions and suggest long-
range transport of the herbicide in addition to local inputs. Atrazine concentrations in precipitation were
not consistently higher surrounding the southern Lake Michigan basin, and in fact, atrazine
concentrations in precipitation were often higher at remote sampling stations in the far north than at
stations surrounding the southern basin.
Atrazine in Tributaries
Atrazine was detected above the method detection limit of 1.25 ng/L in 99% of tributary samples.
Maximum atrazine concentrations in Lake Michigan tributaries ranged from 6.4 ng/L in the Manistique
River to 2700 ng/L in the St. Joseph River, and mean atrazine concentrations ranged from 3.7 ng/L to 350
ng/L in these same two rivers, respectively. Concentrations of atrazine in tributaries were strongly
influenced by geographical location and corn crop acerage. Mean atrazine concentrations in the
Manistique, Pere Marquette, and Menominee Rivers were statistically lower than in the remaining eight
measured tributaries except for the Muskegon River. The watersheds of these three tributaries are more
forested and contain fewer agricultural influences than the other monitored tributaries. Atrazine
concentrations were highest in the St. Joseph, Grand, and Kalamazoo Rivers, where agricultural
influences were much stronger and atrazine use rates were 52 to over 160 lbs/mi2. For these three
tributaries with the highest atrazine levels, distinct peaks in atrazine were observed in mid to late May,
corresponding with the agricultural application of the herbicide. Distinct seasonal patterns of atrazine
concentrations were not observed for the other tributaries. Concentrations of atrazine metabolites (DBA
and DIA) in tributaries were well correlated with atrazine concentrations and generally followed the same
patterns.
Atrazine in the Open-lake Water Column
Within Lake Michigan, atrazine concentrations in open-lake water column samples were relatively
consistent. Average atrazine concentrations at open-lake sampling stations within Lake Michigan ranged
from 33.0 to 48.0 ng/L. Atrazine concentrations within Lake Michigan were statistically greater than
those measured at the reference station on Lake Huron and were statistically lower than concentrations
measured at one Green Bay sampling station. Because the open lake was well-mixed with respect to
atrazine, lake-wide averages could be calculated. Over the course of the study, lake-wide average
atrazine concentrations increased from 37.0 ng/L in April/May 1994 to 39.7 ng/L in March/April 1995.
During this same time period, DBA concentrations increased by 14.9% and DIA concentrations increased
by 54.0%. While atrazine concentrations increased slightly during the study, open-lake average atrazine
levels remained more than 50 times below the maximum contaminant level for drinking water and more
than 300 times less than the proposed ambient water quality criterion for protection of aquatic life from
chronic effects.
-------
Chapter 1
Project Overview
The U.S. Environmental Protection Agency's Great Lakes National Program Office (GLNPO) and
partners instituted the Lake Michigan Mass Balance (LMMB) Study to measure and model the
concentrations of representative pollutants within important compartments of the Lake Michigan
ecosystem. The LMMB Study measured the concentrations of poly chlorinated biphenyls (PCBs), trans-
nonachlor, atrazine, and mercury in the atmosphere, tributaries, lake water, sediments, and food webs of
Lake Michigan. This document summarizes the atrazine data collected as part of the LMMB Study, and
is one in a series of data reports that documents the project. Future documents will present the results of
mass balance modeling.
1.1 Background
The Great Lakes, which contain 20% of the world's freshwater, are a globally important natural resource
that are currently threatened by multiple stressors. While significant progress has been made to improve
the quality of the lakes, pollutant loads from point, non-point, atmospheric, and legacy sources continue
to impair ecosystem functions and limit the attainability of designated uses of these resources. Fish
consumption advisories and beach closings continue to be issued, emphasizing the human health concerns
from lake contamination. Physical and biological stressors such as invasion of non-native species and
habitat loss also continue to threaten the biological diversity and integrity of the Great Lakes.
The United States and Canada have recognized the significance and importance of the Great Lakes as a
natural resource and have taken steps to restore and protect the lakes. In 1978, both countries signed the
Great Lakes Water Quality Agreement (GLWQA). This agreement calls for the restoration and
maintenance of the chemical, physical, and biological integrity of the Great Lakes by developing plans to
monitor and limit pollutant flows into the lakes.
The GLWQA, as well as Section 118(c) of the Clean Water Act, required the development of Lake-wide
Management Plans (LaMPs) for each Great Lake. The purpose of these LaMPs is to document an
approach to reducing inputs of critical pollutants to the Great Lakes and restoring and maintaining Great
Lakes integrity. To assist in developing these LaMPs and to monitor progress in pollutant reduction,
Federal, State, Tribal, and local entities have instituted Enhanced Monitoring Plans. Monitoring is
essential to the development of baseline conditions for the Great Lakes and provides a sound scientific
base of information to guide future toxic load reduction efforts.
The LMMB Study is a part of the Enhanced Monitoring Plan for Lake Michigan. The LMMB Study was
a coordinated effort among Federal, State, and academic scientists to monitor tributary and atmospheric
pollutant loads, develop source inventories of toxic substances, and evaluate the fates and effects of these
pollutants in Lake Michigan. A mass balance modeling approach provides the predictive ability to
determine the environmental benefits of specific load reduction scenarios for toxic substances and the
time required to realize those benefits. This predictive ability will allow Federal, State, Tribal, and local
agencies to make more informed load reduction decisions.
1.2 Description
The LMMB Study used a mass balance approach to evaluate the sources, transport, and fate of
contaminants in the Lake Michigan ecosystem. A mass balance approach is based on the law of
conservation of mass, which states that the amount of a pollutant accumulating in a system is equal to the
amount entering the system, less the amount of that pollutant leaving or chemically changed in the
system. If the system is defined as the Lake Michigan/Green Bay water column, then pollutants may
1-1
-------
Results of the LMMB Study: Atrazine Data Report
enter the system via tributaries, direct runoff, the atmosphere (wet deposition, dry deposition, and
sorption from the vapor phase), the sediment, and the Straits of Mackinac. Pollutants may leave the
system through volatilization to the atmosphere, loss to the sediment, or discharge through the Straits of
Mackinac and the Chicago water diversion. The law of conservation of mass also can be applied to other
systems such as biota, sediment, or air.
System
Moss = Mass
tranrformed
MaSSstored
Massstored + Massout
The LMMB Study measured contaminant What is Mass Balance?
concentrations in various inputs and ecosystem
compartments over spatial and temporal scales.
Mathematical models that track the transport
and fate of contaminants within Lake Michigan
are being developed and calibrated using these
field data. The LMMB Study is the first lake-
wide application of a mass balance
determination for toxics in the Great Lakes and
will serve as the basis of future mass
budget/mass balance efforts.
1.3 Scope
1.3.1 Modeled Pollutants
When EPA published the Water Quality
Guidance for the Great Lakes System (58 FR
20802), the Agency established water quality
criteria for 29 pollutants. Those criteria are
designed to protect aquatic life, terrestrial
wildlife, and human health. PCBs, trans-
nonachlor, and mercury are included in the list of 29 pollutants. The water quality criteria and values
proposed in the guidance apply to all of the ambient waters of the Great Lakes system, regardless of the
sources of pollutants in those waters. The proposed criteria provide a uniform basis for integrating
Federal, State, and Tribal efforts to protect and restore the Great Lakes ecosystem.
The number of pollutants that can be intensively monitored and modeled in the Great Lakes system is
limited by the resources available to collect and analyze thousands of samples, assure the quality of the
results, manage the data, and develop and calibrate the necessary models. Therefore, the LMMB Study
focused on constructing mass balance models for a limited group of pollutants. PCBs, fra«s-nonachlor,
atrazine, and mercury were selected for inclusion in the LMMB Study because these pollutants currently
or potentially pose a risk to aquatic and terrestrial organisms (including humans) in the Lake Michigan
ecosystem. These pollutants also were selected to cover a wide range of chemical and physical properties
and represent other classes of compounds which pose current or potential problems. Once a mass budget
for selected pollutants is established and a mass balance model is calibrated, additional contaminants can
be modeled with limited data and future resources can be devoted to activities such as emission
inventories and dispersion modeling.
1.3.1.1 Polychlorinated Biphenyls
PCBs are a class of man-made, chlorinated, organic chemicals that include 209 congeners, or specific
PCB compounds. The highly stable, nonflammable, non-conductive properties of these compounds have
made them useful in a variety of products including electrical transformers and capacitors, plastics,
rubber, paints, adhesives, and sealants. PCBs were produced for such industrial uses in the form of
1-2
-------
Project Overview
complex mixtures under the trade name "Aroclor" and were commercially available from 1930 through
1977, when EPA banned their production due to environmental and public health concerns. PCBs also
may be produced by combustion processes, including incineration, and can be found in stack emissions
and ash from incinerators.
Seven Aroclor formulations were included in the Priority Pollutant List developed by the EPA Office of
Water under the auspices of the Clean Water Act because they were found by EPA in the effluents from
one or more wastewater treatment facilities. Aroclors may have entered the Great Lakes through other
means, including spills or improper disposal of transformer fluids, contaminated soils washing into the
watershed, or discharges from ships. The PCBs produced by combustion processes may be released to
the atmosphere, where they are transported in both vapor and particulate phases and enter the lakes
through either dry deposition or precipitation events (e.g., rain).
The stability and persistence of PCBs, which made them useful in industrial applications, have also made
these compounds ubiquitous in the environment. PCBs do not readily degrade and thus accumulate in
water bodies and aquatic sediments. PCBs also bioaccumulate, or buildup, in living tissues. Levels of
PCBs in some fish from Lake Michigan exceed U.S. Food and Drug Administration tolerances,
prompting closure of some commercial fisheries and issuance offish consumption advisories. PCBs are a
probable human carcinogen, and human health effects of PCB exposure include stomach, kidney, and
liver damage, liver and biliary tract cancer, and reproductive effects, including effects on the fetus after
exposure of the mother.
PCB congeners exhibit a wide range of physical and chemical properties (e.g., vapor pressures,
solubilities, boiling points), are relatively resistant to degradation, and are ubiquitous. These properties
make them ideal surrogates for a wide range of organic compounds from anthropogenic sources.
In the LMMB Study, PCBs were selected as a model for conservative organic compounds (USEPA,
1997c).
1.3.1.2 Trans-Nonachlor
7>a«s-nonachlor is a component of the pesticide chlordane. Chlordane is a mixture of chlorinated
hydrocarbons that was manufactured and used as a pesticide from 1948 to 1988. Prior to 1983,
approximately 3.6 million pounds of chlordane were used annually in the U.S. In 1988, EPA banned all
production and use of chlordane in the U.S.
Like PCBs, chlordane is relatively persistent and bioaccumulative. Trans-nonachlor is the most
bioaccumulative of the chlordanes. Trans-nonachlor is a probable human carcinogen. Other human
health effects include neurological effects, blood dyscrasia, hepatoxicity, immunotoxicity, and endocrine
system disruption.
Historically, fra«s-nonachlor may have entered the Great Lakes through a variety of means related to the
application of chlordane, including improper or indiscriminate application, improper cleaning and
disposal of pesticide application equipment, or contaminated soils washing into the watershed.
In the LMMB Study, fra«s-nonachlor was selected as a model for the cyclodiene pesticides (USEPA,
1997c).
1-3
-------
Results of the LMMB Study: Atrazine Data Report
1.3.1.3 Atrazine
Atrazine is a herbicide based on a triazine ring structure with three carbon atoms alternating with three
nitrogen atoms. Atrazine is the most widely used herbicide in the U.S. for corn and sorghum production.
Atrazine has been used as an agricultural herbicide since 1959, and 64 to 75 million pounds of atrazine
are used annually in the U.S. Atrazine is extensively used in the upper Midwest, including the Lake
Michigan watershed, where it is primarily associated with corn crops.
Unlike PCBs and trans-nonachlor, atrazine is not extremely persistent or bioaccumulative. Atrazine is
moderately susceptible to biodegradation, with a half-life in soils of about 60-150 days. Atrazine may
persist considerably longer in water and is relatively non-reactive in the atmosphere. Atrazine rarely
exceeds the maximum contaminant level (MCL) set by USEPA as a drinking water standard, but
localized peak values can exceed the MCL following rainfall events after atrazine application. Atrazine
can cause human health effects such as weight loss, cardiovascular damage, muscle and adrenal
degeneration, and congestion of heart, lungs, and kidneys. Atrazine is also toxic to aquatic plants.
In the LMMB Study, atrazine was selected as a model for more reactive, biodegradable compounds in
current use (USEPA, 1997c).
1.3.1.4 Mercury
Mercury is a naturally-occurring toxic metal. Mercury is used in battery cells, barometers, thermometers,
switches, fluorescent lamps, and as a catalyst in the oxidation of organic compounds. Global releases of
mercury in the environment are both natural and anthropogenic (caused by human activity). It is
estimated that about 5,500 metric tons of mercury are released annually to the air, soil, and water from
anthropogenic and natural sources (USEPA 1997e). These sources include combustion of various fuels
such as coal; mining, smelting and manufacturing activities; wastewater; agricultural, animal and food
wastes.
As an elemental metal, mercury is extremely persistent in all media. Mercury also bioaccumulates with
reported bioconcentration factors in fish tissues in the range of 63,000 to 100,000. Mercury is a possible
human carcinogen and causes the following human health effects: stomach, large intestine, brain, lung,
and kidney damage; blood pressure and heart rate increase, and fetus damage.
In the LMMB Study, mercury was selected as a model for bioaccumulative metals (USEPA, 1997c).
1-4
-------
Project Overview
Table 1-1. Characteristics of Lake Michigan Mass Balance Modeled Pollutants
Pollutant
PCBs
trans-
Non-
achlor0
Atrazine
Mercury
Sources
Waste incinerators
(unintentional
byproducts of
combustion)
Industrial
dischargers
Electrical power
Application to crops
and gardens
Application to crops
Waste disposal
Manufacturing
processes
Energy production
Ore processing
Municipal and
medical waste
incinerators
Chloralkali factories
Fuel combustion
Uses
Electrical
transformers and
capacitors
Carbonless copy
paper
Plasticizers
Hydraulic fluids
Pesticide on corn
and citrus crops
Pesticide on lawns
and gardens
Herbicide for corn
and sorghum
production
Battery cells
Barometers
Dental fillings
Thermometers
Switches
Fluorescent lamps
Toxic Effects
Probable human carcinogen
Hearing and vision
impairment
Liver function alterations
Reproductive impairment and
deformities in fish and wildlife
Probable human carcinogen
Nervous system effects
Blood system effects
Liver, kidney, heart, lung,
spleen, and adrenal gland
damage
Weightless
Cardiovascular damage
Muscle and adrenal
degeneration
Congestion of heart, lungs,
and kidneys
Toxic to aquatic plants
Possible human carcinogen
Damage to brain and kidneys
Adverse affects on the
developing fetus, sperm, and
male reproductive organs
Biocon-
cent ration
Factor3
1,800 to
180,000
4,000 to
40,000
2 to 100
63,000 to
100,000
EPA
Regulatory
Standards'1
MCL = 0.5 |jg/L
CCC = 14ng/L
HH = 0.17ng/L
MCL = 2 |jg/L
CMC = 2.4 |jg/L
CCC = 4.3 ng/L
HH = 2.1 ng/L
MCL = 3 |jg/L
CMCd = 350 |jg/L
CCCd = 12|jg/L
MCL = 2 |jg/L
CMC = 1.4|jg/L
CCC = 0.77 |jg/L
HH = 50 ng/L
FWAe = 2.4 |jg/L
FWCe = 12ng/L
Wildlife' =1.3 ng/L
3 From: USEPA. 1995a. National Primary Drinking Water Regulations, Contaminant Specific Fact Sheets, Inorganic Chemicals, Technical
Version. EPA811/F-95/002-T. U.S. Environmental Protection Agency, Office of Water, Washington, D.C.; and USEPA. 1995b. National
Primary Drinking Water Regulations, Contaminant Specific Fact Sheets, Synthetic Organic Chemicals, Technical Version. EPA 811/F-
95/003-T. U.S. Environmental Protection Agency, Office of Water, Washington, DC.
b MCL = Maximum Contaminant Level for drinking water. CMC = Criterion Maximum Concentration for protection of aquatic life from acute
toxicity. CCC = Criterion Continuous Concentration for protection of aquatic life from chronic toxicity. HH = water quality criteria for
protection of human health from water and fish consumption. Data from: USEPA. 1999. National Recommended Water Quality Criteria-
Correction. EPA 822/Z-99/001. U.S. Environmental Protection Agency, Office of Water, Washington, DC.
c Characteristics presented are for chlordane. frans-nonachlor is a principle component of the pesticide chlordane.
d Draft water quality criteria for protection of aquatic life. From: USEPA. 2001a. Ambient Aquatic Life Water Quality Criteria for Atrazine.
U.S. Environmental Protection Agency, Office of Water, Washington, DC.
e FWA = Freshwater acute water quality criterion. FWC = Freshwater chronic water quality criterion. From National Toxics Rule (58 FR
60848).
f Wildlife criterion. From the Stay of Federal Water Quality Criteria for Metals (60 FR 22208), 40 CFR131.36 and the Water Quality Guidance
for the Great Lakes System (40 CFR 132).
1-5
-------
Results of the LMMB Study: Atrazine Data Report
1.3.2 Other Measured Parameters
In addition to the four chemicals modeled in the LMMB Study, many other chemicals and parameters
were measured in the LMMB Study as part of the Enhanced Monitoring Program. A survey of these
chemicals and parameters will aid in understanding the overall ecological integrity of Lake Michigan.
These additional parameters include various biological indicators, meteorological parameters, and
organic, metal, and conventional chemicals in Lake Michigan. A complete listing of all parameters
included in this study is provided in Table 1-2.
Table 1-2. Lake Michigan Mass Balance Study Parameters
Organics
acenaphthene
acenaphthylene
aldrin
anthracene
atrazine
a-BHC
(3-BHC
5-BHC
Y-BHC (Lindane)
benzo[a]anthracene
benzo[g,/y]perylene
benzo[6]fluoranthene
benzo[/(]fluoranthene
benzo[e]pyrene
benzo[a]pyrene
a-chlordane
y-chlordane
chrysene
coronene
p,p'-DDE
p,p'-DDD
p,p'-DDT
endosulfan sulfate
endosulfan I
endosulfan II
endrin
endrin aldehyde
endrin ketone
fluoranthene
fluorene
heptachlor
heptachlor epoxide
hexachlorobenzene (HCB)
indeno[1 ,2,3-ccf]pyrene
mirex
frans-nonachlor
oxychlordane
PCB congeners
phenanthrene
pyrene
retene
toxaphene
Metals
aluminum
arsenic
calcium
cadmium
chromium
cesium
copper
iron
mercury
potassium
magnesium
manganese
sodium
nickel
lead
selenium
thorium
titanium
vanadium
zinc
1-6
-------
Project Overview
Table 1-2. Lake Michigan Mass Balance Study Parameters
Conventionals
alkalinity
ammonia
bromine
chloride
chlorine sulfate
conductivity
dissolved organic carbon
dissolved oxygen
dissolved phosphorous
dissolved reactive silica
dry weight fraction
elemental carbon
nitrate
ort/io-phosphorous
particulate organic carbon
percent moisture
PH
phosphorous
silica
silicon
temperature
total Kjeldahl nitrogen
total organic carbon
total phosphorous
total suspended particulates
total hardness
turbidity
Biologicals
fish species
fish age
fish maturity
chlorophyll a
fish lipid amount
fish weight
fish length
fish taxonomy
fish diet analysis
primary productivity
Meteorological
air temperature
relative humidity
barometric pressure
weather conditions
wind direction
wind speed
visibility
wave height and direction
1.3.3 Measured Compartments
In the LMMB Study, contaminants were measured in the following compartments:
» Open-Lake Water Column The water column in the open lake was sampled and analyzed for the
modeled pollutants
* Tributaries Tributary water columns were sampled and analyzed for the modeled pollutants
* Fish Top predators and forage base species were sampled and analyzed for diet analysis and
contaminant burden. Fish were not analyzed for atrazine, because atrazine is not bioaccumulative.
» Lower Pelagic Food Web Phytoplankton and zooplankton were sampled and analyzed for
species diversity, taxonomy, and contaminant burden. The lower pelagic food web was not analyzed
for atrazine, because atrazine is not bioaccumulative.
» Sediments Cores were collected and trap devices were used to collect Lake Michigan sediment
for determination of contaminants and sedimentation rates. Sediments were not analyzed for atrazine,
because atrazine is relatively water soluble, degradable, and does not generally accumulate in
sediments.
* Atmosphere Vapor, particulate, and precipitation phase samples were collected and analyzed for
the modeled pollutants
For the modeled pollutants, more than 20,000 samples were collected and analyzed, including more than
9000 quality control (QC) samples, at more than 300 sampling locations (Figure 1-1). Field data
collection activities were initially envisioned as a one-year effort. However, it became evident early into
1-7
-------
Results of the LMMB Study: Atrazine Data Report
the project that a longer collection period would be necessary to provide a full year of concurrent
information on contaminant loads and ambient concentrations for modeling purposes. Therefore, field
sampling occurred from April, 1994 to October, 1995.
Figure 1-1. Lake Michigan Mass Balance Study Sampling Locations
Beaver Island
Manitowoc
Sheboygan River
Sleeping Bear
Dunes
Pere Marquette
N River
Muskegon River
vm'LMuskegon
Milwaukee^*.' 'A.*^^
A+ "* ^^'VJ'Grand River
Milwaukee River \J( * ^
*A *4r " IT) Kalamazoo F
Chiwaukee Prairie A ?^CT , ,,
^« * * A South Haven
/^_ * - * ^T*
} St. Joseph River
Benton Harbor
NT Chicago
Chicago SWFP Intake
Grand Calumet Harbor
Indiana Dunes
Atmospheric Station
Tributary Station
Biota Station
Sediment Station
Water Column Station
1.4 Objectives
The goal of the LMMB Study was to develop a sound, scientific base of information to guide future toxic
load reduction efforts at the Federal, State, Tribal, and local levels. To meet this goal, the four following
LMMB Study objectives were developed:
1-8
-------
Project Overview
* Estimate pollutant loading rates Environmental sampling of major media will allow estimation
of relative loading rates of critical pollutants to the Lake Michigan Basin.
* Establish baseline Environmental sampling and estimated loading rates will establish a baseline
against which future progress and contaminant reductions can be gauged.
» Predict benefits associated with load reductions The completed mass balance model will
provide a predictive tool that environmental decision-makers and managers may use to evaluate the
benefits of specific load reduction scenarios.
* Understand ecosystem dynamics Information from the extensive LMMB monitoring and
modeling efforts will improve our scientific understanding of the environmental processes governing
contaminant cycling and availability within relatively closed ecosystems.
1.5 Design
1.5.1 Organization
The Great Lakes National Program Office proposed a mass balance approach to provide coherent,
ecosystem-based evaluation of toxics in Lake Michigan. GLNPO served as the program sponsor for the
LMMB Study. GLNPO formed two committees to coordinate study planning, the Program Steering
Committee and the Technical Coordinating Committee. These committees were comprised of scientists
from Federal, State, academic, and commercial institutions (see Section 1.5.2, Study Participants). The
committees administered a wide variety of tasks including: planning the project, locating the funding,
designing the sample collection, coordinating sample collection activities, locating qualified laboratories,
coordinating analytical activities, assembling the data, assuring the quality of the data, assembling skilled
modelers, developing the models, and communicating interim and final project results. The National
Exposure Research Laboratory (NERL) at Duluth, in cooperation with the National Oceanic and
Atmospheric Administration (NOAA) Great Lakes Environmental Research Laboratory and the
Atmospheric Sciences Modeling Division are supporting the modeling component of the mass balance
study by developing a suite of integrated mass balance models to simulate the transport, fate, and
bioaccumulation of the study target analytes.
1.5.2 Study Participants
The LMMB Study was a coordinated effort among Federal, State, academic, and commercial institutions.
The following agencies and organizations have all played roles in ensuring the success of the LMMB
Study. Except for the three organizations indicated with an asterisk (*), all of the participants were
members of the LMMB steering committee.
Federal and International
* USEPA Great Lakes National Program Office (Program Sponsor)
> USEPA Region 5 Water Division
* USEPA Office of Research and Development Large Lakes Research Station
* USEPA Atmospheric Research and Environmental Analysis Lab
* US Geological Survey
* US Fish and Wildlife Service
* US Department of Energy
» National Oceanic and Atmospheric Administration
> USEPA Office of Air and Radiation*
> USEPA Office of Water*
* Environment Canada*
1-9
-------
Results of the LMMB Study: Atrazine Data Report
State
* Illinois Department of Natural Resources
* Illinois Water Survey
* Indiana Department of Environmental Management
* Michigan Department of Environmental Quality
* Wisconsin Department of Natural Resources
* Wisconsin State Lab of Hygiene
Academic and Commercial
* Indiana University
* Rutgers University
* University of Maryland
* University of Michigan
* University of Minnesota
* University of Wisconsin
* Battelle Labs
* Grace Analytical
1.5.3 Workgroups
Eleven workgroups were formed to provide oversight and management of specific project elements. The
workgroups facilitated planning and implementation of the study in a coordinated and systematic fashion.
The workgroups communicated regularly through participation in monthly conference calls and annual
"all-hands" meetings. Workgroup chairs were selected and were responsible for managing tasks under
the purview of the workgroup and communicating the status of activities to other workgroups. The
workgroups and workgroup chairs are listed below.
* Program Steering Committee Paul Horvatin
* Technical Coordinating Committee Paul Horvatin
* Modeling Workgroup William Richardson
* Air Monitoring Workgroup Jackie Bode
* Biota Workgroup Paul Bertram and John Gannon
* Chemistry Workgroup David Anderson
* Data Management Workgroup Kenneth Klewin and Philip Strobel
* Lake Monitoring Workgroup Glenn Warren
* Tributary Monitoring Workgroup Gary Kohlhepp and Robert Day
* Quality Assurance Workgroup Louis Blume and Michael Papp
* Sediment Monitoring Workgroup Brian Eadie
1.5.4 Information Management
As program sponsor, GLNPO managed information collected during the LMMB Study. Principal
investigators (Pis) participating in the study reported field and analytical data to GLNPO. GLNPO
developed a data standard for reporting field and analytical data and a database for storing and retrieving
study data. GLNPO also was responsible for conducting data verification activities and releasing verified
data to the study modelers and the public. The flow of information is illustrated in Figure 1-2.
1-10
-------
Figure 1-2. Flow of Information in the Lake Michigan Mass Balance Study
Principal
Investigator (PI)
Collect and Analyze
Samples
Report Field and
Analytical Data
(According to LMMB Data
Standard)
Resolve Discrepancies
No
GLNPO Data
Management
Workgroup
Receive, Store, and
Transmit Data
-Yes-
Store, Transmit, and
Upload Data to
GLENDA
GLNPO QA
Workgroup
Conduct Data
Verification (Merge Field
and Analytical Data using
RDMQ)
Produce final verified
data file and provide to
PI for review and
approval
Produce Final Verified
Data File and Transmit
(in GLENDA-compatible
Format)
LMMB Study
Modelers
Input Data to Study
Models
External Parties
1-11
-------
Results of the LMMB Study: Atrazine Data Report
1.5.4.1 Data Reporting
More than twenty organizations produced LMMB data through the collection and analysis of more than
20,000 samples. In the interest of standardization, specific formats (i.e., file formats and codes to
represent certain data values) were established for reporting LMMB data. Each format specified the
"rules" by which data were submitted, and, in many cases, the allowable values by which they were to be
reported. The data reporting formats were designed to capture all pertinent sampling and analytical
information from the field crews and laboratory analysts. Data reporting formats and the resulting Great
Lakes Environmental Monitoring Database (GLENDA, see Section 1.5.4.2,) were designed to be
applicable to projects outside the LMMB as well. For the LMMB Study, special conditions were applied
for reporting analytical results. Because the data were being used for input to study models, principal
investigators were asked to report analytical results as measured even when measurements were below
estimated detection limits. The quality assurance program discussed in Section 1.5.5 included identifying
(i.e., flagging) all analytical results that were below estimated detection limits.
Principal investigators (including sampling crews and the analytical laboratories) supplied sample
collection and analysis data following the standardized reporting formats if possible. LMMB data were
then processed through an automated SAS-based data verification system, Research Data Management
and Quality Control System (RDMQ), for quality assurance/quality control checking. After verification
and validation by the PI, the datasets were output in a form specific for upload to GLENDA. Finally,
these datasets were uploaded to GLENDA.
1.5.4.2 Great Lakes Environmental Monitoring Database
Central to the data management effort is a computerized database system to house LMMB Study and
other project results. That system, the Great Lakes Environmental Monitoring Database (GLENDA), was
developed to provide data entry, storage, access and analysis capabilities to meet the needs of mass
balance modelers and other potential users of Great Lakes data.
Development of GLENDA began in 1993 with a logical model based on the modernized STORET
concept and requirements analysis. GLENDA was developed with the following guiding principles:
» True multi-media scope water, air, sediment, taxonomy, fish tissue, fish diet, and meteorology
data can all be housed in the database
* Data of documented quality data quality is documented by including results of quality control
parameters
» Extensive contextual indicators ensures data longevity by including enough information to allow
future or secondary users to make use of the data
* Flexible and expandable the database is able to accept data from any Great Lakes monitoring
project
» National compatibility GLENDA is compatible with STORET and allows ease of transfer
between these large databases
In an effort to reduce the data administration burden and ensure consistency of data in this database,
GLNPO developed several key tools. Features including standard data definitions, reference tables,
standard automated data entry applications, and analytical tools are (or will soon be) available.
15.4.3 Public Access to LMMB Data
All LMMB data that have been verified (through the QC process) and validated (accepted by the PI) are
available to the public. Currently, GLNPO requires that written requests be made to obtain LMMB data.
1-12
-------
Project Overview
The datasets are available in several formats including WK1, DBF, and SD2. More information about the
datasets is available on the LMMB web site at: http://www.epa.gov/glnpo/lmmb/datafaqs.html.
The primary reason for requiring an official request form for LMMB data is to keep track of requests.
This allows GLNPO to know how many requests have been made, who has requested data, and what use
they intend for the data. This information assists GLNPO in managing and providing public access to
Great Lakes data and conducting public outreach activities. As of November 2000, 38 requests for
LMMB data have been made: 8 from EPA, 5 from other federal agencies, 5 from state agencies, 5 from
universities, 10 from consultants, 3 from international agencies, and 2 from non-profit or other groups. In
the future, after all data are verified and validated, GLNPO intends to make condensed versions of the
datasets available on the LMMB web site for downloading. This will allow easy public access to LMMB
data.
Additional details of the information management for the LMMB study can be found in The Lake
Michigan Mass Balance Study Quality Assurance Report (USEPA, 200 li).
1.5.5 Quality Assurance Program
At the outset of the LMMB Study, managers recognized that the data gathered and the models developed
from the study would be used extensively by decision makers responsible for making environmental,
economic, and policy decisions. Environmental measurements are never true values and always contain
some level of uncertainty. Decision makers, therefore, must recognize and be sufficiently comfortable
with the uncertainty associated with data on which their decisions are based. In recognition of this
requirement, LMMB Study managers established a QA program goal of ensuring that data produced
under the LMMB Study would meet defined standards of quality with a specified level of confidence.
The QA program prescribed minimum standards to which all organizations collecting data were required
to adhere. Data quality was defined, controlled, and assessed through activities implemented within
various parameter groups (e.g., organic, inorganic, and biological parameters). QA activities included the
following:
* QA Program Prior to initiating data collection activities, plans were developed, discussed, and
refined to ensure that study objectives were adequately defined and to ensure that all QA activities
necessary to meet study objectives were considered and implemented.
* QA Workgroup EPA established a QA Workgroup whose primary function was to ensure that the
overall QA goals of the study were met.
* QA Project Plans (QAPPs) EPA worked with Pis to define program objectives, data quality
objectives (DQOs), and measurement quality objectives (MQOs) for use in preparing QAPPs.
Principal investigators submitted QAPPs to EPA for review and approval. EPA reviewed each QAPP
for required QA elements and soundness of planned QA activities.
* Training Before data collection activities, Pis conducted training sessions to ensure that
individuals were capable of properly performing data collection activities for the LMMB Study.
* Monthly Conference Calls and Annual Meetings EPA, Pis, and support contractors participated
in monthly conference calls and annual meetings to discuss project status and objectives, QA issues,
data reporting issues, and project schedules.
* Standardized Data Reporting Format Principal investigators were required to submit all data in
a standardized data reporting format that was designed to ensure consistency in reporting and
facilitate data verification, data validation, and database development.
* Intercomparison Studies EPA conducted studies to compare performance among different Pis
analyzing similar samples. The studies were used to evaluate the comparability and accuracy of
program data.
1-13
-------
Results of the LMMB Study: Atrazine Data Report
* Technical Systems Audits During the study, EPA formally audited each Pi's laboratory for
compliance with their QAPPs, the overall study objectives, and pre-determined standards of good
laboratory practice.
» Data Verification Pis and EPA evaluated project data against pre-determined MQOs and DQOs
to ensure that only data of acceptable quality would be included in the program database.
* Statistical Assessments EPA made statistical assessments of the LMMB study data to estimate
elements of precision, bias, and uncertainty.
» Data Validation EPA and modelers are evaluating the data against the model objectives.
Comparability of data among Pis participating in the LMMB Study was deemed to be important for
successful completion of the study. Therefore, Measurement Quality Objectives (MQOs) for several data
attributes were developed by the Pis and defined in the QAPPs. MQOs were designed to control various
phases of the measurement process and to ensure that the total measurement uncertainty was within the
ranges prescribed by the DQOs. MQOs were defined in terms of six attributes:
* Sensitivity/Detectability The determination of the low-range critical value that a method-specific
procedure can reliably discern for a given pollutant. Sensitivity measures included, among others,
method detection limits (MDLs) as defined at 40 CFR Part 136, system detection limits (SDLs), or
instrument detection limits (IDLs).
» Precision A measure of the degree to which data generated from replicate or repetitive
measurements differ from one another. Analysis of duplicate samples was used to assess precision.
* Bias The degree of agreement between a measured and actual value. Bias was expressed in terms
of the recovery of an appropriate standard reference material or spiked sample.
» Completeness The measure of the number of samples successfully analyzed and reported
compared to the number that were scheduled to be collected.
* Comparability The confidence with which one data set can be compared to other data sets.
* Representativeness The degree to which data accurately and precisely represent characteristics of
a population, parameter variations at a sampling point, a process condition, or an environmental
condition.
The Pi-defined MQOs also were used as the basis for the data verification process. GLNPO conducted
data verification through the LMMB QA Workgroup. The workgroup was chaired by GLNPO's Quality
Assurance Manager and consisted of quality control coordinators that were responsible for conducting
review of specific data sets. Data verification was performed by comparing all field and QC sample
results produced by each PI with their MQOs and with overall LMMB Study objectives. If a result failed
to meet predefined criteria, the QC Coordinator contacted the PI to discuss the result, verify that it was
correctly reported, and determine if corrective actions were feasible. If the result was correctly reported
and corrective actions were not feasible, the results were flagged to inform data users of the failure.
These flags were not intended to suggest that data were not useable; rather they were intended to caution
the user about an aspect of the data that did not meet the predefined criteria. Data that met all predefined
requirements were flagged to indicate that the results had been verified and were determined to meet
applicable MQOs. In this way, every data point was assigned one or more validity flags based on the
results of the QC checks. GLNPO also derived data quality assessments for each LMMB study dataset for
a subset of the attributes listed above, specifically sensitivity, precision, and bias. The LMMB study
modelers and the Large Lakes Research Station Database Manager also perform data quality assessments
prior to inputting data into study models. Such activities include verifying the readability of electronic
files, identifying missing data, checking units, and identifying outliers. A detailed description of the
quality assurance program is included in The Lake Michigan Mass Balance Study Quality Assurance
Report (USEPA, 200 li). A brief summary of quality implementation and assessment is provided in each
of the following chapters.
1-14
-------
Project Overview
1.6 Project Documents and Products
During project planning, LMMB participants developed study tools including work plans, a methods
compendium, quality assurance project plans, and data reporting standards. Through these tools, LMMB
participants documented many aspects of the study including information management and quality
assurance procedures. Many of these documents are available on GLNPO's website at
http: //www. epa.gov/glnpo/lmmb.
LMMB Work Plan
Designers of the LMMB Study have documented their approach in a report entitled Lake Michigan Mass
Budget/Mass Balance Work Plan (USEPA, 1997c). The work plan describes the essential elements of a
mass balance study and the approach used to measure and model these elements in the Lake Michigan
system. This document was developed based upon the efforts of many Federal and State scientists and
staff who participated in the initial planning workshop, as well as Pis.
Quality Assurance Program/Project Plans
The Lake Michigan Mass Balance Project Quality Assurance Plan for Mathematical Modeling, Version
3.0 (USEPA, 1998) documents the quality assurance process for the development and application of
LMMB models, including hydrodynamic, sediment transport, eutrophication, transport chemical fate, and
food web bioaccumulation models.
The Enhanced Monitoring Program Quality Assurance Program Plan
The Enhanced Monitoring Program Quality Assurance Program Plan (USEPA, 1997d) was developed in
1993 to ensure that data generated from the LMMB study supports its intended use.
LMMB Methods Compendium
The Lake Michigan Mass Balance Project (LMMB) Methods Compendium (USEPA, 1997a, 1997b)
describes the sampling and analytical methods used in the LMMB Study. The entire three volumes are
available on GLNPO's website mentioned above.
LMMB Data Reporting Formats and Data Administration Plan
Data management for the LMMB Study was a focus from the planning stage through data collection,
verification, validation, reporting, and archiving. The goal of consistent and compatible data was a key to
the success of the project. The goal was met primarily through the development of standard formats for
reporting environmental data. The data management philosophy is outlined on the LMMB website
mentioned above.
Lake Michigan LaMP
"Annex 2" of the 1972 Canadian-American Great Lakes Water Quality Agreement (amended in 1978,
1983, and 1987) prompted development of Lakewide Area Management Plans (LaMPs) for each Great
Lake. The purpose of these LaMPs is to document an approach to reducing input of critical pollutants to
the Great Lakes and restoring and maintaining Great Lakes integrity. The Lake Michigan LaMP calls for
basin-wide management of toxic chemicals.
GLENDA Database
Central to the data management effort is a computerized data system to house Lake Michigan Mass
Balance and other project results. That system, the Great Lakes Environmental Monitoring Database
(GLENDA), was developed to provide data entry, storage, access and analysis capabilities to meet the
needs of mass balance modelers and other potential users of Great Lakes data.
1-15
-------
Results of the LMMB Study: Atrazine Data Report
LMMB Data Reports
This report is one in a series of data reports that summarize the data from monitoring associated with
EPA's Lake Michigan Mass Balance Study. In addition to this data report on atrazine, data reports are
being published for PCBs and frara-nonachlor (USEPA, 200If) and mercury (USEPA, 200lg).
Future Documents and Products
Following the completion of modeling efforts associated with the LMMB Study, GLNPO anticipates
publishing reports summarizing the modeling results. In 2005, GLNPO also anticipates conducting a
reassessment of Lake Michigan to calibrate and confirm modeling results with data collected 10-years
after the initial LMMB sampling.
1-16
-------
Chapter 2
Atrazine Study Overview
Figure 2-1. Chemical Structure of Atrazine
Cl
2.1 Atrazine Introduction
2.1.1 Physical/Chemical Properties
Atrazine, or 6-chloro-N-ethyl-N'-(l-methylethyl)-l,3,5-
triazine-2,4-diamine, is a triazine herbicide with a chemical
formula of C8H14C1N5 and the chemical structure displayed in
Figure 2-1. Atrazine is a white crystalline solid with a
molecular weight of 215.7, a melting point of 171-174°C, and
a boiling point of 279°C. The vapor pressure of atrazine is
0.04mPa (at 20°C), the solubility of atrazine in water is 30
mg/L (at 20°C), and the octanol-water partition coefficient
(log Kow) for atrazine is 2.34 (Worthing, 1991).
2.1.2 Atrazine Use
Atrazine is a systemic herbicide that inhibits photosynthetic electron transport and is used for control of
broadleaf weeds and some grassy weeds. Atrazine is the most widely used herbicide in the U.S. for corn
and sorghum production, with approximately three-fourths of all field corn and sorghum treated with
atrazine (USEPA, 200 Ih). Atrazine is also used for weed control on crops of sugarcane, wheat, guava,
macadamia nuts, orchard grass and hay, range grasses, and southern turf grasses. Atrazine is primarily
applied directly to the soil during pre-planting or pre-emergence. Less commonly, atrazine may be
applied post-emergence by foliar application directly to target plants.
Approximately 64 to 75 million pounds of atrazine are applied per year in the U.S. (USEPA, 2001h). It is
primarily used in the "Corn Belt," the region stretching from eastern Nebraska to Ohio, immediately
south and west of the
Figure 2-2. Estimated Annual Usage Rates of Atrazine in the United States for Great Lakes (Goolsby et
1991-1995 (from U.S. Geological Survey, Pesticide National Synthesis Project, al, 1997). Some of the
1998) most intensive use of
atrazine is in the upper
Midwest directly
surrounding Lake
Michigan (Figure 2-2).
In total pounds of atrazine
applied to corn crops,
Illinois ranks first and
Indiana ranks third
(Ribaudo and Bouzaher,
1994). Michigan and
Wisconsin, which also
border Lake Michigan,
rank seventh and eighth,
respectively, in total
pounds of atrazine
applied to corn crops.
Atrazine use,
in pounds per square mile
I I o I I <1.080
1.080-5.587
21.212-66.515
5.588-21.211
>66.516
2-1
-------
Results of the LMMB Study: Atrazine Data Report
2.1.3 Regulatory Background
Atrazine was initially registered as an herbicide in 1958 by Ciba-Geigy and has been used in the U.S. for
over 40 years. Due to the extensive use of this herbicide and potential ecological and human health
effects, EPA began to institute controls in the early nineties. In 1990, atrazine was classified as a
restricted-use pesticide, application rates were reduced, most non-crop land uses were prohibited,
application through irrigation systems was prohibited, and well-head protection plans were instituted.
These controls were strengthened in 1992 through further reductions in application rates, expansion of
setback requirements, and institution of construction requirements for bulk storage facilities (USEPA,
2001h).
To provide human health protection, EPA set a maximum contaminant level of 3 i-ig/L for atrazine in
drinking water. Atrazine was originally classified as a possible human carcinogen, and based on this
cancer risk, EPA initiated a special review for atrazine, simazine, and cyanazine in 1994. EPA has
completed a preliminary risk assessment for atrazine including ecological and human health risks
(USEPA, 200 Ib). As part of this review and assessment, the Health Effects Division's Cancer
Assessment Review Committee recently concluded that atrazine is "not likely to be carcinogenic to
humans" (USEPA, 200Ih).
To protect aquatic life, EPA recently proposed ambient water quality criteria for atrazine (USEPA, 200la;
USEPA, 200Id). The draft criterion for protection against acute toxicity in freshwater (Criterion
Maximum Concentration) was set at 350 i-ig/L. The draft criteria for protection against chronic toxicity in
freshwater (Criterion Continuous Concentration) was set at 12 i-ig/L.
2.1.4 Fate and Effects
Applied atrazine may reach surface water resources through spray drift, runoff, contaminated
groundwater discharge to surface water, or atmospheric deposition in precipitation, vapor, or particulate
phases. The relatively low adsorption characteristics of atrazine and its solubility in water make it
relatively mobile and susceptible to leaching and transport in runoff (USEPA, 200 le). Atrazine loadings
from runoff are highest on highly sloped lands and when intense rain events directly follow herbicide
application. Atrazine concentrations in agricultural field runoff have been measured in the low mg/L
range, but these concentrations are typically diluted to the i-ig/L range once they enter receiving streams
(USEPA, 200 la). Atmospheric sources of atrazine can originate from spray drift during application,
airborne transport with associated soil particles suspended by wind erosion or planting and tilling
operations, and volatilization from soil or foliage surfaces (Glotfelty etal., 1989). In agricultural regions,
atrazine is generally present in the atmosphere throughout the growing season and available for removal
via wet and dry deposition (Goolsby et al., 1997; Williams et al., 1992).
Atrazine is moderately susceptible to aerobic degradation in soils, with a half-life of 60-150 days
(Ribaudo and Bouzaher, 1994). Under anaerobic conditions, this degradation rate slows dramatically to a
half-life of 660 days. Studies on the persistence of atrazine in water have produced varied results with
half-lives of several days reported in some experimental wetland systems and artificial streams to half-
lives of over 300 days in larger lake systems (USEPA, 2001a). The persistence of atrazine in surface
waters with relatively long hydraulic residence times and relatively low microbial activity is due to its
resistence to abiotic hydrolysis and direct aqueous photolysis, its limited volatilization potential, and its
moderate susceptibility to biodegradation (USEPA, 200le).
2-2
-------
Atrazine Study Overview
Figure 2-3. Chemical Structure for Atrazine
Breakdown Products, Deethyl-atrazine (DEA) and
Deisopropyl-atrazine (DIA)
DEA
NH7
DIA
H2N
Degradation products of atrazine include deethyl-
atrazine (DEA), deisopropyl-atrazine (DIA),
diaminochloro-triazine (DACT), and hydroxy-
atrazine (HA). The relative concentration of these
degradates in soil is atrazine » DEA > DIA >
DACT ~ HA. In surface water, the relative
concentrations are similar, with atrazine » DEA >
DIA ~ DACT (USEPA, 200 Ic). The chemical
structures of the two most common degradation
products (DEA and DIA), which were studied in the
Lake Michigan Mass Balance Study, are shown in
Figure 2-3. These same metabolites are also
produced in the atmosphere, where abiotic
degradation of atrazine is primarily carried out by
OH radical attack (Van Dijk and Guicherit, 1999).
In general, atrazine is not very toxic to aquatic
animals. Species mean values for acute toxicity
ranged from 720 |ig/L of atrazine for the midge,
Chironomus tentans, to 49,000 i-ig/L for Daphnia
magna (USEPA, 200la). Species mean values for
chronic toxicity ranged from 88.32 i-ig/L for brook
trout, Salvelinus fontinalis, to 2935 i-ig/L for
Ceriodaphnia dubia (USEPA, 200la). Because
atrazine inhibits photosynthesis, atrazine is
considerably more toxic to aquatic plants than aquatic invertebrates and vertebrates. Toxicity to aquatic
plants, including algae and macrophytes, commonly occurs at 10 i-ig/L of atrazine and above (USEPA,
200la). Recent studies also have suggested that atrazine may cause endocrine disruption and
reproductive abnormalities in amphibians at doses as low as 0.1 i-ig/L (Hayes et al, 2002).
Preliminary environmental risk assessments for atrazine, conducted by the EPA Office of Pesticides
Program's Environmental Fate and Effect Division, have indicated the possibility of chronic effects on
mammals, birds, fish, aquatic invertebrates, and non-target plants at maximum, and in some cases, typical
use rates (USEPA, 200 le). The data strongly suggest that atrazine will have direct negative impact on
freshwater and estuarine plants as well as indirect effects on aquatic invertebrate and fish populations that
rely on aquatic plants for habitat and a food chain base (USEPA, 200le).
2.2 Study Design
2.2.1 Description
Atrazine was selected as one of four contaminants to be modeled in the Lake Michigan Mass Balance
Study. In this study, atrazine and atrazine metabolites (DEA and DIA) were measured in atmospheric,
tributary, and open-lake water column components of the Lake Michigan ecosystem from March 1994
through October 1995. The data generated from this study was used to estimate an overall mass balance
of atrazine in Lake Michigan and meet the defined objectives of the LMMB Project (see Section 1.4).
2.2.2 Scope
To develop a mass balance of atrazine in Lake Michigan, all significant sources and stores of atrazine in
the environment were measured. Significant sources included tributary inputs and atmospheric inputs
2-3
-------
Results of the LMMB Study: Atrazine Data Report
Figure 2-4. Sources and Stores of Atrazine Measured in the
LMMB Study
Atmospheric
Vapor
(294 samples)
Parti cu late
(243 samples)
Precipitation
(207 samples)
Tributaries
(108 samples)
from the vapor phase, particulate phase, and
precipitation (Figure 2-4). The open-lake
water column compartment was determined
to be the only significant store of atrazine.
The sediment compartment was not
included in this study because atrazine is
relatively water soluble. Atrazine also does
not bioaccumulate in living tissue, so biotic
compartments were not included.
Atmospheric samples were collected from
March 15, 1994 to October 20, 1995. These
atmospheric samples were collected from 8
shoreline sampling stations, 3 out-of-basin
sampling stations, and 18 open-lake
sampling stations. Atmospheric samples
were collected in three separate sampling
mediums or phases: vapor, particulate, and
precipitation. A total of 294 vapor phase
samples, 226 particulate samples, and 207
precipitation samples were collected and
analyzed for atrazine. All but 64 of these
samples were analyzed for DBA and DIA as well as atrazine. Samples from Sleeping Bear Dunes that
were processed at Indiana University were not analyzed for DBA and DIA.
Tributary samples were collected from April 4, 1995 to October 31, 1995. A total of 108 samples were
collected from 11 tributaries that flow into Lake Michigan. Open-lake water column samples from Lake
Michigan were collected from April 25, 1994 to April 17, 1995. A total of 234 open-lake samples were
collected from 35 sampling stations located throughout Lake Michigan, 2 sampling stations located in
Green Bay, and 1 sampling station located on Lake Huron. All tributary and open-lake samples were
analyzed for atrazine, DBA, and DIA.
2.2.3 Organization/Management
The responsibility for collecting and analyzing atrazine samples from the various components was
divided among several principal investigators (Figure 2-5). All tributary and open-lake samples were
collected and analyzed at the University of Minnesota Gray Freshwater Biological Institute and Rutgers
University under the supervision of principal investigator, Steve Eisenreich. From April 1994 through
July 1994, atmospheric samples from the Sleeping Bear Dunes site were collected and analyzed at the
Illinois State Water Survey under the supervision of principal investigator, Clyde Sweet. From August
1994 through October 1995, atmospheric samples from the Sleeping Bear Dunes site were collected and
analyzed at Indiana University under the supervision of principal investigator, Ron Hites. Atmospheric
samples from all other sites also were collected and analyzed at the Illinois State Water Survey.
Each principal investigator developed a Quality Assurance Project Plan (QAPP) that was submitted to
EPA's Great Lakes National Program Office. The QAPPs detailed the project management, study design,
and sampling and analysis procedures that would be used in the study and the quality control elements
that would be implemented to protect the integrity of the data. The LMMB quality assurance (QA)
program is further discussed in Section 2.6, and detailed information on the quality assurance activities
and data quality assessment specific to each ecosystem component are discussed in Chapters 3-5.
2-4
-------
Atrazine Study Overview
Figure 2-5. Principal Investigators Responsible for Sampling
and Analysis of Atrazine in Atmospheric, Tributary, and Open-
lake Water Column Components
Study Component
Atmospheric
Sleeping Bear Dunes(4/94-7/94)
All other sites
2.3 Sampling Locations
2.3.1 Atmospheric Components
Eight shoreline sampling stations were
established on Lake Michigan for the
collection of atmospheric samples. In
addition, three out-of-basin sampling
stations were established as regional
background sites to represent air coming
over Lake Michigan during periods of
southwest or northwest prevailing winds.
Figure 2-6 shows the location of each of
these sampling locations. Sampling
locations for the LMMB Project were
selected through workshop discussions.
Site selection criteria considered
predominant annual wind directions,
source areas, and episodic summer events.
In general, sites were selected to be
regionally representative of the following
land use categories:
remote - no urban areas or major
sources of air pollutants within 50 km
(Beaver Island, Sleeping Bear Dunes,
Eagle Harbor, and Brule River sites)
urban - major urban sources within 1 km (Chicago IIT site)
urban-influenced - major urban sources within 10 km (Muskegon, Manitowoc
and Indiana Dunes sites)
rural - urban sources generally more than 10 km away, but agriculture sources
Haven and Bondville sites)
Sleeping Bear Dunes(8/94-10/95)
Tributary
Open Water
Principal
Investigator
Clyde Sweet -
Illinois State
Water Survey
Ron Hites -
Indiana University
Steve Eisenreich -
Rutgers
University
, Chiwaukee Prairie,
within 1 km (South
Sampling at the shoreline and background sites was conducted on a regular schedule between April, 1994
and October, 1995. Atrazine, DEA, and DIA also were measured in air and precipitation collected at 14
locations over Lake Michigan during periodic cruises of the Research Vessel (R/V), Lake Guardian
(April-May, 1994; June, 1994; August, 1994; October-November, 1994; January, 1995; March-April,
1995; August, 1995; September-October, 1995).
2-5
-------
Figure 2-6. Atmospheric Sampling Stations
I Lake
Michigan
«Q
Beaver rj
Island
Sleeping Bear
Dunes
H
Lake
Michigan
Must
uskegon
Wisconsin
Illinois
Chiwaukee
Prairie
NT Chicago1^
n South Haven
25 ii
2-6
-------
Atrazine Study Overview
2.3.2 Tributaries
Tributary water column samples were collected from eleven rivers that flow into Lake Michigan (Figure
2-7). These tributaries included the Menoninee, Fox, Sheboygan, and Milwaukee Rivers in Wisconsin;
the Grand Calumet River in Indiana; and the St. Joseph, Kalamazoo, Grand, Muskegon, Pere Marquette,
and Manistique Rivers in Michigan. With the exception of the Pere Marquette River, these tributaries
were selected for the LMMB Study because of elevated concentrations of PCBs and mercury in resident
fish. The Pere Marquette River was selected because it has a fairly large and pristine watershed. The 11
monitored tributaries represent greater than 90% of the total river flow into Lake Michigan and an even
higher percentage of the total tributary load of pollutants into Lake Michigan.
Table 2-1 describes specific
watershed characteristics and
impairment information for each
of the monitored tributaries. Of
the 11 tributaries, 6 (the
Kalamazoo, Manistique,
Menominee, Fox, Sheboygan,
and Grand Calumet Rivers) are
classified as Great Lakes areas of
concern (AOCs). Areas of
concern are severely degraded
geographic areas within the
Great Lakes Basin. They are
defined by the US-Canada Great
Lakes Water Quality Agreement
(Annex 2 of the 1987 Protocol)
as "geographic areas that fail to
meet the general or specific
objectives of the agreement
where such failure has caused or
is likely to cause impairment of
beneficial use or the area's
ability to support aquatic life."
Most of the 11 tributaries are
also listed on the Clean Water
Act Section 303(d) list of
impaired water bodies due to
contamination from mercury,
PCBs, and other pollutants.
Figure 2-7. Tributary Sampling Stations
? Michigan
' --V-
\^_ Indiana
2-7
-------
Table 2-1. Watershed Characteristics for Tributaries Monitored in the LMMB Study
Tributary
St. Joseph
Kalamazoo
Grand (lower)
Muskegon
Pere Marquette
Manistique
Menominee
Fox (lower)
Sheboygan
Milwaukee
Grand Calumet
Watershed
area
(mi2)
4685
2047
2003
2686
2644
1464
2306
442
2201
864
1039
Total river
miles in
watershed
3743
1560
2014
1886
1356
1061
1660
700
1699
802
760
Riparian Habitat
Forested
25-50%
25-50%
25-50%
25-50%
25-50%
>75%
>75%
25-50%
25-50%
25-50%
25-50%
Agricultural
/ Urban
>50%
>50%
>50%
>50%
>50%
20-50%
20-50%
>50%
>50%
>50%
>50%
IWI Score3
3- less serious problems, low
vulnerability
3- less serious problems, low
vulnerability
5- more serious problems, low
vulnerability
5- more serious problems, low
vulnerability
3- less serious problems, low
vulnerability
1- better quality, low
vulnerability
1- better quality, low
vulnerability
6- more serious problems, high
vulnerability
5- more serious problems, low
vulnerability
5- more serious problems, low
vulnerability
5- more serious problems, low
vulnerability
Impaired forb
E. coli, mercury, PCBs, pathogens,
macro-invertebrate community
mercury, PCBs
PCBs, pathogens
mercury, PCBs
mercury, PCBs, pathogens
dioxin, PCBs, mercury, pathogens
PCBs, organic enrichment, dissolved
oxygen
PCBs, mercury
PCBs
PCBs, pesticides, lead, mercury,
dissolved oxygen, cyanide, chlorides,
impaired biotic community, oil and
grease, copper
Area of
Concern
X
X
X
X
X
X
a EPA's Index of Watershed Indicators Score for assessing the health of aquatic resources.
b Based on 1998 listing of Clean Water Act Section 303(d) impaired waters.
2-8
-------
Atrazine Study Overview
Figure 2-8. Open-lake Water Column Sampling Stations
2.3.3 Open Lake
Open-lake water column samples were
collected from 35 sampling locations
on Lake Michigan, 2 sampling
locations in Green Bay, and 1
sampling location on Lake Huron
(Figure 2-8). Ten of these stations
were identified as master stations
where increased resolution sampling
was conducted. Open-lake stations
were selected to represent good spatial
coverage throughout the lake and to
include both nearshore (<25-30 meters
in depth) and offshore locations.
Many of the open-lake sampling
stations were located at existing
NOAA (National Oceanic and
Atmospheric Administration) weather
buoys or sites where additional
information had previously been
collected.
Open-lake samples were collected
during six cruises of the R/VLake
Guardian. These cruises were
conducted from April 25, 1994 to
April 17, 1995. The first survey
occurred in the early spring just after ice out (April/May 1994). The second survey was in early summer
(June 1994) after the onset of stratification and following the spring runoff period of agricultural
chemicals from cropland. The third survey was in late summer (August 1994) during later stages of
stratification. The fourth and fifth surveys, conducted in October 1994 and January 1995, sampled only a
few of the Lake Michigan sites. The final survey occurred in March/April 1995 just after ice out.
2.4 Sampling Methods
Full details of the sampling methods used in the LMMB Study have been published by EPA in a Methods
Compendium (USEPA, 1997a; USEPA, 1997b). A brief summary is provided below.
2.4.1 Atmospheric Components
Twenty-eight-day composite precipitation samples were collected using a MIC-B wet only sampler
(Meterological Instruments of Canada, Thornhill, ONT) with a 0.212 m2 stainless steel catch basin. The
sampler was modified for all-weather operation by enclosing and insulating the space underneath the
sampler. The temperature in the enclosure was maintained at 10 to 15°C during the winter using a small
space heater. The collector also was fitted with a precipitation sensor and a retractable cover. The catch
basin remained covered to prevent evaporation until precipitation was detected by the sensor. Rain or
melted snow collected in the catch basin passed by gravity flow through a 30-cm XAD-2 resin column
that absorbed atrazine in the precipitation sample. Glass wool plugs inserted before and after the XAD-2
2-9
-------
Results of the LMMB Study: Atrazine Data Report
resin trapped any particles in the sample. Resin columns and plugs were changed at 28-day intervals and
shipped to the lab for analysis. At this time, the precipitation collection runnel was rinsed with water and
wiped with a piece of clean quartz fiber filter paper to remove adhering particles. The filter paper and
rinsings were then included as part of the sample. Samples were sealed with Teflon caps, transported to
the testing laboratory, and stored in air-tight containers at -18°C until analysis.
Twenty-four-hour composite air samples for atrazine analysis were collected every 3 to 12 days using a
modified high-volume sampler. The sampler was modified to include an aluminum tube behind the filter
holder that accommodated a vapor trap consisting of a stainless steel cartridge of XAD-2 resin. Air flow
was maintained at a rate of 34 m3/hr during sampling. Particulate phase atmospheric samples were
collected on pre-fired quartz fiber filters, and vapor phase atmospheric samples were collected in the
XAD-2 resin vapor trap. Samples were wrapped in aluminum foil, sealed in air tight metal cans,
transported to the testing laboratory, and stored at -18°C until analysis.
For most atmospheric sampling sites, individual 24-hour samples were combined in the laboratory to
provide monthly composite vapor trap and filter samples. This monthly composite sample, which
consisted of multiple 24-hour composites, was then analyzed at the testing laboratory. At the Sleeping
Bear Dunes site and occasionally at other sites, individual 24-hour XAD-2 samples were analyzed
without compositing. In these cases, the results from individual 24-hour samples were mathematically
composited (by volume weighting) to obtain monthly atrazine values.
Each shoreline site had a 10-meter meteorological tower and a number of meteorological instruments
including wind speed and wind direction sensors at a height of 10 m (Met-One, Grants Pass, OR), a solar
radiation sensor (Ll-Cor, model LI 200S, Lincoln, NE), temperature and relative humidity sensors
(Campbell Scientific, Logan, UT), and a standard Belfort rain gauge (Belfort Instrument, Baltimore, MD)
with a Nipher wind shield. All of the meteorological sensors were automatically recorded every six
seconds using a datalogger (Campbell Scientific, model 2IX, Logan, UT).
2.4.2 Tributaries
The number and timing of sampling events were dependent upon the stability of the tributary and the
timing of increased flow events. Tributaries with greater stability (i.e., those that are less responsive to
precipitation events) were sampled less frequently than those that were more variable. Sampling was also
timed to collect approximately one-third of samples during base flow conditions and approximately two-
thirds of samples when flows were above the 20th percentile.
Tributary samples were collected as near to river mouths as possible without being subject to flow
reversals that are common near river mouths in Lake Michigan. Composite samples were obtained using
the USGS quarter-point sampling procedure. In this procedure, the stream is visually divided into three
equal flow areas. At the center of each flow area, samples were collected from 0.2 and 0.8 times the
depth. All six samples were then composited and pumped (using a peristaltic pump) through a 0.7 |im
glass fiber filter. The filtrate was then passed through a large, 250 g, XAD-2 resin column to trap
dissolved organics. Samples were then chilled and delivered to the testing laboratory. While the
measured atrazine concentrations in tributary samples represent the filtered phase, these concentrations
should generally approximate total atrazine concentrations due to the solubility of atrazine.
2-10
-------
Atrazine Study Overview
2.4.3 Open Lake
Open-lake samples were collected from various depths depending upon the stratification conditions.
During stratification, open-lake stations were sampled at the mid-epilimnion and mid-hypolimnion.
During non-stratified periods, samples were collected at mid-water column depth and two meters below
the surface. Master stations, during times of non-stratification, were sampled at mid-water column depth,
one meter below the surface, and two meters off the bottom. During times of stratification, master
stations were sampled at one meter below the surface, mid-epilimnion, mid-hypolimnion, and two meters
off the bottom. In addition, stations 18 and 41 were sampled at the thermocline and 2 meters below the
surface during stratification.
Water samples were collected using a General Oceanics (Model 1015) rosette sampler on board the R/V
Lake Guardian. Water was transferred from individual rosette canisters to amber one liter bottles, and
stored at 4°C until processing at the testing laboratory.
2.5 Analytical Methods
Full details of the analytical methods used in the LMMB Study have been published by EPA in a methods
compendium (USEPA, 1997a; USEPA 1997b). The sample collection, extraction, and analysis methods
for the atmospheric components and water samples from the tributaries and open lake are modifications of
the methods used for the PCBs and chlorinated pesticides. These modifications include the changes to the
extraction conditions needed to extract more polar analytes like atrazine, DEA, and DIA, and changes to
chromatographic conditions such as columns and temperature programs necessary for the instrumental
analysis. A brief summary is provided below.
2.5.1 Atmospheric Components
Atrazine and atrazine metabolites were extracted from XAD-2 resin or filter samples by Soxhlet
extraction with 300 mL of a 1:1 hexane and acetone mixture. The extract was concentrated by rotary
evaporation to about 3 mL and then the solvent was exchanged to hexane with rotary evaporation. The
concentrated extract was subjected to clean-up on 3% deactivated silica with a sodium sulfate cap to
remove most of the non-target, interfering compounds. Atrazine, DEA, and DIA were eluted in a
methanol fraction. The volume of each fraction was reduced to between 0.3 and 1 mL by rotary
evaporation and concentrated in a stream of high-purity nitrogen. Samples were transferred to vials,
capped, and stored at -20°C. Analysis of atrazine, DEA, and DIA was conducted using gas
chromatography coupled to a mass spectrometer detector.
2.5.2 Tributaries and Open Lake
Atrazine, DEA, and DIA were isolated from filtered water samples using 250 mg Carbopack (Supelco
Corp) solid-phase extraction (SPE) cartridges. Analytes were eluted from the SPE using 7 mL of a 90%
dichloromethane, 10% methanol solution (vol:vol), followed by 5 mL of methanol. The eluent was then
passed through clean anhydrous sodium sulfate to remove excess water. Extracts were concentrated to
<100 |^L under a nitrogen gas stream. Analysis of atrazine, DEA, and DIA was conducted using gas
chromatography coupled to a mass spectrometer detector.
2-11
-------
Results of the LMMB Study: Atrazine Data Report
2.6 Quality Implementation and Assessment
As described in Section 1.5.5, the LMMB QA program prescribed minimum standards to which all
organizations collecting data were required to adhere. The goal of the QA program was to ensure that all
data gathered during the LMMB Study met defined standards of quality with specified levels of
confidence. Data quality was defined, controlled, and assessed through activities that included
development of study QAPPs, use of SOPs, and data verification. These activities are described in detail
in The Lake Michigan Mass Balance Study Quality Assurance Report (USEPA, 200 li). Specific quality
control elements implemented in the sampling and analysis of atrazine included:
use of Standard Operating Procedures and trained personnel for field sampling and laboratory
analysis;
determination of method sensitivity through calculation of method detection limits;
establishment and maintenance of sample holding times;
preparation and analysis of a variety of blanks to characterize contamination associated with specific
sample handling, storage, and analysis processes including field blanks, lab reagent blanks, bottle
blanks, trip blanks, and lab procedural blanks;
collection and analysis of field or laboratory duplicate samples;
preparation and analysis of a variety of quality control samples including performance standards;
use of a standardized data reporting format; and
for atmospheric samples, preparation and analysis of matrix spike samples to characterize the
applicability of the analytical method to the study sample matrices.
Performance or intercomparison studies were not conducted for the atrazine analyses. However, all of the
Pis obtained the atrazine calibration standard from the same vendor and the sample preparation,
extraction, and instrumental analysis (GC/MS with Selected Ion Monitoring) procedures were similar
among all Pis. In addition, each researcher's laboratory was audited during an on-site visit at least once
during the time LMMB samples were being analyzed. The auditors reported positive assessments and did
not identify issues that adversely affected the quality of the data. Prior to data submission, each
researcher submitted electronic test files containing field and analytical data according to the LMMB data
reporting standard. GLNPO reviewed these test data sets for compliance with the data reporting standard
and provided technical assistance to the researchers.
Prior to sample collection, Quality Assurance Project Plans (QAPPs) were developed by the Pis and
submitted to GLNPO for review. Because the open-lake and tributary monitoring were conducted by the
same PI, a separate QAPP was not prepared for the tributary monitoring, and GLNPO and the PI agreed
to implement the procedures outlined in the open-lake QAPP for the tributary sampling and analysis. In
the QAPPs, the Pis defined MQOs in terms of six attributes: sensitivity, precision, accuracy,
representativeness, completeness, and comparability. The MQOs were designed to control various phases
of the measurement process and to ensure that the total measurement uncertainty was within the ranges
prescribed by the DQOs. The MQOs for atrazine are listed in Section 4 of The Lake Michigan Mass
Balance Study Quality Assurance Report (USEPA, 200 li).
The Pi-defined MQOs also were used in the data verification process. GLNPO conducted data
verification through the LMMB QA Workgroup. The workgroup was chaired by GLNPO's Quality
Assurance Manager and consisted of quality control coordinators that were responsible for verifying the
quality of specific data sets. Data verification was performed by comparing all field and QC sample
results produced by each PI with their MQOs and with overall LMMB Study objectives. If the results
failed to meet MQOs and corrective actions were not feasible, the results were flagged to inform data
users of the failure. These flags were not intended to suggest that data were not useable; rather they were
intended to caution the user about an aspect of the data that did not meet the predefined criteria. In this
2-12
-------
Atrazine Study Overview
data report, the summary and analysis of atrazine data represent all results with the exception of those
flagged as "invalid" by the QC coordinator in concert with the PI.
In addition to flags related to the evaluation of MQOs, a wide variety of flags were applied to the data to
provide detailed information to data users. For example, the flag LAC (laboratory accident, no result
reported) was applied to sample results to document that a sample was collected, but no result was
reported due to a laboratory accident. The frequencies of flags applied to atrazine study data are provided
in the Quality Implementation Sections of each of the following chapters. The flag summaries include the
flags that directly relate to evaluation of the MQOs to illustrate some aspects of data quality, but do not
include all flags applied to the data to document sampling and analytical information (such as LAC). In
order to provide detailed quality information to data users, the study data are maintained in the GLENDA
database with all applied flags. Detailed definitions of the flags can be found in the Allowable Codes
Table on GLNPO's website at: www.epa.gov/glnpo/glenda/codes/codes.html under Result Remark, List
of QC flags (lab_rmrk).
The Pis participating in the study also conducted real-time data verification. Pis applied best professional
judgement during sampling, analysis, and data generation, based on their experience monitoring atrazine
in the environment. In most cases, when sample results were questionable, the PI reanalyzed the sample
or clearly documented the data quality issues in the database through the application of data quality flags
or by including comments in the database field, "Exception to Method, Analytical." Because the flags
and comments are maintained in the database for each sample result, data users are fully informed of data
quality and can evaluate quality issues based on their intended use of the data. The level of
documentation that GLNPO is maintaining in the study database is unprecedented for a database of this
size and will serve as a model for future efforts.
GLNPO also conducted data quality assessments in terms of three of the six attributes used as the basis
for the MQOs, specifically sensitivity, precision, and bias. These three attributes could be assessed from
the results of QC samples analyzed in the study. For example, system precision was estimated as the
mean relative percent difference (RPD) between results for field duplicate pairs. Similarly, analytical
precision was estimated as the mean RPD between results for laboratory duplicate pairs. Bias was
estimated using the mean recovery of spiked field samples or other samples of known concentration such
as laboratory performance standards. A summary of data quality assessments is provided for the atrazine
study data in the Quality Implementation Section of each of the following chapters.
2-13
-------
Chapter 3
Atrazine in Atmospheric Components
3.1 Results
From March 15, 1994 to October 20, 1995, atmospheric samples were collected from 8 shoreline
sampling stations, 3 out-of-basin sampling stations, and 18 sampling stations in the open lake (over-
water) (Table 3-1). Atmospheric samples were collected, analyzed, and results reported in three separate
sampling mediums or phases: vapor (in pg/m3), particulate (in pg/m3), and precipitation (in ng/L). A total
of 294 vapor phase samples, 226 particulate samples, and 207 precipitation samples were collected and
analyzed for atrazine. All but 64 of these samples also were analyzed for deethyl-atrazine (DBA) and 6-
deisopropyl-atrazine (DIA). Samples from Sleeping Bear Dunes that were processed at Indiana
University were not analyzed for DBA and DIA.
Table 3-1. Number of Atmospheric Samples Analyzed for Atrazine, PEA, and DIA
Sampling Station
Shoreline
Atmospheric
Stations
Out-of-basin
Atmospheric
Stations
Beaver Island
Indiana Dunes
I IT Chicago
Muskegon
Manitowoc
Sleeping Bear
Dunes3
South Haven
Chiwaukee
Prairie
Brule River
Bondville
Eagle Harbor
Sampling Dates
3/15/94-10/8/95
3/15/94-10/20/95
3/15/94-10/2/95
3/15/94-10/13/95
3/15/94-10/8/95
4/1/94-10/19/95
3/15/94-10/8/95
3/15/94-10/2/95
4/1/94-10/8/95
3/15/94-10/20/95
4/1/94-7/31/94
Vapor
Samples
Analyzed
17
29
25
17
18
45b
21
20
17
21
8
Particulate
Samples
Analyzed
17
29
24
16
18
19C
18
20
17
20
4
Precipitation
Samples
Analyzed
20
21
17
20
20
18d
21
20
19
21
4
Total
Samples
Analyzed
54
79
66
53
56
82
60
60
53
62
16
3-1
-------
Results of the LMMB Study: Atrazine Data Report
Sampling Station
Over-water
Atmospheric
Stations
1
11
110
18M
23M
27M
280
310
380
40M
41
47M
5
6
GB17
GB24M
MB19M
geographical
composite6
Sampling Dates
5/10/94-10/11/95
5/8/94
4/8/95 - 9/23/95
5/6/94-10/9/95
5/4/94-10/3/95
5/2/94 - 9/27/95
10/26/94-10/1/95
3/28/95-10/8/95
10/31/94-1/23/95
10/18/94-9/25/95
4/30/94-8/12/94
8/7/94-9/19/95
11/4/94-10/10/95
8/25/94-10/12/95
4/12/95
10/17/94-9/20/95
1/24/95
4/30/94-10/11/95
Total
Vapor
Samples
Analyzed
4
1
3
5
4
5
4
3
1
4
2
5
6
4
4
1
294
Particulate
Samples
Analyzed
1
1
1
1
3
1
2
14
226
Precipitation
Samples
Analyzed
1
1
1
1
1
1
207
Total
Samples
Analyzed
6
1
3
5
6
5
4
3
2
5
3
5
10
5
1
7
1
14
727
a Samples from Sleeping Bear Dunes that were processed at Indiana University (8/2/94 -10/19/95) were not
analyzed for DBA and DIA.
b 45 samples were analyzed for atrazine, and 10 samples were analyzed for DBA and DIA.
019 samples were analyzed for atrazine, and 4 samples were analyzed for DBA and DIA.
d 18 samples were analyzed for atrazine, and 4 samples were analyzed for DBA and DIA.
e Samples were collected while the R/VLake Guardian was traveling between stations. These samples are
considered geographical composites.
3.1.1 Vapor Fraction
Atrazine was not detected in the vapor fraction of most atmospheric samples collected during the 1994-
1995 sampling campaign. Of 294 vapor samples tested, atrazine was detected (i.e., measured above zero)
in only 40 samples (14%), and measured atrazine concentrations were above sample-specific detection
limits in only 11 (3.7%) samples (Figure 3-1). Method detection limits (MDLs) were calculated from
results of seven spiked samples according to the procedures specified at 40 CFRpart 136, Appendix B.
3-2
-------
Atrazine in Atmospheric Components
The MDL was then adjusted for each
sample based on the analyzed sample
volume and surrogate recovery
factors to obtain sample-specific
detection limits. Sample-specific
detection limits ranged from 13 to
104 pg/m3 for samples analyzed at the
Illinois Water Survey, and from 8.89
to 49.7 pg/m3 for samples analyzed at
Indiana University. Sample-specific
detection limits averaged 32 pg/m3
and 20.7 pg/m3 for samples analyzed
at the Illinois Water Survey and
Indiana University, respectively.
Figure 3-1. Detection Frequency of Atrazine in Vapor Phase
Samples
Atrazine not detected
Atrazine measured
below sample-
specific detection limit
Atrazine measured
above sample-specific
detection limit
Table 3-2 lists the 11 vapor samples
that contained measured atrazine
concentrations above sample-specific
detection limits. These samples included one individual sample from South Haven, four composite
samples from Bondville, and six individual samples from Sleeping Bear Dunes. Four of the six samples
from Sleeping Bear Dunes that contained atrazine above sample-specific detection limits (those collected
on 10/23/94, 11/4/94, 8/8/95, and 10/7/95) were flagged for contamination of corresponding field or
laboratory blanks. The concentrations measured in these four samples were only 1.0, 3.6, 1.4, and 4.5
times the concentration measured in corresponding blanks, so results for these four samples are likely
biased high.
Table 3-2. Atrazine Concentrations in the Vapor Fraction Measured Above Sample-specific Detection Limits
Sampling Location
South Haven
Bondville
Sleeping Bear Dunes
Sampling Date
7/7/94 - 7/8/94
5/7/94 - 5/21/94
6/1/94 - 6/26/94
6/9/95 - 6/22/95
7/3/95 - 7/28/95
10/23/94-10/24/94
11/4/94-11/5/94
11/16/94-11/17/94
8/8/95 - 8/9/95
9/13/95-9/14/95
10/7/95-10/8/95
Sample Description
single 24-hr composite sample
manual composite of 2, 24-hr composite samples
manual composite of 3, 24-hr composite samples
manual composite of 2, 24-hr composite samples
manual composite of 3, 24-hr composite samples
single 24-hr composite sample
single 24-hr composite sample
single 24-hr composite sample
single 24-hr composite sample
single 24-hr composite sample
single 24-hr composite sample
Atrazine
Concentration
(pg/m3)
70
330
420
480
160
18.6
71.1
22.1
11.6
31.5
34.5
3-3
-------
Results of the LMMB Study: Atrazine Data Report
3.1.1.1 Seasonal Variation
Atrazine was detected in an insufficient number of samples to evaluate seasonal trends of atrazine
concentrations in the atmospheric vapor fraction. Atrazine in the vapor fraction was measured above
sample-specific detection limits in only 11 samples, which were collected during the months of May,
June, July, August, September, October, and November (Table 3-2).
3.1.1.2 Geographical Variation
Atrazine was detected in an insufficient number of samples to evaluate geographical trends of atrazine
concentrations in the atmospheric vapor fraction. During the March 1994 through October 1995 sampling
campaign, atrazine was not detected above sample-specific detection limits in any vapor phase samples
from the following stations: Beaver Island, Eagle Harbor, Brule River, Indiana Dunes, IIT Chicago,
Muskegon, Manitowoc, and Chiwaukee Prairie. Vapor-phase atrazine concentrations were measured
above sample-specific detection limits only in samples from South Haven, Bondville, and Sleeping Bear
Dunes (Table 3-2). Atrazine concentrations in these samples ranged from 11.6 to 480 pg/m3, and
maximum detected concentrations were 70, 480, and 71.1 pg/m3 at the South Haven, Bondville, and
Sleeping Bear Dunes sites, respectively.
Two of the three sites where atrazine was detected in the vapor phase were classified as rural sampling
locations (South Haven and Bondville), where local agricultural influences were prevalent. The third site
(Sleeping Bear Dunes) was classified as a remote site, however, four of the six samples from this site that
had atrazine concentrations above sample-specific detection limits were likely biased high due to
laboratory or field contamination (see Section 3.1.1). The two remaining samples from Sleeping Bear
Dunes that contained detectable levels of atrazine (those collected on 11/16/95 and 9/13/95) were not
influenced by blank contamination, but were only slightly above (1.2 and 1.4 times) sample-specific
detection limits. Only samples from Bondville were greater than five times sample-specific detection
limits. These samples from Bondville also represented monthly composites of multiple 24-hr samples,
whereas samples from South Haven and Sleeping Bear Dunes that contained atrazine above sample-
specific detection limits were individual 24-hr samples. This indicates that atrazine concentrations well
above detection levels may be typical in May, June, and July at the Bondville site. At all other sites,
atrazine concentrations are generally below detection, but occasional peaks in atrazine concentration
above detection levels may be experienced in individual samples.
3.113 Analysis of Breakdown Products
DBA and DIA were not detected in any of the 259 vapor samples analyzed. Sample-specific detection
limits for DBA in the vapor phase ranged from 24 to 199 pg/m3 and averaged 62 pg/m3 for samples
analyzed at the Illinois Water Survey. Sample-specific detection limits for DIA in the vapor phase ranged
from 24 to 204 pg/m3 and averaged 63 pg/m3 for samples analyzed at the Illinois Water Survey. Samples
processed at Indiana University were not analyzed for DBA or DIA.
3.1.2 Particulate Fraction
Of 226 particulate samples tested, atrazine was detected (i.e., measured above zero) in 65 samples (29%),
and measured atrazine concentrations were above sample-specific detection limits in 52 (23%) samples
(Figure 3-2). Method detection limits (MDLs) were calculated from results of seven spiked samples
according to the procedures specified at 40 CFR part 136, Appendix B. The MDL was then adjusted for
each sample based on the analyzed sample volume and surrogate recovery factors to obtain sample-
specific detection limits. Sample-specific detection limits ranged from 3.0 to 68 pg/m3 for samples
analyzed at the Illinois Water Survey, and from 26.8 to 284 pg/m3 for samples analyzed at Indiana
3-4
-------
Atrazine in Atmospheric Components
Atrazine not detected
University. Sample-specific detection
limits averaged 17 pg/m3 and 70.7 pg/m3 Figure 3-2. Detection Frequency of Atrazine in Particulate Phase
for samples analyzed at the Illinois Samples
Water Survey and Indiana University,
respectively. Particle size distribution
analysis was not conducted to identify ' 161
the size of atmospheric particles
associated with the detection of atrazine.
3.12.7 Seasonal Variation
Atrazine measured
below sample-
specific detection limit
Atrazine measured
above sample-specific
detection limit
The presence and concentration of
atrazine in the particulate phase was
highly influenced by season. Of the 52
particulate phase samples that contained
atrazine above sample -specific detection
limits, 5 1 were collected during the
months of April through July (Figure 3-
3). Only one particulate phase sample collected from August through March contained atrazine levels
above method detection limits. During the late spring, a majority of samples contained atrazine. Atrazine
was detected above sample-specific detection limits in 92% of particulate samples collected in May and
80% of samples collected in June (Figure 3-3).
Atrazine concentrations in the particulate phase peaked during both the spring of 1994 and the spring of
1995 at all shoreline and out-of basin sampling stations (Figure 3-4). For all sampling stations except for
two (Indiana Dunes and South Haven), peak particulate phase atrazine concentrations were higher in the
spring of 1994 than the spring of 1995. Peak concentrations were observed during the month of May at 9
of the 1 1 sampling stations. Peak concentrations were observed in April and June at the IIT Chicago and
Bondville stations, respectively.
Due to the dramatic seasonal variation in atrazine concentrations, station averages were calculated only
for measurements during the spring and summer growing season, when atrazine was primarily detected in
atmospheric samples. Spring/summer mean atrazine concentrations were calculated by averaging
monthly composite samples that were collected between March 20 (the first day of spring) and September
23 (the first day of fall), based
Figure 3-3. Frequency Distribution of Atrazine Detection in Particulate
Phase Atmospheric Samples
II
Q> "S
Q.T3
EA
0)
_
BJ
100%
80%
60%
40%
2 20%
0%
UJ
UJ
on the midpoint of the
compositing period. When
individual samples were
collected during a month, rather
than monthly composite
samples, the results were
mathematically composited (by
volume weighting) to obtain
monthly composite results,
which were then used to
calculate spring/summer mean
atrazine concentrations. All
results were included in mean
calculations as reported,
meaning that measurements
3-5
-------
Results of the LMMB Study: Atrazine Data Report
below sample-specific detection limits were included and not censored. Results reported as zero also
were included in mean calculations.
Particulate atrazine concentrations in the spring/summer of 1994, 1995, and both years combined are
presented in Table 3-3. For all sampling stations except for two (Sleeping Bear Dunes and South Haven),
mean spring/summer atrazine levels in the particulate phase were higher in 1994 than 1995.
3.12.2 Geographical Variation
Atrazine was detected in the particulate phase at all shoreline and out-of-basin sampling stations.
Maximum monthly atrazine concentrations at these locations ranged from 160 pg/m3 at Sleeping Bear
Dunes to 1400 pg/m3 at Bondville. Mean spring/summer atrazine levels for the two-year sampling
campaign (1994-1995) ranged from 25 pg/m3 at Beaver Island to 370 pg/m3 at Bondville (Table 3-3).
Mean particulate atrazine concentrations differed by more than a factor of 10 among sites, but these
differences were not statistically significant (based on a Kruskal-Wallis test (p = 0.395)), due to the high
variability of measured concentrations at each station.
While mean atrazine concentrations did not differ significantly among stations, it was observed that
atrazine levels in the particulate phase were generally higher at shoreline sampling stations surrounding
the southern Lake Michigan basin than shoreline sampling stations surrounding the northern basin. Mean
spring/summer atrazine levels at all of the southern basin sampling stations (Chiwaukee Prairie, IIT
Chicago, Indiana Dunes, South Haven, and Muskegon) were higher than at any of the northern basin
sampling stations (Manitowoc, Sleeping Bear Dunes, and Beaver Island) (Figure 3-5). This trend is
consistent with the increased use of atrazine in the southern Lake Michigan basin compared to the
northern basin (Figure 2-2). Out-of-basin sampling stations also fit this trend with the exception of the
Eagle Harbor site. The southern out-of-basin sampling station (Bondville), which is in the center of an
intensive corn-growing region, had the highest mean particulate atrazine concentration of any site. The
remote, northern Brule River sampling station had a lower mean atrazine concentration than any of the
southern sites. The northern Eagle Harbor sampling station was the only site that did not fit this trend.
The mean atrazine concentration at the Eagle Harbor sampling station was higher than at three of the
southern basin sampling stations (Chiwaukee Prairie, Muskegon, and South Haven), however, only four
monthly composite samples were collected from the Eagle Harbor site (between April and July 1994), and
atrazine was detected in only one of those monthly composites (530 pg/m3 in May 1994). In a 1995
study, Foreman et al. (2000) detected atrazine in 35% of weekly composite air samples collected at the
Eagle Harbor site.
In addition to samples collected at land-based sampling stations, atmospheric particulate samples were
collected from seven sampling locations in the open lake (over-water). A total of 10 samples were
collected from the following over-water sites: 1, 5, 6, 41, 23M, 40M, and GB24M. Fourteen additional
samples were geographical composites that were collected as the R/V Lake Guardian traveled between
open-lake sampling stations. Of the 24 analyzed samples, atrazine was detected in only two samples. In
May 1994, an atrazine concentration of 560 pg/m3 was measured at station 1, and in May 1995, an
atrazine concentration of 280 pg/m3 was measured at station 5. Both of these stations are relatively close
to the shore, and measured concentrations are comparable to the nearest shoreline stations (IIT Chicago
and Indiana Dunes).
3-6
-------
Atrazine in Atmospheric Components
Figure 3-4. Sea
600 -
~ 500 -
"S
a 400 -
0 300-
= ^5»S
a<2~'OCj3 ^^r^C^CTQ-T?
(U.xoajroajroQ-ro^-^^aj.i;
U)U-ZO->U-S
-------
Table 3-3. Mean Spring/Summer Atrazine Concentrations Measured in the Participate Phase
Sampling
Location
Beaver Island
Eagle Harbor
Sleeping Bear
Dunes
Brule River
Indiana Dunes
1 IT Chicago
Muskegon
Manitowoc
Chiwaukee
Prairie
Bondville
South Haven
Spring/Summer 1994"
Nd
6
4
5
5
6
6
6
6
6
6
5
Mean
(pg/m3)
45
130
39
71
280
200
87
64
76
520
100
Range
(pg/m3)
0.0-200
0.0-530
0.0-160
0.0-290
0.0-840
0.0-630
0.0-440
0.0-340
0.0-270
0.0-1400
0.0-520
SDe
(pg/m3)
82
270
67
120
340
290
170
140
110
610
230
RSDf
(%)
180
200
170
180
120
140
200
210
140
120
220
Spring/Summer 19951
Nd
6
-
7
6
6
6
5
6
6
6
6
Mean
(pg/m3)
4.1
-
44
9.1
200
53
66
16
38
210
140
Range
(pg/m3)
0.0-25
-
1.5-120
0.0-54
0.0-1200
0.0-220
0.0-230
0.0-78
0.0-120
0.0-830
0.0-610
SDe
(pg/m3)
10
-
46
22
470
91
100
31
53
350
250
RSD
(%)
240
-
100
240
240
170
150
200
140
170
180
Spring/Summer 1994-1995°
Nd
12
4
12
11
12
12
11
12
12
12
11
Mean
(pg/m3)
25
130
42
37
240
130
78
40
57
370
120
Range
(pg/m3)
0.0-200
0.0-530
0.0-160
0.0-290
0.0-1200
0.0-630
0.0-440
0.0-340
0.0-270
0.0-1400
0.0-610
SDe
(pg/m3)
60
270
53
87
400
220
140
98
82
500
230
RSDf
(%)
240
200
130
230
160
170
180
240
140
140
190
a Samples collected from March 20, 1994 through September 23, 1994, based on the midpoint of the compositing period.
b Samples collected from March 20, 1995 through September 23, 1995, based on the midpoint of the compositing period.
0 Samples collected from March 20 through September 23, 1994 and March 20 through September 23, 1995, based on the midpoint of the compositing period.
d N = number of monthly composite samples.
e SD = standard deviation.
f RSD = relative standard deviation.
3-8
-------
Atrazine in Atmospheric Components
Figure 3-5. Mean Spring/Summer Atrazine Concentrations Measured in the Particulate Phase
Chiwaukee ri
Wisconsin __PiairjeM
3.12.3 Analysis of Breakdown Products
DBA was only detected above sample-specific detection limits in five particulate phase samples. Sample-
specific detection limits for DBA in the particulate phase ranged from 4.5 to 102 pg/m3 and averaged 25
pg/m3 for samples analyzed at the Illinois Water Survey. DBA was detected above sample-specific
detection limits in the June 1995 monthly composite sample from Bondville (67 pg/m3); the June 1994
monthly composite sample from Indiana Dunes (78 pg/m3); the May 1995 monthly composite sample
from Indiana Dunes (230 pg/m3); a 12-hour composite sample from over-water station 1 on May 10, 1994
(230 pg/m3); and a 12-hour composite sample from over-water station 5 on May 11, 1995 (100 pg/m3).
All samples that contained detectable amounts of DBA corresponded to samples containing >280 pg/m3
atrazine. In those samples with DBA above sample-specific detection limits, DEA/atrazine ratios
averaged 0.25.
DIA was not detected in any particulate phase samples analyzed. Sample-specific detection limits for
DIA in the particulate phase ranged from 4.5 to 102 pg/m3 and averaged 25 pg/m3 for samples analyzed at
the Illinois Water Survey. Samples processed at Indiana University were not analyzed for DBA or DIA.
3.1.3 Precipitation Fraction
Of 207 precipitation samples tested, atrazine was detected (i.e., measured above zero) in 124 samples
(60%), and measured atrazine concentrations were above sample-specific detection limits in 104 (50%)
samples (Figure 3-6). Method detection limits (MDLs) were calculated from results of seven spiked
samples according to the procedures specified at 40 CFR part 136, Appendix B. The MDL was then
adjusted for each sample based on the analyzed sample volume and surrogate recovery factors to obtain
sample-specific detection limits. Sample-specific detection limits ranged from 0.37 to 20 ng/L for
samples analyzed at the Illinois Water Survey, and from 0.726 to 47.6 ng/L for samples analyzed at
3-9
-------
Results of the LMMB Study: Atrazine Data Report
Indiana University. Sample-specific detection limits averaged 3.0 ng/L and 8.96 ng/L for samples
analyzed at the Illinois Water Survey and Indiana University, respectively.
3.13.7 Seasonal Variation
The presence and concentration of atrazine
in precipitation samples was strongly
influenced by season. Atrazine was
primarily detected during the spring and
summer months (Figure 3-7). All
precipitation samples collected during
April and May contained atrazine above
the sample-specific detection limits, and a
majority of samples collected in March,
June, and July contained atrazine above
sample-specific detection limits. Atrazine
was detected in less than half of
precipitation samples collected in August,
September, and October; and atrazine was
not detected in any samples collected from
November through February.
Figure 3-6. Detection Frequency of Atrazine in Precipitation
Samples
Atrazine not detected
Atrazine measure
above sample-specific
detection limit
Atrazine measured
below sample-
specific detection limit
Atrazine concentrations in precipitation peaked during both the spring of 1994 and the spring of 1995 at
all shoreline and out-of basin sampling stations (Figure 3-8). For all sampling stations except for
Sleeping Bear Dunes, peak atrazine concentrations in precipitation were higher in the spring of 1994 than
the spring of 1995. Peak concentrations were observed during the month of March at five sampling
stations (Beaver Island, Muskegon, Manitowoc, Chiwaukee Prairie, and Bondville). At four sampling
stations (Sleeping Bear Dunes, Indiana Dunes, IIT Chicago, and South Haven), peak concentrations were
Figure 3-7. Frequency Distribution of Atrazine Detection in observed in May. Peak
Precipitation Samples concentrations were observed in
April and June at Eagle Harbor and
Brule River, respectively.
II
Q> "5
0. -C
C A
m «
TO r-
s-
ft
/
s~
7
Z7I
flJ^e'
Jan
Mar May July Sept Nov
DBA and DIA concentrations
followed the same seasonal pattern as
atrazine concentrations, with
detection frequency and
concentrations peaking in the spring
and summer months. DBA was only
detected in samples collected from
March through August, and DIA was
only detected in samples collected
from April through June. Similarly
to seasonal trends for atrazine, DBA
and DIA concentrations peaked in the
early spring. Peak concentrations of
DBA and DIA at all sampling stations were higher in the spring of 1994 than the spring of 1995.
3-10
-------
Atrazine in Atmospheric Components
Figure 3-8. Seasonal Trend of Atrazine Concentrations in Precipitation
Remote Sites
400
o> 300
Beaver Island
Eagle Harbor
Sleeping Bear Dunes
Brule River
Urban and Urban-Influenced Sites
3000
2500
Indiana Dunes
NT Chicago
Muskegon
Manitowoc
Chiwaukee Prairie
iOiommmiomif)
O5O5CDO5O5O5O5O5
500
400
O)
o 300
o
o
200
'N
TO
~ 100
Rural Sites
< w O
O5O5O5O5CDO5O5O5O5O5
(DroLLS
-------
Results of the LMMB Study: Atrazine Data Report
Due to the dramatic seasonal variation in atrazine, DBA, and DIA concentrations, station averages were
calculated only for measurements during the spring and summer growing season, when atrazine and
atrazine metabolites were primarily detected in atmospheric samples. Spring/summer mean atrazine
concentrations were calculated by volume-weighted averaging of 28-day composite samples that were
collected between March 20 (the first day of spring) and September 23 (the first day of fall), based on the
midpoint of the compositing period. Station averages were calculated as volume-weighted means to
avoid bias from small precipitation events with very high measured concentrations or large precipitation
events with low measured concentrations. Volume-weighted mean concentrations were calculated as:
ci x vt
i=l
z=l
where:
ct = measured concentration in the rth sample,
vt = volume of the rth sample, and
n = number of samples.
All results were included in volume-weighted mean calculations as reported, meaning that measurements
below sample-specific detection limits were included and not censored. Results reported as zero also
were included in volume-weighted mean calculations.
Volume-weighted mean atrazine concentrations measured in precipitation during the spring/summer of
1994, 1995, and both years combined are presented in Table 3-4. For all sampling stations except for
Chiwaukee Prairie and Sleeping Bear Dunes, volume-weighted mean spring/summer atrazine levels in the
precipitation phase were higher in 1994 than 1995 (Table 3-4). Volume-weighted mean spring/summer
DBA and DIA concentrations were higher in 1994 than 1995 at all sites (Table 3-5).
3-12
-------
Table 3-4. Volume-weighted Mean Spring/Summer Atrazine Concentrations Measured in Precipitation
Sampling
Location
Beaver Island
Eagle Harbor
Sleeping Bear
Dunes
Brule River
Indiana Dunes
1 IT Chicago
Muskegon
Manitowoc
Chiwaukee
Prairie
Bondville
South Haven
Spring/Summer 1994"
Nd
7
4
6
6
7
5
7
7
7
7
7
Mean6
(ng/L)
85
19
17
31
130
53
43
40
29
120
39
Range
(ng/L)
0.0-220
0.0-100
0.84-120
0.0-160
0.0-2800
0.0-380
0.84-340
0.0-210
0.0-240
0.0-270
0.0-300
SDf
(ng/L)
120
48
50
61
1100
160
120
80
100
120
120
RSDg
(%)
120
170
160
180
240
120
170
140
150
130
140
Spring/Summer 19951
Nd
7
-
7
7
7
6
6
6
7
7
7
Mean
(ng/L)
46
-
47
19
110
8.2
8.5
7.4
34
56
12
Range
(ng/L)
0.0-210
-
1.4-300
0.0-72
4.0-320
0.0-17
0.0-29
0.79-22
0.0-160
0.0-160
0.0-28
SDf
(ng/L)
79
-
110
26
130
6.5
13
9.1
57
59
12
RSDB
(%)
170
-
190
170
130
88
120
98
160
150
110
Spring/Summer 1994-1995°
Nd
14
4
13
13
14
11
13
13
14
14
14
Mean6
(ng/L)
67
19
35
25
120
24
26
26
32
83
28
Range
(ng/L)
0.0-220
0.0-100
0.84-300
0.0-160
0.0-2800
0.0-380
0.0-340
0.0-210
0.0-240
0.0-270
0.0-300
SDf
(ng/L)
98
48
84
44
750
120
92
62
80
97
88
RSD9
(%)
140
170
180
190
280
190
210
180
160
150
190
a Samples collected from March 20, 1994 through September 23, 1994, based on the midpoint of the compositing period.
b Samples collected from March 20, 1995 through September 23, 1995, based on the midpoint of the compositing period.
0 Samples collected from March 20 through September 23, 1994 and March 20 through September 23, 1995, based on the midpoint of the compositing period.
d N = number of 28-day composite samples.
6 Mean was calculated as a volume-weighted mean by dividing the total mass of atrazine measured at a site (individual concentrations x sample volumes) by the
total precipitation volume collected at the site.
f SD = standard deviation.
g RSD = relative standard deviation.
3-13
-------
Results of the LMMB Study: Atrazine Data Report
Table 3-5. Volume-weighted Mean Spring/Summer PEA and DIA Concentrations Measured in Precipitation3
Sampling Location
Beaver Island
Eagle Harbor
Sleeping Bear
Dunes
Brule River
Indiana Dunes
1 IT Chicago
Muskegon
Manitowoc
Chiwaukee Prairie
Bondville
South Haven
Nb
14
4
4
13
14
11
13
13
14
14
14
DEA
Mean0
(ng/L)
18
9.7
13
4.7
20
2.3
20
17
12
9.2
12
Range
(ng/L)
0.0-110
0.0-19
0.0-80
0.0-21
0.0-620
0.0-85
0.0-190
0.0-220
0.0-280
0.0-100
0.0-640
SDd
(ng/L)
36
7.9
38
6.2
160
25
65
60
78
27
170
RSDe
(%)
170
73
140
140
310
270
190
230
220
210
330
DIA
Mean0
(ng/L)
3.7
1.2
0.0
0.38
3.5
0.96
2.6
0.28
1.5
0.81
1.2
Range
(ng/L)
0.0-50
0.0-6.4
0.0-0.0
0.0-3.0
0.0-150
0.0-92
0.0-32
0.0-9.3
0.0-95
0.0-18
0.0-110
SDd
(ng/L)
14
3.2
0.0
0.97
39
28
9.9
2.6
25
4.9
28
RSDe
(%)
300
200
0.0
250
350
330
250
360
360
330
370
a Samples collected from March 20 through September 23, 1994 and March 20 through September 23, 1995, based
on the midpoint of the compositing period.
b N = number of 28-day composite samples.
0 Mean was calculated as a volume-weighted mean by dividing the total mass of DEA or DIA measured at a site
(individual concentrations x sample volumes) by the total precipitation volume collected at the site.
d SD = standard deviation.
e RSD = relative standard deviation.
3.13.2 Geographical Variation
Atrazine was detected in precipitation samples from all shoreline and out-of-basin sampling stations.
Maximum atrazine concentrations at these locations ranged from 100 ng/L at Eagle Harbor to 2800 ng/L
at Indiana Dunes. The maximum concentration measured at Indiana Dunes in May 1994 is a suspected
outlier. The high concentration in this sample, which is more than six times the value for other samples,
may be due to a relatively small precipitation event (low volume) that may have coincided with emissions
from nearby fields. To avoid bias of station averages by such events, station averages were compared on
a volume-weighted basis.
Volume-weighted mean spring/summer atrazine levels for the two-year sampling campaign (1994-1995)
ranged from 19 ng/L at Eagle Harbor to 120 ng/L at Indiana Dunes (Table 3-4 and Figure 3-9).
Differences in mean atrazine levels among sites were not statistically significant due to the high
variability of measured concentrations at each station, based on the Kruskal-Wallis test (p=0.913).
While particulate atrazine levels were generally higher at shoreline sampling stations surrounding the
southern Lake Michigan basin than shoreline sampling stations surrounding the northern basin, this trend
was not observed for atrazine levels in precipitation. In fact, the volume-weighted mean spring/summer
atrazine concentrations at Beaver Island and Sleeping Bear Dunes, the two northern-most shore-line
sampling station, were higher than at any of the southern basin sampling stations except for Indiana
3-14
-------
Atrazine in Atmospheric Components
Dunes (Figure 3-9). Volume-weighted mean spring/summer atrazine concentrations at the remote out-of-
basin sampling stations (Eagle Harbor and Brule River) were comparable to concentrations at the
shoreline stations. At the agricultural Bondville station, atrazine was higher than shoreline sampling
stations, except for Indiana Dunes.
Similarly to atrazine in precipitation, DBA was detected in precipitation at all shoreline and out-of-basin
sampling stations. Maximum DBA concentrations at each site ranged from 19 ng/L at Eagle Harbor to
640 ng/L at South Haven. Volume-weighted spring/summer DBA concentrations ranged from 2.3 ng/L at
IIT Chicago to 20 ng/L at both Indiana Dunes and Muskegon (Table 3-5). Differences in mean DBA
levels among sites were not statistically significant, based on the Kruskal-Wallis test (p=0.957), and did
not represent any distinct geographical trends.
DIA was detected at all shoreline and out-of-basin sampling locations with the exception of the Sleeping
Bear Dunes site, which was not sampled for DBA or DIA during the 1995 sampling campaign. Maximum
non-zero DIA concentrations ranged from 3.0 ng/L at Brule River to 150 ng/L at Indiana Dunes.
Volume-weighted spring/summer DIA concentrations ranged from 0.0 ng/L at Sleeping Bear Dunes to
3.7 ng/L at Beaver Island (Table 3-5). Differences in mean DIA levels among sites were not statistically
significant, based on the Kruskal-Wallis test (p=0.994), and did not represent any distinct geographical
trends.
Figure 3-9. Mean Spring/Summer Atrazine Concentrations Measured in Precipitation
[^f- ^f Raawar
Beaver
Island; 1_
1 100-
£ 80 -
'I 60-
§ 40 -
20 -
0 -
Manitowoc
Lake
^1 Michigan
Chiwaukeefl
Wisconsin __PraJl!eM
'0
IIT Chicago if
Six precipitation samples were collected from the following over-water sampling stations: 1, 5, 380, 23M,
GB24M, and GB17. Atrazine was detected in only two of these samples. The precipitation sample
collected on August 20, 1994 at station 23M contained 7.5 ng/L atrazine, and the precipitation sample
collected on April 12, 1995 at station GB17 contained 29 ng/L atrazine. Atrazine was not detected in any
other precipitation samples from over-water stations, however, the remaining samples were collected in
October and November, when atrazine also was not detected at shore-line sampling stations. DBA and
DIA were not detected in precipitation samples collected from over-water sampling stations, however,
3-15
-------
Results of the LMMB Study: Atrazine Data Report
Figure 3-10. Detection Frequency of DEA in Precipitation Samples
DEA not detected
DEA measured
above sample-specific
detection limit
DEA measured
below sample-
specific detection limit
Figure 3-11. Detection Frequency of DIA in Precipitation Samples
four of these six samples were
collected in October and November,
when DEA and DIA also were not
detected at shore-line sampling
stations.
3.13.3 Analysis of Breakdown
Products
Of 193 precipitation samples tested,
DEA was detected (i.e., measured
above zero) in 76 samples (39%), and
measured DEA concentrations were
above sample-specific detection limits
in 58 (30%) samples (Figure 3-10).
DIA was only detected in 17 of 193
(8.8%) samples (Figure 3-11).
Sample-specific detection limits for
DEA and DIA ranged from 0.55 to
30 ng/L and averaged 4.5 ng/L for
samples analyzed at the Illinois
Water Survey. Samples processed at
Indiana University were not analyzed
for DEA or DIA.
DEA and DIA concentrations were
highly correlated with atrazine
concentrations in precipitation
samples (Figure 3-12). All samples
that contained detectable levels of
DEA or DIA also contained
detectable levels of atrazine, and as
atrazine concentrations in
precipitation samples increased, DEA and DIA concentrations also generally increased. Correlations
between atrazine and DEA concentrations and between atrazine and DIA concentrations were statistically
significant at the 95% confidence level. Pearson correlation coefficients for log transformed data were
0.80 and 0.84 for atrazine concentrations compared to DEA and DIA concentrations, respectively. For
samples with concentrations greater than sample-specific detection limits, DEA/atrazine ratios in
precipitation averaged 0.52. DIA/atrazine ratios averaged 0.22 among samples with concentrations
greater than sample-specific detection limits.
DIA not detected
DIA measured
above sample-specific
detection limit
DIA measured
below sample-
specific detection limit
3-16
-------
Atrazine in Atmospheric Components
Figure 3-12. Correlation of DEA and DIA Concentrations with Measured Atrazine
Concentrations in Precipitation Samples
1000 n
I!
ii
0) Q.
e E
8 8
< i
C Q.
* o
rf
-------
Results of the LMMB Study: Atrazine Data Report
The predominant atmospheric source of atrazine, DBA, and DIA measured in this study was precipitation.
Atrazine was detected in 50% of precipitation samples overall and up to 100% of precipitation samples
collected during spring and summer months for a given station. Goolsby et al. (1997) found that atrazine
was the most frequently detected herbicide in rainfall collected from the midwest and northeast in 1990-
91, with 30.2% of precipitation samples containing detectable levels of atrazine. This compares well to
the 50% of samples with detectable atrazine in this study, since this study focused on the midwest, where
Goolsby et al. cited more frequent detection of atrazine. Goolsby et al. detected DBA and DIA in 17.4%
and 2.6% of samples, respectively. This also is comparable to the 30% and 8.8% of samples collected in
the LMMB Study that contained DBA and DIA, respectively.
Atrazine concentrations as high as 2.8 i-ig/L were measured in precipitation at the Indiana Dunes sampling
station. This is similar to the maximum concentration of 3 i-ig/L measured by Goolsby et al. (1997) in
individual samples collected from the midwest in 1990-91, and the maximum concentration of 2.9 i-ig/L
measured by Capel et al. (1997) in Minnesota from 1989-1994. Nations and Hallberg (1992) measured
considerably higher maximum atrazine concentrations (up to 40 i-ig/L) in precipitation from intense
agricultural areas in Iowa. Goolsby et al. (1997) also found that volume-weighted mean concentrations of
0.2 - 0.4 i-ig/L were common in the midwest during mid-April through mid-July. These reported mean
values were slightly higher than the volume-weighted mean concentrations observed in the LMMB Study.
Volume-weighted mean atrazine concentrations measured in the LMMB Study ranged from 0.019 to 0.12
I-ig/L, however, these means were calculated for the March 20 through September 23 time period while
Goolsby et al. (1997) reported means just for a 13-week period during peak atrazine detection. Volume-
weighted mean atrazine concentrations calculated in the LMMB Study could also be lower due to the
significant number of non-detect results that were reported and factored into mean calculations.
3.2.2 Seasonality
Atrazine, DBA, and DIA concentrations in all atmospheric phases were highly seasonal. Atrazine and
atrazine metabolites were generally not detectable in winter air and precipitation samples. Atrazine
detection frequency and concentrations peaked in the late spring to early summer corresponding to the
agricultural application of atrazine during spring planting of corn and sorghum crops. Atrazine
concentrations then generally decreased to below detection by mid-August.
This finding agrees with other researchers who have identified the absence of atrazine in winter samples
(Williams et al., 1992) and the increase in atrazine detection frequency in mid-April with peaks in May or
June (Goolsby et al, 1997; Nations and Hallberg, 1992; Capel et al., 1997; Thurman and Cromwell,
2000). This highly seasonal nature of atmospheric atrazine concentrations suggests that atmospheric
loadings to Lake Michigan can be quickly affected by alterations in use patterns. The general absence of
atrazine in winter atmospheric samples indicates that atrazine is not extremely persistent in the
atmosphere nor is it appreciably transported to the atmosphere during periods of non-use.
3.2.3 Regional Considerations
Atrazine concentrations in the particulate phase appeared to be higher in areas of more intensive
agricultural land use, where atrazine usage is greater (see Figure 2-2). The highest mean concentration of
atrazine in the particulate phase was measured at the Bondville sampling station, which is in the center of
an intensive corn-growing region, where atrazine usage is greater than 163.73 pounds per square mile. In
addition, atmospheric sampling stations surrounding the southern Lake Michigan basin, where atrazine
use is highest, had higher mean particulate atrazine levels than stations surrounding the northern basin
(Figure 3-9). Mean particulate atrazine concentrations at the more northern and remote Beaver Island,
Sleeping Bear Dunes, and Brule River sites were lower than at any of the stations surrounding the
southern Lake Michigan basin and were only 7-11% of the mean particulate atrazine concentration at the
3-18
-------
Atrazine in Atmospheric Components
Bondville site. These findings indicate that while long-range (e.g., regional) transport of atrazine in the
particulate phase does occur, as demonstrated by the detection of particulate-phase atrazine at remote
sites, atrazine concentrations in the particulate phase are generally reflective of local conditions. As
suggested by Miller et al. (2000), airborne particles associated with atrazine are likely suspended by
mechanical processes such as farm machinery and are likely to be in the larger size range (3-20 [im
diameter). These larger particles are less likely to undergo long-range (e.g., regional) atmospheric
transport and are generally deposited locally. Particle size distribution analysis was not conducted to
confirm this hypothesis, but this explanation is consistent with the finding that particulate phase atrazine
concentrations are reflective of local conditions.
In contrast, concentrations of atrazine in precipitation were less reflective of local conditions. While
concentrations of atrazine in precipitation were highest at the Bondville and Indiana Dunes sites that are
located in highly agricultural areas, precipitation at remote sites also contained high concentrations of
atrazine. Concentrations of atrazine exceeding 100 ng/L were measured in individual precipitation
samples collected at all of the LMMB sampling stations including those in remote areas hundreds of
kilometers away from areas of intensive atrazine use. Volume-weighted mean atrazine concentrations in
precipitation at the remote Beaver Island and Sleeping Bear Dunes sites were higher than all other stations
except for Indiana Dunes and Bondville. This suggests that precipitation samples do not only reflect the
atrazine from local sources that is scavenged by rainfall, but also may reflect atrazine that is brought into
cloud formations and transported over long distances. Similarly, Capel et al. (1997) concluded that there
is a significant content of pesticides in rain regionally in the Midwest, and that superimposed on this
regional background is a local influence.
In addition, peak atrazine concentrations at Chiwaukee Prairie, Manitowoc, and Beaver Island in 1994
were observed in mid-March, even though corn planting had not yet begun in southern Wisconsin during
this period (Wisconsin Department of Agriculture, 1994). Nations and Hallberg (1992) also observed that
atrazine presence in precipitation began before application in the local area. Both the early appearance of
atrazine around Lake Michigan and its detection in samples at remote sites on northern Lake Michigan
and on Lake Superior (Brule River), far removed from agricultural source areas, suggest long-range (e.g.,
regional) transport. This conclusion is consistent with the findings of other researchers who have also
detected atrazine in precipitation at sites remote from cropland. Goolsby et al. (1997) found atrazine in
precipitation samples from Maine, Isle Royale in northern Lake Superior, and 4% of background sites in
the Rocky Mountains and Alaska. Majewski et al. (2000) detected atrazine in 76% of samples collected
from April through September 1995 at the remote Eagle Harbor site. Thurman and Cromwell (2000)
measured atrazine concentrations as high as 1.8 i-ig/L in precipitation at the Isle Royale National Park in
northern Lake Superior.
3.2.4 Atrazine Breakdown Products
Almost no DBA or DIA residues were detected in samples of airborne particles while detection of these
breakdown products was common in rain samples containing atrazine. This may be simply due to the
differences in methodologies used to collect, concentrate, and extract atrazine and atrazine metabolites
from the different atmospheric phases. This finding may also suggest more efficient scavenging of
atrazine metabolites by precipitation. Van Dijk and Guicherit (1999) agreed that atrazine metabolites are
more polar than their parent compound, more water soluble, and more readily removed by wet deposition
processes.
Goolsby et al. (1997) also suggested that significant degradation of atrazine could be occurring in the
atmosphere. This hypothesis was based on the finding that DBA to atrazine ratios in precipitation
samples (median of 0.5) were higher than these ratios in streams draining cornbelt states (<0.1 to 0.4).
DBA to atrazine ratios calculated in precipitation samples from the LMMB Study (median of 0.18),
3-19
-------
Results of the LMMB Study: Atrazine Data Report
however, were somewhat lower than ratios reported by Goolsby (median of 0.5), and ratios in stream
samples from the LMMB Study (median of 0.77) were higher than reported by Goolsby.
Conclusions concerning the breakdown of atrazine are difficult to make simply from measured atrazine
and DBA concentrations. Inputs, outputs, and transformations of atrazine and DBA all affect measured
concentrations. To fully assess atrazine degradation, specific studies using labeled compounds would be
required.
3.3 Quality Implementation and Assessment
As described in Section 1.5.5, the LMMB QA program prescribed minimum standards to which all
organizations collecting data were required to adhere. The quality activities implemented for the atrazine
monitoring portion of the study are further described in Section 2.6 and included use of SOPs, training of
laboratory and field personnel, and establishment of MQOs for study data. A detailed description of the
LMMB QA program is provided in The Lake Michigan Mass Balance Study Quality Assurance Report
(USEPA, 200 li). A brief summary of data quality issues for the atmospheric atrazine data is provided
below.
As discussed in Section 2.5, the sample collection, extraction, and analysis methods for atrazine
monitoring in this study are modifications of the methods used for the PCBs and chlorinated pesticides.
Pis used surrogate spikes to monitor the bias of the analytical procedure. Surrogate recoveries were
reported with the data generated at Indiana University, but were not reported with the data generated at
the Illinois Water Survey. The Illinois Water Survey PI noted that spike recoveries for atrazine were
generally 70 to 120%, similar to those obtained for PCBs. In a few cases, however, atrazine recoveries
were in the range of 10 to 25%, well below the normal recovery range of 70 to 120%. This occurrence is
documented for each affected sample in the database in the field "Exception to Method, Analytical." DBA
and DIA recoveries were usually less than 50%, and sometimes as low as 20%. As a consequence, the PI
characterized DBA and DIA results as semi-quantitative. For the Indiana University data, surrogate
correction was applied to a subset of the atrazine data. Approximately 25% of the study samples were
already extracted when the use of a surrogate was included in the analytical procedure. Sample results
that were not surrogate corrected are indicated as such in the database in the field "Exception to Method,
Analytical."
As discussed in Section 2.6, data verification was performed by comparing all field and QC sample
results produced by each PI with their MQOs and with overall LMMB Study objectives. Analytical
results were flagged when pertinent QC sample results did not meet acceptance criteria as defined by the
MQOs. These flags were not intended to suggest that data were not useable; rather they were intended to
caution the user about an aspect of the data that did not meet the predefined criteria. Table 3-6 provides a
summary of flags applied to the atmospheric atrazine data. The summary includes the flags that directly
relate to evaluation of the MQOs to illustrate some aspects of data quality, but does not include all flags
applied to the data to document sampling and analytical information, as discussed in Section 2.6. In this
data report, the summary and analysis of atrazine data represent all results with the exception of those
flagged as "invalid" by the QC coordinator in concert with the PI.
As illustrated in Table 3-6 and discussed in previous sections, atrazine was not detected in the majority of
atmospheric samples. Ninety eight percent of vapor samples analyzed at Illinois Water Survey and 86%
of vapor samples analyzed at Indiana University contained atrazine at concentrations below detection
limits and were flagged MDL (less than method detection limit) or UND (analyte not detected).
Precipitation samples most frequently contained atrazine concentrations above detection, with only 43%
of precipitation samples analyzed at Illinois Water Survey and 50% of precipitation samples analyzed at
Indiana University containing atrazine below detection and flagged MDL or UND. As discussed in
3-20
-------
Atrazine in Atmospheric Components
previous sections, the presence and concentration of atrazine was highly influenced by season with peak
concentrations in spring corresponding to agricultural application and the lowest observed concentrations
(below detection) throughout the winter months.
A significant portion of study samples for some fractions were flagged for exceeding sample holding
times. For example, 71% of vapor samples and 53% of particulate samples analyzed at Indiana
University exceeded the established holding time. Overall, only 16% of atmospheric samples exceeded
holding times because samples analyzed at the Illinois Water Survey rarely exceeded holding times.
However, the holding times for atrazine, and many other environmental pollutants, are not well-
characterized and the effects on the sample results are unknown.
To characterize contamination associated with field and analytical activities, field blanks were obtained
for precipitation and high-volume samples at a subset of monitoring stations. In each case, filters and/or
absorbent were installed in the samplers for the normal sampling period but were not exposed to
precipitation or air flow. The precipitation field blank included a water rinse of the collector surfaces to
check for contamination by dry-deposition that might have penetrated the cover and seal on the
precipitation collector. Atrazine was reported in three particulate field blanks. Laboratory blanks
analyzed at Indiana University (and associated with samples collected from Sleeping Bear Dunes from
August 1994-October 1995) showed consistent contamination. Fourteen precipitation blanks and 13
particulate blanks contained atrazine at concentrations ranging from 62.2 to 912 pg/L and 0.437 to 5.19
pg/m3. Of the 18 vapor phase laboratory blanks, 17 contained detectable atrazine at concentrations
ranging from 0.590 to 9.54 pg/m3. The source of this contamination was never identified. Project
samples associated with contaminated blanks were qualified with the FBS (Failed Blank Sample) flag and
also as likely biased high (HIB) when the sample concentration was less than five times the concentration
found in the associated blank. Only 6% of vapor samples and 7% of precipitation samples were flagged
HIB. No particulate samples required this flag. Blank contamination is an issue when evaluating vapor
sample results at Sleeping Bear Dunes that showed concentrations above the detection limit. As
discussed in Section 3.1.1, four of the six detected atrazine concentrations from Sleeping Bear Dunes
(those collected on 10/23/94, 11/4/94, 8/8/95, and 10/7/95) were flagged for contamination of
corresponding field or laboratory blanks.
3-21
-------
Table 3-6. Summary of Routine Field Sample Flags for the Analysis of Atrazine in Atmospheric Samples
Fraction
Vapor (Illinois)
Precipitation (Illinois)
Particulate (Illinois)
Vapor (Indiana)
Precipitation (Indiana)
Particulate (Indiana)
FLAGS3
Sensitivity
MDL
NA
NA
NA
86% (30)
50% (7)
87% (13)
UNO
98% (254)
43% (83)
76% (161)
NA
NA
NA
Holding Time
EHT
21% (55)
8% (15)
6% (13)
71% (25)
21% (3)
53% (8)
Contamination
FBS
0
0
0
69% (24)
29% (4)
60% (9)
FFR
0
0
0
NA
NA
NA
Precision
FDL
0
0
0
NA
NA
NA
FFD
NA
NA
NA
14% (5)
21% (3)
20% (3)
Bias
FPC
0
14% (27)
0
NA
NA
NA
FSS
NA
NA
NA
6% (2)
21% (3)
20% (3)
FMS
0
5% (9)
4% (8)
0
21% (3)
7%(1)
LOB
0
11% (21)
4% (8)
0
0
0
HIB
0
0
0
6% (2)
7%(1)
0
a The number of routine field samples flagged is provided in parentheses. The summary provides only a subset of applied flags and does not represent the full suite of flags applied to the data.
MDL= Less than method detection limit (analyte produced an instrument response but reported value is below the calculated method detection limit). Validity of reported value may be
compromised.
UND= Analyte not detected (analyte produced no instrument response above noise).
EHT= Exceeded holding time (sample or extract was held longer than the approved amount of time before analysis). Validity of reported value may be compromised.
FBS= Failed blank sample (a blank sample associated with this analysis failed the acceptance criteria). Validity of reported value may be compromised.
FFR= Failed field blank (a field blank sample associated with this analysis failed the acceptance criteria). Validity of reported value may be compromised.
FDL= Failed laboratory duplicate (a laboratory duplicate associated with this analysis failed the acceptance criteria). Validity of reported value may be compromised.
FFD= Failed field duplicate (a field duplicate associated with this analysis failed the acceptance criteria). Validity of reported value may be compromised.
FPC= Failed performance check (a laboratory performance check sample associated with this analysis failed the acceptance criteria). Validity of reported value may be compromised.
FSS= Failed surrogate (surrogate recoveries associated with this analysis failed the acceptance criteria). Validity of reported value may be compromised.
FMS= Failed matrix spike (a matrix spike associated with this analysis failed the acceptance criteria). Validity of reported value may be compromised.
LOB= Likely biased low (reported value is probably biased low as evidenced by LMS (lab matrix spike) results, SRM (standard reference material) recovery or other internal lab QC data).
Reported value is not considered invalid.
HIB= Likely biased high (reported value is probably biased high as evidenced by LMS (lab matrix spike) results, SRM (standard reference material) recovery, blank contamination, or other
internal lab QC data). Reported value is not considered invalid.
NA= Not applicable. This flag was not applied to this data set or this type of QC sample was not prepared.
3-22
-------
Atrazine in Atmospheric Components
Matrix spike samples results showed acceptable results for the majority of study samples, where 5% of
precipitation samples, 4% of particulate samples, and no vapor samples analyzed at Illinois Water Survey
failed the matrix spike MQO criteria. For samples analyzed at Indiana University, 21% of precipitation
samples, 7% of particulate samples, and no vapor samples failed the MQO criteria.
Based on the evaluation of multiple QC sample results and surrogate recoveries, the PI and the QC
coordinator assigned HIB and low bias (LOB) flags to sample results. From Illinois Water Survey, 11%
of precipitation samples and 4% of particulate samples were flagged LOB. No samples analyzed at
Indiana University were flagged LOB, but 6% of vapor samples and 7% of precipitation samples were
flagged HIB. No samples analyzed at Illinois Water Survey were flagged HIB.
As discussed in Section 1.5.5, MQOs were defined in terms of six attributes: sensitivity, precision,
accuracy, representativeness, completeness, and comparability. GLNPO derived data quality assessments
based on a subset of these attributes. For example, system precision was estimated as the mean relative
percent difference (RPD) between the results for field duplicate pairs. Similarly, analytical precision was
estimated as the mean RPD between the results for laboratory duplicate pairs. Table 3-7 provides a
summary of data quality assessments for several of these attributes for the atmospheric atrazine study
data.
System precision, estimated as the mean RPD between field duplicate results varied greatly by fraction.
For example, particulate samples analyzed at Illinois Water Survey had a mean RPD of 11% and
precipitation samples analyzed at Illinois Water Survey had a mean RPD of 115%. This may reflect the
difficulty in collecting precipitation samples, however, the mean RPD for precipitation samples analyzed
by Indiana University was 28.1%. Duplicate pair samples with a reported concentration of zero for either
one of both samples could not be used in this assessment. Because of the large number of results reported
as zero, the system precision estimate is based on only a small number of field duplicates and may not
accurately reflect the system. Analytical precision, estimated as the mean RPD between laboratory
duplicates, could only be estimated for precipitation samples analyzed at Indiana University (due to the
large number of zero results in the Illinois Water Survey data). The mean RPD between laboratory
duplicates of precipitation samples was 23%. Analytical precision was lower, but not much lower, than
the mean RPD of 28.1% for system precision. This suggests the majority of the variability associated
with the measurement system, as implemented at Indiana University, was associated with the analytical
process and less variability was associated with the sampling process. This is not uncommon for
environmental monitoring activities. Analytical precision at Indiana University also can be evaluated by
the variability in surrogate correction factors. The standard deviations of surrogate correction factors for
vapor, precipitation, and particulate samples were 0.227, 0.411, and 0.188, respectively. This suggests
that the analytical results are highly reproducible.
Evaluation of matrix spike sample (LMS) results and surrogate recoveries shows a slight low bias overall
for all phases of atmospheric samples. In most cases, the PI and QC coordinator determined that the bias
was not strong enough to warrant flagging the data as HIB or LOB. A small portion of samples generated
at Illinois Water Survey were flagged LOB, due in part to low surrogate recoveries, and a small portion of
samples generated at Indiana University were flagged HIB, due in part to contamination in laboratory
blanks.
3-23
-------
Table 3-7. Data Quality Assessment for the Analysis of Atrazine in Atmospheric Samples3
Fraction
Vapor
(Illinois)
Precipitation
(Illinois)
Particulate
(Illinois)
Vapor
(Indiana)
Precipitation
(Indiana)
Particulate
(Indiana)
Number of
Routine
Samples
Analyzed
259
193
211
35
14
15
Number of
Field
Duplicates
Analyzed
26
10
15
13
12
8
System Precision
Mean Field Duplicate RPD (%)
MDL
all results=0
115%
(3)
11%
(2)
(0)
28.1%
(5)
51.3%
(2)
Analytical Precision
Mean Lab
Duplicate RPD
(%),
-------
Chapter 4
Atrazine in Tributaries
Table 4-1. Tributary Samples Collected and
Analyzed for Atrazine, PEA, and DIA
4.1 Results
From April 4, 1995 to October 31, 1995, filtered samples were collected from 11 tributaries that flow into
Lake Michigan. A total of 108 filtered samples were collected and analyzed for atrazine, DBA, and DIA
(Table 4-1). Unlike atmospheric samples that often did not contain detectable levels of atrazine and
atrazine breakdown products, most tributary samples contained detectable levels of atrazine, DBA, and
DIA. All tributary samples except for one (collected on October 18, 1995 from the Pere Marquette River)
contained atrazine above method detection limits (MDLs). DBA was measured above the MDL in all but
10 (9.3%) tributary samples (2 from the Pere Marquette River, 2 from the Menominee River, and 6 from
the Manistique River). DIA was measured above
the MDL in all but 16 (14.8%) tributary samples (1
from the Kalamazoo River, 1 from the St. Joseph
River, 4 from the Manistique River, 5 from the
Menominee River, and 5 from the Pere Marquette
River). Method detection limits were calculated
from results of seven spiked samples according to
the procedures specified at 40 CFRpart 136,
Appendix B. The calculated MDLs were 1.25
ng/L for atrazine, 2.46 ng/L for DBA, and 8.27
ng/L for DIA.
4.1.1 Seasonal Variation
Tributary samples were only collected over a seven
month period, so seasonal trends could not be fully
evaluated. For the three tributaries with the
highest mean atrazine concentrations (the St.
Joseph River, the Kalamazoo River, and the Grand
River), peaks in atrazine concentration were
observed during mid to late May (Figure 4-1).
Peaks in atrazine concentration were much less
distinct for the remaining tributaries and occurred
in April, May, July, or August (Figure 4-1).
Seasonal patterns of DBA and DIA concentrations
were very similar to those for atrazine. DBA
concentrations in the St. Joseph and Grand Rivers
peaked in mid to late May, and DBA
concentrations in the Kalamzaoo River peaked in
April. DIA concentrations in the St. Joseph and
Kalamazoo Rivers peaked in mid to late May, and DIA concentrations in the Grand River peaked in
April. For other tributaries, peaks in DBA and DIA concentrations were less defined and occurred in
May, June, August, September, or October.
To statistically evaluate seasonal trends, monthly log-transformed atrazine, DBA, and DIA concentrations
were compared using a two-way ANOVA model that considered station variability and monthly
variability. Atrazine concentrations were significantly higher in June than in most other months
Tributary
Fox
Grand
Calumet
Grand
Kalamazoo
Manistique
Menominee
Milwaukee
Muskegon
Pere
Marquette
Sheboygan
St. Joseph
Sampling Dates
4/5/95-10/12/95
4/20/95-10/18/95
4/10/95-10/31/95
4/11/95-10/30/95
4/11/95-10/26/95
4/11/95-10/11/95
4/4/95-10/6/95
4/13/95-10/17/95
4/14/95-10/18/95
4/6/95-10/24/95
4/12/95-10/27/95
Total
Number of
Samples
13
5
14
11
7
6
16
7
7
11
11
108
4-1
-------
Results of the LMMB Study: Atrazine Data Report
Figure 4-1. Seasonal Trend of Atrazine Concentrations in Lake Michigan Tributaries
3000
kSt. Joseph
Kalamazoo
Grand
120
-. 100
O)
u
Ǥ
0)
c
'N
Fox
Milwaukee
-Grand Calumet
-Sheboygan
O)
u
o
o
0)
c
'N
2
Muskegon
Menominee
Pere Marquette
Manistique
4-2
-------
Atrazine in Tributaries
(April, July, August, and October), and atrazine concentrations were significantly lower in October than
in most other months (April, May, June, and August). This finding of higher tributary atrazine
concentrations in June and lower tributary atrazine concentrations in October (the last month sampled in
the study) is consistent with application of the herbicide in the spring and degradation of the herbicide
throughout the summer. A similar pattern was seen for DBA, with DBA concentrations significantly
higher in June than in most other months (April, May, July, and October), and DBA concentrations
significantly lower in October than in most other months (April, May, June, and August). DIA
concentrations in June were significantly higher than concentrations in two other months (October and
July), and DIA concentrations in October were significantly lower than concentrations in two other
months (June and August). There was no significant interaction between month and station for any of the
three analytes (atrazine: p = 0.975; DBA: p = 0.973; DIA: p = 0.568).
In addition to season, tributary atrazine concentrations are potentially affected by tributary flows. Higher
flows associated with runoff events may increase atrazine concentrations as atrazine is flushed from the
watershed. Conversely, during longer duration high flow events, atrazine concentrations could decrease
as a result of dilution. The correlations between tributary atrazine concentrations and flow are presented
in Table 4-2. In only one tributary (the Manistique River) was atrazine concentration significantly
correlated with flow, and in this tributary, the correlation was negative (i.e., atrazine decreasing with
increasing flow). Associations between flow and atrazine are difficult to make with such small data sets
(n of 5 to 16), but in general, atrazine peaks appear to be much more influenced by the timing of atrazine
application than tributary flows. Figure 4-2 shows the flow patterns and atrazine concentrations for the
three tributaries with the highest atrazine levels. Peaks in atrazine occurred in the spring after atrazine
application in the watershed. These peaks were often associated with small peaks in flow during the
spring but were relatively independent of the magnitude of flow peaks. As Richards and Baker (1993)
noted for Lake Erie tributaries, the annual pattern of atrazine concentrations is one of storm event peaks
modified by an annual pattern of availability of the pesticide in late spring.
Table 4-2. Correlation of Tributary Atrazine Levels with Tributary Flow
Tributary
Manistique
Muskegon
Milwaukee
Grand Calumet
St. Joseph
Grand
Fox
Menominee
Kalamazoo
Sheboygan
Pere Marquette
N
7
7
16
5
11
14
13
6
11
11
7
Correlation Coefficient
-0.76
-0.52
-0.33
0.019
0.060
0.27
0.33
0.47
0.53
0.58
0.73
i2
0.57
0.27
0.11
0.00036
0.0036
0.073
0.11
0.22
0.28
0.34
0.53
p-value
0.048
0.230
0.210
0.976
0.862
0.350
0.274
0.347
0.091
0.062
0.063
4-3
-------
Results of the LMMB Study: Atrazine Data Report
Figure 4-2. Seasonal Flow Patterns and Atrazine Concentrations in Selected Lake Michigan
Tributaries
10000
10000
o
Jan-95 Feb-95 Apr-95 May-95 Jul-95 Sep-95 Oct-95
6000
1000
1
Jan-95 Feb-95 Apr-95 May-95 Jul-95 Sep-95 Oct-95
t 1000
100
1
o
c
01
o
o
O
Jan-95 Feb-95 Apr-95 May-95 Jul-95 Sep-95 Oct-95
4-4
-------
Atrazine in Tributaries
4.1.2 Geographical Variation
Atrazine concentrations measured in tributaries ranged nearly 4 orders of magnitude, from 0.50 ng/L in
the Pere Marquette River to 2700 ng/L in the St. Joseph River (Figure 4-3). With the exception of one
sample from the St. Joseph River that measured 2700 ng/L of atrazine, all other tributary samples
contained 550 ng/L atrazine or less. Eighty-six percent of tributary samples contained less than 100 ng/L
of atrazine, and all samples above 100 ng/L of atrazine were from the St. Joseph, Kalamazoo, or Grand
Rivers.
Figure 4-3. Atrazine Concentrations Measured in Lake Michigan Tributaries3
O)
c
O
"co
CD
o
c
o
O
10000g
1000E
100E
10=
IE
A
ABC
T
-n
8
"5"
M
n
§
c
m
H
II
I
CD
7J
>
z
D
^
II
^
7;
5
1AZOO (n
g
z
en
TIQUE (n
i
S
m
z
z
m
m
g
F
^
m
m
g
c
en
S
II
-D
>
3]
QUETTE
i
en
I
m
n
§
1
3
CO
H
b
en
m
-D
I
3*
M
aBoxes represent the 25th (box bottom), 50th (center line), and 75th (box top) percentile results. Bars represent the results nearest 1.5 times
the inter-quartile range (IQR=75th-25th percentile) away from the nearest edge of the box. Circles represent results beyond 1.5*IQR from the
box. Xs represent results beyond 3*IQR from the box. Letters above the boxes represent results of analysis of variance and multiple
comparisons test. Tributaries with the same letter were not statistically different (at alpha = 0.05).
Mean atrazine concentrations during the 1995 sampling campaign ranged from 3.7 ng/L in the Manistique
River to 350 ng/L in the St. Joseph River (Table 4-3 and Figure 4-4). The St. Joseph, Grand, and
Kalamazoo Rivers had the highest mean atrazine concentrations (103 - 350 ng/L); and the Menominee,
Manistique, and Pere Marquette Rivers had the lowest atrazine concentrations (3.7 - 5.3 ng/L). Those
tributaries with the highest mean atrazine levels are located along the shore of the southern Lake
Michigan basin, where agricultural influences are greatest. Those tributaries with the lowest mean
atrazine levels are located in the northern portions of the lake, where land use is less dominated by
agriculture (Figure 4-4).
4-5
-------
Results of the LMMB Study: Atrazine Data Report
Table 4-3. Mean Atrazine Concentrations Measured in Lake Michigan Tributaries
Tributary
Fox
Grand Calumet
Grand
Kalamazoo
Manistique
Menominee
Milwaukee
Muskegon
Pere Marquette
Sheboygan
St. Joseph
Na
13
5
14
11
7
6
16
7
7
11
11
Mean
(ng/L)
59
47
150
103
3.7
5.3
30
17
4.1
28
350
Range
(ng/L)
41-81
34-57
36-550
8.5-440
1.6-6.4
1.6-12
11-58
8.9-22
0.50-8.7
12-48
35-2700
SDb
(ng/L)
13
9.7
140
140
1.6
4.3
15
4.4
3.3
10
780
RSDC
(%)
22
21
96
140
44
80
51
25
82
36
220
a N = number of samples.
b SD = standard deviation.
c RSD = relative standard deviation.
Two-way analysis of variance controlling for month performed on log transformed tributary atrazine
concentrations revealed that mean atrazine concentrations differed significantly (at the 95% confidence
level) among tributaries. Results of Tukey's Studentized Range Test for multiple comparisons showed
that mean atrazine concentrations in the St. Joseph and Grand Rivers were significantly higher (at the
95% confidence level) than mean concentrations in the Milwaukee, Sheboygan, Muskegon, Menominee,
Manistique, and Pere Marquette Rivers and St. Joseph was greater than Kalamazoo (Figure 4-3). The
mean atrazine concentration in the Fox River was significantly higher (at the 95% confidence level) than
mean concentrations in the Muskegon, Menominee, Manistique, and Pere Marquette Rivers. These later
three tributaries (the Menominee River, the Manistique River, and the Pere Marquette River) exhibited
significantly lower (at the 95% confidence level) mean atrazine concentrations than any other site. Mean
atrazine concentrations in these three rivers were from 3.3 to 94 times lower than mean atrazine
concentrations in any other tributary.
DBA concentrations in tributary samples ranged from 1.1 ng/L in the Manistique River to 220 ng/L in the
St. Joseph River. Mean DBA concentrations ranged from 1.7 ng/L to 55 ng/L in these same tributaries
(Table 4-4). Similarly to atrazine concentrations, DBA concentrations were significantly lower (at the
95% confidence level) in the Menominee, Manistique, and Pere Marquette Rivers than in all other
tributaries (Figure 4-5). Mean DBA concentrations were from 3.4 to 33 times lower in these three
tributaries than in any other tributary. DBA concentrations in the Muskegon River were also significantly
lower (at the 95% confidence level) than in the remaining tributaries (Figure 4-5).
DIA concentrations ranged from below detection to as high as 130 ng/L in the St. Joseph River. Mean
DIA concentrations ranged from 6.1 ng/L in the Pere Marquette River to 45 ng/L in the Grand Calumet
River (Table 4-4). DIA concentrations also were lower in the Manistique, Menominee, and Pere
Marquette Rivers than in all other tributaries, however, the difference between the Pere Marquette River
and the St. Joseph, Fox, and Kalamazoo were not statistically significant (at the 95% confidence level)
4-6
-------
Atrazine in Tributaries
(Figure 4-5). Mean DIA concentrations in the Manistique, Menominee, and Pere Marquette Rivers were
from 3.1 to 7.4 times lower than in other tributaries.
Figure 4-4. Mean Atrazine Concentrations Measured in Lake Michigan
Tributaries
Pere Marquette
Muskegon
4-7
-------
Results of the LMMB Study: Atrazine Data Report
Table 4-4. Mean PEA and DIA Concentrations Measured in Lake Michigan Tributaries
Tributary
Fox
Grand Calumet
Grand
Kalamazoo
Manistique
Menominee
Milwaukee
Muskegon
Pere Marquette
Sheboygan
St. Joseph
Na
13
5
14
11
7
6
16
7
7
11
11
DEA
Mean
(ng/L)
47
31
49
28
1.7
4.0
29
14
4.0
32
55
Range
(ng/L)
37-62
24-39
26-88
13-50
1.1-2.6
1.6-6.9
17-60
6.0-23
1.8-8.3
28-44
23-220
SDb
(ng/L)
6.5
6.5
19
13
0.62
2.2
10
5.8
2.2
4.4
55
RSDC
(%)
14
21
40
46
37
56
34
43
56
14
99
DIA
Mean
(ng/L)
23
45
41
20
6.5
6.3
28
40
6.1
38
35
Range
(ng/L)
14-34
27-94
26-60
7.9-37
0.0-23
0.0-23
15-56
17-84
0.0-19
25-47
3.0-130
SDb
(ng/L)
6.0
27
11
10
8.3
8.5
11
23
7.0
7.2
35
RSDC
(%)
27
60
27
51
130
140
39
56
110
19
100
a N = number of samples.
b SD = standard deviation.
c RSD = relative standard deviation.
4.1.3 Analysis of Breakdown Products
DEA and DIA concentrations were highly correlated with atrazine concentrations in tributary samples.
As atrazine concentrations increased, DEA and DIA concentrations also generally increased (Figure 4-6).
Correlations between atrazine and DEA concentrations and between atrazine and DIA concentrations
were statistically significant (p< 0.0001). Pearson correlation coefficients for log transformed data were
0.88 and 0.66 for atrazine concentrations compared to DEA and DIA concentrations, respectively.
The ratio of DEA to atrazine in surface water samples has been used to evaluate the degradation of
atrazine in receiving systems. The median DEA/atrazine ratio calculated for tributary samples in this
study was 0.77. Ratios ranged from 0.08 to 3.7. Two-way analysis of variance was performed on log-
transformed DEA/atrazine ratios to evaluate the effect of month and tributary. This analysis revealed that
DEA/atrazine ratios differed significantly among tributaries and differed significantly among months.
The interaction of month and tributary did not account for a significant portion of the variance (p=0.777).
Mean DEA/atrazine ratios were the highest in the Pere Marquette, Sheboygan, and Milwaukee Rivers.
Mean ratios were above 1.0 at each of these sites and were significantly greater than DEA/atrazine ratios
in the Kalamazoo, Manistique, Grand, and St. Joseph Rivers. Mean DEA/atrazine ratios were 0.76 in the
Kalamazoo River and below 0.5 in the Manistique, Grand, and St. Joseph Rivers. Higher DEA/atrazine
ratios, especially those exceeding 1.0, indicate significant degradation of atrazine within the watershed or
the tributary itself. Significant differences between DEA/atrazine ratios among tributaries suggest that
either average residence times of atrazine within the watershed differ among tributaries, or degradation
rates in those watersheds differ.
4-8
-------
Atrazine in Tributaries
Figure 4-5. DEA and DIA Concentrations Measured in Lake Michigan Tributaries3
1000q
DEA
AB
O
O)
c
O
"ro
§
o
O
10=
A
o
r^, ABC
PI
L=J
i i
n n
° g
1 i
-y m
"3"
II
^J1
A
T
i i i
-GRAND(n=14)
p
~r
I
i
I
3"
II
p
-MANISTIQUE (n=7)
BC
I . p
Lnr-' -j-
E
nr
i i i
s s s
m p c
z < en
8 1 m
1 i n
m m 0
m m ^.
1 1 ?
^^ n
ABC r^
I
ds
^
E
T
_L
-SHEBOYGAN(n=11
-P MARQUETTE (n=7
p
r
0
H
0
n
D
:
'
^
^
O)
o
"ro
I
o
O
140n
120J
100J
80J
60J
40J
20=
0=
-20
DIA
AB
CD
ABC
T
8
n
e
n
D
en
D
c
m
z
m
m
|
c
en
^
en
m
m
Tl
I
"Boxes represent the 25th (box bottom), 50th (center line), and 75th (box top) percentile results. Bars represent the results nearest 1.5 times
the inter-quartile range (IQR=75th-25th percentile) away from the nearest edge of the box. Circles represent results beyond 1.5*IQR from the
box. Xs represent results beyond 3*IQR from the box. Letters above the boxes represent results of analysis of variance and multiple
comparisons test. Tributaries with the same letter were not statistically different (at alpha = 0.05).
4-9
-------
Results of the LMMB Study: Atrazine Data Report
Figure 4-6. Correlation of DEA and DIA Concentrations with Measured Atrazine Concentrations in
Tributary Samples
re
+j
o IT
It
o
C 0.
o S
1000
100
10
<
Q
C
-------
Atrazine in Tributaries
Other studies that have focused on smaller, more agriculturally influenced streams or that have focused on
measuring atrazine concentrations in storm events following herbicide application, have found much
higher concentrations of atrazine than measured in this study. Kolpin and Kalkhoff (1993) measured
atrazine concentrations as high as 8940 ng/L in small Iowa streams, and Stamer et al. (1994) estimated
atrazine levels of 19,000 to 62,000 ng/L in Nebraska streams. In a 1989-1990 survey of mid-western
streams, Thurman et al. (1992) measured a maximum concentration of 108,000 ng/L and a median
concentration of 3800 ng/L during post-planting. Similarly, in a 1998 survey of mid-western streams,
Battaglin et al. (2000) measured a maximum concentration of 224,000 ng/L and a median of 3970 ng/L
shortly after herbicide application. As previously mentioned, several of these studies focused on the
collection of samples in storm events following herbicide application. As Thurman et al. (1992) noted,
median atrazine levels in mid-western streams dropped to 230 ng/L during pre-planting and harvest time
periods. Atrazine concentrations also are more moderated in large river systems than in small streams
that directly drain agricultural fields. For instance, Pereira and Hostettler (1993) measured atrazine
concentrations ranging from 54 to 4735 ng/L in the Mississippi River and its major tributaries, and Clark
et al. (1999) measured a median concentration of approximately 500 ng/L in the Mississippi River.
4.2.2 Comparison to Regulatory Limits
Measured atrazine concentrations in Lake Michigan tributaries were below regulatory limits for human
health concerns and proposed ambient water quality criteria. The highest mean atrazine concentration in
a tributary (350 ng/L) was approximately 10 times below the maximum contaminant level for drinking
water (MCL) of 3000 ng/L. This contrasts with a 1989 survey of mid western streams, which found that
55% of streams exceeded the MCL during post-planting (Thurman et al., 1992).
When compared to the proposed ambient water quality criterion maximum concentration of 350 i-ig/L,
maximum atrazine concentration measured in this study were more than 100 times less than the criterion.
Mean atrazine concentrations were more than 34 times less than the proposed ambient water quality
criterion continuous concentration of 12 i-ig/L.
4.2.3 Seasonality
For the three tributaries with the highest atrazine levels (St. Joseph River, Kalamazoo River, and Grand
River), distinct peaks in atrazine concentrations were observed in mid to late May. Distinct peaks in
atrazine concentrations were not observed in other tributaries, however, it should be noted that measured
tributary concentrations reflect not only atrazine levels but tributary flows as well. While tributary
concentrations may be relatively constant, atrazine loads may peak in the late spring due to higher
tributary flows. The modeling phase of the LMMB Study will combine atrazine concentration data with
tributary flow data to estimate atrazine loads and the seasonal timing of these loads.
Most researchers have noted that the bulk of total atrazine loads in tributaries occurs during the early
summer just after atrazine application and in correlation with major rain events (Muller et al., 1997;
Schottler and Eisenreich, 1997; Thurman etal, 1992; Lemieux etal, 1995; Clark etal, 1999). Richards
and Baker (1993) noted that the annual pattern of atrazine concentrations in Lake Erie tributaries was one
of storm event peaks modified by an annual pattern of availability of the pesticide in late spring.
Schottler and Eisenreich (1997) estimated the monthly contribution to annual atrazine loads to be 50% in
June, 20% in May, 15% in July, 5% in April, 3% in August, and 1% in all other months.
4.2.4 Regional Considerations
Concentrations of atrazine and atrazine metabolites in tributaries were strongly influenced by
geographical location and regional land use patterns. Atrazine, DEA, and DIA concentrations were
4-11
-------
Results of the LMMB Study: Atrazine Data Report
significantly lower in the Menominee River, the Manistique River, and the Pere Marquette River than any
other tributaries. This finding is consistent with land use and atrazine use patterns in these watersheds.
These watersheds generally consist of less cropland and are more forested (Figure 4-7). In fact, EPA has
listed the riparian habitat for the Manistique and Menominee Rivers as greater than 75% forested (Table
2-1). The lower percentage of cropland in the Manistique, Menominee, and Pere Marquette River
watersheds also translates into lower atrazine use rates for these watersheds (Figure 4-8) and lower
tributary atrazine levels (Figure 4-4). These rivers flow through counties with some of the lowest
estimated atrazine use rates in the Lake Michigan region (estimated below 2.32 pounds per square mile in
1987-1989). The Muskegon River, which had the fourth lowest mean tributary atrazine concentration
flows through counties with estimated atrazine use rates of less than 52.35 pounds per square mile. The
remaining tributaries all flow through counties with estimated atrazine use rates of greater than 52.35
pounds per square mile. The St. Joseph River, which had the highest measured atrazine concentrations,
flows through Indiana counties with estimated atrazine use rates of greater than 163.73 pounds per square
mile. In conclusion, tributary atrazine concentrations are reflective of land use and atrazine use rates
within the watershed. In a study of Nebraska streams, Stamer et al. (1994) similarly found that the largest
yields and mean concentrations of atrazine in surface water were associated from drainage areas with the
highest percentage of cropland, and the smallest was associated with the smallest amount of cropland.
Richards and Baker (1993) also found that atrazine peak concentrations were strongly affected by land
use and by soil type. Capel and Larson (2001) added that the load of atrazine to receiving streams as a
Figure 4-7. Mean Tributary Atrazine Concentrations (in parenthesis) and Land Use Patterns in the Lake
Michigan Watershed
Muskegon .t^
(17 ng/L)
Grand
Kalamazoo
(103 ng/L)
Urban and Built-Up Land
Cropland and Pasture
Cropland/Grassland Mosaic
Cropland/Woodland Mosaic
Mixed Forest
Evergreen Needle leaf Forest
Water Bodies
-------
Atrazine in Tributaries
percentage of atrazine use within Figure 4-8. Mean Tributary Atrazine Concentrations (in parenthesis)
the basin was relatively constant and Atrazine Use in the Lake Michigan Watershed Estimated for 1987
regardless of the scale of the to 1989 (graphic modified from Battaglin, U.S. Geological Survey,
watershed. Lastly, seasonal and 1994)
annual weather patterns and the _^f r
timing of precipitation events in
relation to agricultural activities are
important in influencing tributary
concentrations and loads of atrazine
(Richards et al, 1996).
4.2.5 Atrazine Breakdown
Products
Ratios of DBA to atrazine in
tributary samples measured in this
study indicated significant
degradation of atrazine.
DEA/atrazine ratios averaged as
high as 1.6 for some tributaries
(Pere Marquette River). This is
higher than mean DEA/atrazine
ratios reported by Thurman et al.
(1992) for mid-western tributaries.
Ratios of DEA to atrazine also
increased over the summer months
(from 0.63 in May to 1.4 in
October) as spring-applied atrazine
degraded within the watershed.
Thurman et al (1992)
also reported an increase in
DEA/atrazine ratios from <0.1
shortly after application to 0.4 later
in the year. Thurman et al. (1992)
hypothesized that increases in
DEA/atrazine ratios later in the
season were due to an influx of
groundwater, which contains more
degraded atrazine metabolites,
during stream base flow conditions. Schottler et al. (1994) also found that late in the season, subsurface
transport of atrazine and atrazine metabolites was predominant and led to increased DEA/atrazine ratios.
4.3 Quality Implementation and Assessment
As described in Section 1.5.5, the LMMB QA program prescribed minimum standards to which all
organizations collecting data were required to adhere. The quality activities implemented for the atrazine
monitoring portion of the study are further described in Section 2.6 and included use of SOPs, training of
laboratory and field personnel, and establishment of MQOs for study data. A detailed description of the
LMMB QA program is provided in The Lake Michigan Mass Balance Study Quality Assurance Report
(USEPA, 2001i). A brief summary of data quality issues for the tributary atrazine data is provided below.
Menominee
(5.3 ng/L)
Atrazine Use in Pounds per Square Mile
Missing or 0
0.01 -2.32
2.33-13.46
13.47-52.35
52.36-163.73
>163.73
4-13
-------
Results of the LMMB Study: Atrazine Data Report
Because the open-lake and tributary monitoring were conducted by the same PI, a separate QAPP was not
prepared for the tributary monitoring, and GLNPO and the PI agreed to implement the procedures
outlined in the open-lake QAPP for the tributary sampling and analysis. Field or trip blanks were not
collected for tributary atrazine, however, a wide variety of blanks were prepared and analyzed for the
open-lake water column data sets. As mentioned above, a single PI was responsible for collecting and
analyzing the open-lake and tributary samples and these samples were extracted and analyzed together.
Therefore, the results of the blank samples for the open-lake data may provide some information
regarding system and analytical contamination (see Section 5.3). Six laboratory blanks were extracted
and analyzed with the open-lake and tributary samples and are reported with both data sets. Results for
atrazine for all of the six laboratory blanks were reported as zero.
As discussed in Section 2.5, the sample collection, extraction, and analysis methods for atrazine
monitoring in this study are modifications of the methods used for the PCBs and chlorinated pesticides.
In addition, during an audit of analytical procedures, two changes to the analytical method described in
the approved QAPP were observed: addition of sodium sulfate to the extract in the centrifuge tube to
remove interfering water, and use of d5-ethylatrazine rather than d10-anthracene as the internal standard.
These changes improved analyte recovery and quantitation and were included in the analytical SOP
(USEPA, 1997a; USEPA, 1997b).
As discussed in Section 2.6, data verification was performed by comparing all field and QC sample
results produced by each PI with their MQOs and with overall LMMB Study objectives. Analytical
results were flagged when pertinent QC sample results did not meet acceptance criteria as defined by the
MQOs. These flags were not intended to suggest that data were not useable; rather they were intended to
caution the user about an aspect of the data that did not meet the predefined criteria. Table 4-5 provides a
summary of flags applied to the tributary atrazine data. The summary provided below includes the flags
that directly relate to evaluation of the MQOs to illustrate some aspects of data quality, but does not
include all flags applied to the data to document sampling and analytical information, as discussed in
Section 2.6. In this data report, the summary and analysis of atrazine data represent all results with the
exception of those flagged as "invalid" by the QC coordinator in concert with the PI.
All of the tributary samples contained detectable concentrations of atrazine so none of the tributary
samples were flagged as "analyte not detected" (UND). Eighteen percent of tributary samples were
flagged for exceeding the sample holding time of 30 days to extraction. However, the holding times for
atrazine, and many other environmental pollutants, are not well-characterized and the effects on the
sample results are unknown. All of the results for the field duplicate pairs were within acceptance criteria
(i.e., none of the sample results were flagged as failed field duplicate), and only 8% of tributary sample
results showed surrogate recoveries outside acceptance criteria. Based on the evaluation of multiple QC
sample results and surrogate recoveries, the PI and the QC coordinator did not assign high bias (HIB) or
low bias (LOB) flags to any sample results.
Table 4-5. Summary of Routine Field Sample Flags for the Analysis of Atrazine in Tributary Samples
Flag3
UND, Analyte not detected
EHT, Exceeded holding time
FSS, Failed surrogate
FFD, Failed field duplicate
Percentage of Samples Flagged (%)b
0
18% (19)
8% (9)
0
a The summary provides only a subset of applied flags and does not represent the full suite of flags applied to the data.
b The number of routine field samples flagged is provided in parentheses.
4-14
-------
Atrazine in Tributaries
As discussed in Section 1.5.5, MQOs were defined in terms of six attributes: sensitivity, precision,
accuracy, representativeness, completeness, and comparability. GLNPO derived data quality assessments
based on four of these attributes. For example, system precision was estimated as the mean relative
percent difference (RPD) between the results for field duplicate pairs. Similarly, analytical precision was
estimated as the mean relative percent difference (RPD) between the results for laboratory duplicate pairs.
Table 4-6 provides a summary of data quality assessments for several of these attributes for the tributary
atrazine study data.
System precision, estimated as the mean RPD between field duplicate results, was 9%. Analytical
precision, estimated as the standard deviation of surrogate correction factors, was 0.3. These measures
suggest that analytical results are highly reproducible. Analytical bias, estimated as the mean surrogate
recovery for the study samples, was 87% suggesting an overall slight low bias.
Table 4-6. Data Quality Assessment for the Analysis of Atrazine in Tributary Samples
Parameter3
Number of Routine Samples Analyzed
Number of Field Duplicates Analyzed
System Precision, Mean Field Duplicate RPD (%), >MDL
Analytical Precision, SCF Variability (SD)
Analytical Bias, Mean Surrogate Recovery (%)
Analytical Sensitivity , Samples Reported as < MDL (%)
Assessment13
108
5
8.5 %
(5)
0.25
(108)
87%
(108)
0 .9 %
a RPD = Relative percent difference.
MDL = Method detection limit.
SCF = Surrogate correction factor.
b Number of sample/duplicate pairs used in the assessment is provided in parentheses.
4-15
-------
Chapter 5
Atrazine in the Open-Lake Water Column
5.1 Results
Open-lake water column samples were collected during six cruises of the R/VLake Guardian (Table 5-1).
These cruises were conducted from April 25, 1994 to April 17, 1995. Open-lake samples were collected
from 35 sampling locations on Lake Michigan, 2 sampling locations in Green Bay, and 1 sampling
location on Lake Huron. Samples were collected at depths ranging from 1 to 257 m.
Table 5-1. Summary of Open-lake Samples Collected
Cruise
1
2
3
4
5
6
Sampling Dates
4/25/94-5/11/94
6/17/94-6/25/94
8/4/94 - 8/26/94
10/18/94-10/30/94
1/17/95
3/23/95-4/17/95
Total
Number of
Sites
Sampled
25
8
31
3
1
26
38
Total
Number of
Samples
54
25
64
10
2
79
234
A total of 234 samples were collected and
analyzed for atrazine, DBA, and DIA. All
open-lake samples contained levels of
atrazine and DBA above method detection
limits (MDLs), and all but 12 samples
(5.1%) contained DIA above the MDL.
MDLs were calculated from results of
seven spiked samples according to the
procedures specified at 40 CFRpart 136,
Appendix B. The calculated MDLs were
1.25 ng/L for atrazine, 2.46 ng/L for
DBA, and 8.27 ng/L for DIA.
5.1.1 Geographical Variation
Mean atrazine concentrations were
calculated for each individual open-lake
sampling station. Analysis of variance
revealed that mean atrazine levels differed
significantly among stations (at the 95%
confidence level), however, these differences were primarily due to the inclusion of sampling stations
located on Lake Huron and Green Bay. Atrazine concentrations at the Lake Huron sampling station
(LH54M) were significantly lower than atrazine concentrations at all but three Lake Michigan sampling
stations (MB63, MB72M, and 280), based on Tukey pairwise comparisons. Atrazine concentrations in
southern Green Bay (station GB17) were significantly higher than atrazine concentrations at 18 Lake
Michigan sampling stations.
Within Lake Michigan, atrazine concentrations were relatively consistent. Individual sample results
ranged from 22.0 to 58.0 ng/L, and sampling station mean atrazine concentrations only ranged from 33.0
to 48.0 ng/L. Analysis of variance and Tukey's Studentized Range Test for multiple comparisons
revealed that the only statistically significant difference in atrazine concentrations was between site 52
and three other sites. Atrazine concentrations at site 52 in northern Lake Michigan (near Beaver Island)
were significantly greater than atrazine concentrations at sites 18M (located in south central Lake
Michigan), 23M (located in south central Lake Michigan), MB72M (located near the Straits of
Mackinac), and 280 (located in central Lake Michigan). While this difference may be statistically
significant, it may be of little environmental significance. The mean atrazine concentration at site 52 was
calculated from only two samples (42.0 and 54.0 ng/L atrazine).
Similar patterns of consistency among sampling stations were observed for DBA and DIA concentrations.
The only statistically significant differences in DBA or DIA concentrations were between stations in
5-1
-------
Results of the LMMB Study: Atrazine Data Report
Green Bay and stations in Lake Michigan. Sampling station GB17 in Green Bay contained significantly
higher DBA concentrations than three Lake Michigan sites (18M, 23M, and 280). Sampling stations
GB17 and GB24M in Green Bay contained significantly higher DIA concentrations than five Lake
Michigan sites (18M, 23M, 41, 3, and 5). Within Lake Michigan proper, no statistically significant
differences were observed in DBA or DIA concentrations among sites. Individual DBA concentrations in
Lake Michigan ranged from 14.0 to 36.0 ng/L, and sampling station means ranged from 18.5 to 30.3
ng/L. Individual DIA concentrations in Lake Michigan ranged from 0.00 to 30.3 ng/L, and sampling
station means ranged from 9.00 to 20.6 ng/L.
Due to the geographical consistency of atrazine, DBA, and DIA concentrations within Lake Michigan,
lake-wide mean concentrations can be calculated to reliably represent the lake. Lake-wide mean
concentrations calculated for the study were 38.1, 25.8, and 14.9 ng/L for atrazine, DBA, and DIA,
respectively. These concentrations represent average values for the entire April 1994 through April 1995
LMMB sampling campaign, however, it was found that average lake-wide concentrations differed
significantly overtime during the campaign (see Section 5.1.2 below).
5.1.2 Seasonal Variation
Seasonal evaluations of open-lake atrazine concentrations are limited by the timing and intensity of
sampling conducted in the LMMB Study. Open-lake atrazine concentrations were measured during six
sampling cruises, however, limited data was collected during two of these cruises. Only ten samples from
three sites were collected during the fourth cruise in October 1994, and only two samples from one site
were collected during the fifth cruise in January 1995. Due to the limited number of samples and sites
evaluated during the fourth and fifth cruises, data from these cruises were not deemed to be representative
of the entire lake and were not used to evaluate seasonal trends. Sufficient data were present from the
remaining four cruises (April/May 1994, June 1994, August 1994, and March/April 1995) to adequately
compare mean Lake Michigan atrazine concentrations (Figure 5-1) overtime.
An analysis of variance was conducted to test for statistical differences in mean atrazine concentrations
measured during cruises one, two, three, and six. Statistically significant differences among cruises were
identified at the 95% confidence level. Tukey's Studentized Range Test for multiple comparisons
showed that mean atrazine concentrations measured during cruises three and six were significantly higher
than atrazine concentrations measured during cruises one and two (Figure 5-1). Mean atrazine
concentrations during cruises one (April/May 1994) and two (June 1994) were 37.0 and 36.0 ng/L,
respectively; and mean atrazine concentrations during cruises three (August 1994) and six (March/April
1995) were 39.2 and 39.7 ng/L, respectively. Mean atrazine concentrations during cruises four and five,
which were not included in this analysis, were 35.8 and 39.3 ng/L, respectively.
Similar patterns of increasing concentrations also were observed for DBA and DIA. Mean DBA
concentrations measured during cruises three and six were significantly higher than DBA concentrations
measured during cruises one and two (Figure 5-2). Mean DBA concentrations during cruises one
(April/May 1994) and two (June 1994) were 24.1 and 22.2 ng/L, respectively; and mean DBA
concentrations during cruises three (August 1994) and six (March/April 1995) were 26.9 and 27.7 ng/L,
respectively. Mean DBA concentrations during cruises four and five, which were not included in this
analysis, were 24.6 and 22.0 ng/L, respectively.
5-2
-------
Atrazine in the Open-Lake Water Column
Figure 5-1. Atrazine Concentrations Measured in Lake Michigan During Four Sampling
Cruises (Cruise 1 = April/May 1994, Cruise 2 = June 1994, Cruise 3 = August 1994, Cruise 6
= March/April 1995)a
O)
c
c
O
"ro
-t->
CD
O
O
O
55;
50|
40]
35-
Atrazine
B
A
o
25;
9Di
O
ZU~| i
o
'
o
E
'
O
1
E?
O
ii
~*i
o
a Boxes represent the 25th (box bottom), 50th (center line), and 75th (box top) percentile results. Bars represent the results nearest 1.5 times
the inter-quartile range (IQR=75th-25th percentile) away from the nearest edge of the box. Circles represent results beyond 1.5*IQR from the
box. Xs represent results beyond 3*IQR from the box. Letters above the boxes represent results of analysis of variance and multiple
comparisons test. Boxes with the same letter were not statistically different (at alpha = 0.05).
The mean DIA concentration measured during cruise six was significantly higher than mean DIA
concentrations measured during cruises one, two, or three (Figure 5-2). Mean DIA concentrations during
cruises one (April/May 1994), two (June 1994), and three (August 1994) were 12.4, 12.9, and 13.1 ng/L,
respectively; the mean DIA concentration during cruise six (March/April 1995) was 19.1 ng/L. Mean
DIA concentrations during cruises four and five, which were not included in this analysis, were 17.6 and
6.90 ng/L, respectively.
In conclusion, open-lake atrazine, DBA, and DIA concentrations increased during the one year sampling
campaign. Increases from April/May 1994 to March/April 1995 were statistically significant for each
analyte. During this period, lake-wide mean atrazine concentrations increased by 2.7 ng/L or 7.30%.
Mean DBA concentrations increased by 3.6 ng/L or 14.9%, and mean DIA concentrations increased by
6.7 ng/L or 54.0%.
5-3
-------
Results of the LMMB Study: Atrazine Data Report
Figure 5-2. DEA and DIA Concentrations Measured in Lake Michigan During Four
Sampling Cruises (Cruise 1 = April/May 1994, Cruise 2 = June 1994, Cruise 3 = August
1994, Cruise 6 = March/April 1995)a
en
g
'-i*
ro
-------
Atrazine in the Open-Lake Water Column
5.1.3 Vertical Variation
Open-lake samples were collected at various depths ranging from 1 to 257 m. Overall, there was no
correlation between atrazine concentrations and depth of samples collected. This is not surprising,
considering that the majority of samples collected during the LMMB Study were not collected during
periods of thermal stratification in Lake Michigan. Only cruise three in August 1994 was conducted
during highly stratified conditions.
Further investigation of vertical trends during stratified conditions revealed a statistically significant trend
within the northern Lake Michigan basin (p=0.034, based on a two-sample t-test). Within the northern
basin (above approximately 44° N latitude), atrazine concentrations were significantly higher (at the 95%
confidence level) in the epilimnion than in the hypolimnion. The mean atrazine concentration in the
epilimnion (at 20 m depth and above) was 42.0 ng/L, and the mean atrazine concentration in the
hypolimnion (below 20 m depth) was 35.9 ng/L. Figure 5-3 shows depth profiles for atrazine, DBA, and
DIA at site 41 in the northern Lake Michigan basin. This trend of increased concentrations in the
epilimnion during stratified conditions was not observed in the southern basin. Within the southern basin,
the mean atrazine concentration in the epilimnion (38.8 ng/L) was not significantly different from the
mean atrazine concentration in the hypolimnion (39.0 ng/L) (p=0.873, based on a two-sample t-test).
Figure 5-3. Depth Profile of Atrazine, DEA, and DIA Concentrations at
Station 41 in Lake Michigan on August 12,1994
Vertical trends of DIA were similar
to those for atrazine. Within the
northern basin, the mean DIA
concentration in the epilimnion
(15.7 ng/L) was significantly (at the
95% confidence level) higher than
the mean DIA concentration in the
hypolimnion (13.3 ng/L) (p=0.046).
Within the southern basin, the mean
DIA concentration in the epilimnion
(12.0 ng/L) was not significantly
different from the mean DIA
concentration in the hypolimnion
(12.1 ng/L) (p=0.873). Vertical
trends in DEA concentrations were
not statistically significant in either
basin (Northern: p=0.143; Southern:
p=0.478).
5.1.4 Analysis of Breakdown Products
DEA and DIA concentrations were correlated with atrazine concentrations in open-lake samples. As
atrazine concentrations in open-lake samples increased, DEA and DIA concentrations also generally
increased (Figure 5-4). This correlation was significant at the 95% confidence level, but the correlations
among atrazine, DEA, and DIA were not as strong in open-lake samples as observed for these analytes in
precipitation and tributary samples. This is most likely due to the smaller range of concentrations
measured in open-lake samples. Pearson correlation coefficients were 0.61 and 0.32 for atrazine
concentrations in open-lake samples compared to DEA and DIA concentrations, respectively.
Depth (m)
u
Art
sn
i nn
1 UU
ion
i Art
11-U
iRrt
m A
A
A
A Atrazine
DIA
A
10
20 30
Concentration (ng/L)
40
50
5-5
-------
Results of the LMMB Study: Atrazine Data Report
Figure 5-4. Correlation of DEA and DIA Concentrations with Measured Atrazine Concentrations in
Open-lake Samples
c
0)
Q.
O
c
O O
1s
£ 8
8 "S-
e E
O
-------
Atrazine in the Open-Lake Water Column
(2002) measured atrazine concentrations of 1000 to 4000 ng/L in a Kansas reservoir (Perry Lake). Even
remote lakes in the Midwest with presumably only atmospheric inputs of atrazine have measurable
atrazine concentrations. In lakes on the Isle of Royale (an island in northern Lake Superior), Thurman
and Cromwell (2000) measured 5 to 22 ng/L of atrazine in August 1992.
5.2.2 Comparison to Regulatory Limits
Measured atrazine concentrations in Lake Michigan were well below regulatory limits for human health
concerns and proposed ambient water quality criteria. The lake-wide mean atrazine concentration of 38.3
ng/L was more than 75 times below the 3000 ng/L maximum contaminant level for drinking water. When
compared to the proposed ambient water quality criterion maximum concentration of 350 i-ig/L, maximum
atrazine concentrations measured in this study were more than 6000 times less than the criterion. Mean
atrazine concentrations were more than 300 times less than the proposed ambient water quality criterion
continuous concentration of 12 i-ig/L.
5.2.3 Lateral Variation
Atrazine concentrations measured at stations located in Green Bay were statistically higher than several
Lake Michigan stations. This is likely due to the smaller volume of water available for dilution of
pollutants in Green Bay and the limited mixing between the bay and Lake Michigan. Within Lake
Michigan itself, atrazine concentrations were relatively consistent throughout the lake. Station averages
only ranged from 33.0 to 48.0 ng/L. Schottler and Eisenreich (1994) also found no lateral variation in
atrazine concentrations in Lake Michigan. They concluded that the lake was well mixed spatially with
respect to atrazine. This study confirms that conclusion.
5.2.4 Temporal Trends
This study found that atrazine, DEA, and DIA concentrations in Lake Michigan all increased from the
spring of 1994 to the spring of 1995. The measured increases of 7.30% in atrazine concentrations, 14.9%
in DEA concentrations, and 54.0% in DIA concentrations were all statistically significant. This agrees
with the increasing trend in atrazine concentrations reported by Schottler and Eisenreich (1994), who
found that Lake Michigan atrazine concentrations were statistically greater in 1992 than in 1991.
Combining results from the two studies, lake-wide atrazine concentrations in Lake Michigan have
increased from 34 ng/L in 1991 to 39.4 ng/L in 1995.
While significantly longer studies are needed to confirm long-term trends of increasing Lake Michigan
atrazine concentrations, the increases observed over the short 1991-1995 time period is consistent with
the modeling efforts of Rygwelski et al. (1999). Rygwelski et al. (1999) estimated that if continued use
of atrazine is sustained, atrazine concentrations in Lake Michigan would continue to rise until a
concentration of approximately 160 ng/L is reached in about 300 years. Depending upon atrazine
degradation half-life estimates ranging from two years to no degradation, Tierney et al. (1999) estimated
achievement of steady state conditions in 11 to 307 years. Tierney et al. (1999) estimated that steady
state concentrations would be 33, 77, 140, or 710 ng/L using degradation half-lives of 2 years, 5 years, 10
years, and no degradation, respectively. The two-year degradation half-life estimate is likely an
underestimate, based on Lake Michigan atrazine concentrations measured in this study, which have
already increased past 33 ng/L. In contrast to models that predict long-term increases in Lake Michigan
atrazine concentrations, Schottler and Eisenreich (1997) estimated that atrazine concentrations in Lake
Michigan are currently near steady state.
5-7
-------
Results of the LMMB Study: Atrazine Data Report
5.2.5 Vertical Trends
This study found that in the northern Lake Michigan basin, atrazine concentrations during stratification
were significantly higher in the epilimnion than in the hypolimnion. This pattern was not observed in the
southern Lake Michigan basin or when data from all sites (northern and southern basins) were combined.
In previous studies, Schottler and Eisenreich (1994) concluded that concentration profiles demonstrated
no consistent vertical trends, and epilimnetic atrazine concentrations were not statistically different from
hypolimnetic concentrations. The findings of these two studies do not necessarily contradict one another.
Both studies found that when all sites were combined, atrazine concentrations in the epilimnion were not
statistically different than hypolimnetic atrazine concentrations. The LMMB study went on to observe
that in just the northern basin, epilimnetic atrazine concentrations were statistically different from
hypolimnetic atrazine concentrations. This observation may have been allowed by the greater coverage
of lake stations in the current study (29 sites evaluated in the August 1994 cruise). Schottler and
Eisenreich (1994) investigated only ten Lake Michigan sites, six of which were in the southern basin.
Additional investigations are likely needed to determine if the observed trend in the northern basin
observed in this study is a statistical anomaly or environmentally relevant.
Muller et al. (1997) also observed increased atrazine concentrations in the epilimnion of small Swiss
lakes during stratification, and attributed this finding to increased atrazine loadings into the epilimnion
during spring and summer months when the lake was stratified.
5.3 Quality Implementation and Assessment
As described in Section 1.5.5, the LMMB QA program prescribed minimum standards to which all
organizations collecting data were required to adhere. The quality activities implemented for the atrazine
monitoring portion of the study are further described in Section 2.6 and included use of SOPs, training of
laboratory and field personnel, and establishment of MQOs for study data. A detailed description of the
LMMB QA program is provided in The Lake Michigan Mass Balance Study Quality Assurance Report
(USEPA, 2001 i). A brief summary of data quality issues for the open lake water column atrazine data is
provided below.
As discussed in Section 2.5, the sample collection, extraction, and analysis methods for atrazine
monitoring in this study are modifications of the methods used for the PCBs and chlorinated pesticides.
In addition, during an audit of analytical procedures, two changes to the analytical method described in
the approved QAPP were observed: addition of sodium sulfate to the extract in the centrifuge tube to
remove interfering water, and use of d5-ethylatrazine rather than dlO-anthracene as the internal standard.
These changes improved analyte recovery and quantitation and were included in the analytical SOP
(USEPA, 1997a; USEPA, 1997b).
Bottle blanks, laboratory blanks, field blanks, and trip blanks were prepared and analyzed. Atrazine,
DEA, and DIA were not detected in any of these blanks, indicating that contamination of study samples
did not occur from these sources. Field blanks and trip blanks were not collected at all study sites, and
therefore, sample contamination from site-specific sources cannot be evaluated. However, the variety of
blanks collected and the fact that all blank samples contained undetectable levels of atrazine, DEA, and
DIA suggests that contamination of study samples may be unlikely.
As discussed in Section 2.6, data verification was performed by comparing all field and QC sample
results produced by each PI with their MQOs and with overall LMMB Study objectives. Analytical
results were flagged when pertinent QC sample results did not meet acceptance criteria as defined by the
MQOs. These flags were not intended to suggest that data were not useable; rather they were intended to
caution the user about an aspect of the data that did not meet the predefined criteria. Table 5-2 provides a
5-8
-------
Atrazine in the Open-Lake Water Column
summary of flags applied to the open-lake atrazine data. The summary provided below includes the flags
that directly relate to evaluation of the MQOs to illustrate some aspects of data quality, but does not
include all flags applied to the data to document sampling and analytical information, as discussed in
Section 2.6. In this data report, the summary and analysis of atrazine data represent all results with the
exception of those flagged as "invalid" by the QC coordinator in concert with the PI.
Table 5-2. Summary of Routine Field Sample Flags for the Analysis of Atrazine in Open-lake Samples
Flag3
UNO, Analyte not detected
EHT, Exceeded holding time
FSS, Failed surrogate
FFD, Failed field duplicate
Percentage of Samples Flagged (%)b
0
42% (99)
9% (21)
0
a The summary provides only a subset of applied flags and does not represent the full suite of flags applied to the data.
b The number of routine field samples flagged is provided in parentheses.
All of the open-lake samples contained detectable concentrations of atrazine so none of the open-lake
samples were flagged as "analyte not detected". Forty-two percent of open-lake samples were flagged for
exceeding the sample holding time of 30 days to extraction. However, the holding times for atrazine, and
many other environmental pollutants, are not well-characterized and the effects on the sample results are
unknown. All of the results for the field duplicate pairs were within acceptance criteria (i.e., none of the
sample results were flagged as failed field duplicate), and only 9% of open-lake sample results showed
surrogate recoveries outside acceptance criteria. . Based on the evaluation of multiple QC sample results
and surrogate recoveries, the PI and the QC coordinator did not assign high bias (HIB) or low bias (LOB)
flags to any sample results.
As discussed in Section 1.5.5, MQOs were defined in terms of six attributes: sensitivity, precision,
accuracy, representativeness, completeness, and comparability. GLNPO derived data quality assessments
based on four of these attributes. For example, system precision was estimated as the mean relative
percent difference (RPD) between the results for field duplicate pairs. Similarly, analytical precision was
estimated as the mean relative percent difference (RPD) between the results from laboratory duplicate
pairs. Table 5-3 provides a summary of data quality assessments for several of these attributes for the
open-lake atrazine study data.
System precision, estimated as the mean RPD between field duplicate results, was 6.08%. Analytical
precision, estimated as the standard deviation of surrogate correction factors, was 0.250. These measures
suggest that analytical results are highly reproducible. Analytical bias, estimated as the mean surrogate
recovery for the study samples, was 91.3% suggesting an overall slight low bias.
5-9
-------
Results of the LMMB Study: Atrazine Data Report
Table 5-3. Data Quality Assessment for the Analysis of Atrazine in Open-lake Samples
Parameter3
Number of Routine Samples Analyzed
Number of Field Duplicates Analyzed
System Precision, Mean Field Duplicate RPD (%), >MDL
Analytical Precision, SCF Variability (SD)
Analytical Bias, Mean Surrogate Recovery (%)
Analytical Sensitivity , Samples reported as < MDL (%)
Assessment13
234
59
6.08%
(59)
0.250
(234)
91.3%
(234)
0%
a RPD = Relative percent difference.
MDL = Method detection limit.
SCF = Surrogate correction factor.
b Number of sample/duplicate pairs used in the assessment is provided in parentheses.
5-10
-------
References
Battaglin, W.A., E.T. Furlong, M.R. Burkhardt, and C.J. Peter. 2000. Occurrence of sulfonylurea,
sulfonamide, imidazolinone, and other herbicides in rivers, reservoirs and ground water in the
Midwestern United States, 1998. The Science of the Total Environment. 248: 123-133.
Capel, P.D., M.Lin, and P.J. Wotzka. 1997. Wet Atmospheric Deposition of Pesticides in Minnesota,
1989-94. Water Resources Investigations Report 97-4026. U.S. Geological Survey.
Capel, P.D. and S.J. Larson. 2001. Effect of scale on the behavior of atrazine in surface waters.
Environmental Science and Technology. 35: 648-657.
Clark, G.M., D.A. Goolsby, and W.A. Battaglin. 1999. Seasonal and annual load of herbicide from the
Mississippi River Basin to the Gulf of Mexico. Environmental Science and Technology. 33:
981-986.
Fallen, J.D., D.P. Tierney, and E.M Thurman. 2002. Movement of atrazine and deethylatrazine through
a midwestern reservoir. Journal of the American Water Works Association. 94(11): 54-66.
Foreman, W.T., M.S. Majewski, D.A. Goolsby, F.W. Wiebe, and RH. Coupe. 2000. Pesticides in the
atmosphere of the Mississippi River Valley, part II - air. The Science of the Total Environment.
248: 213-226.
Glotfelty, D.E., M.M. Leech, J. Jersey, and A.W. Taylor. 1989. Volatilization and wind erosion of
surface soil applied atrazine, simazine, alachlor, and toxaphene. J. Agric. Food Chem. 37: 546-
551.
Goolsby, D.A., E.M. Thurman, M.L. Pomes, M.T. Meyer, and W.A. Battaglin. 1997. Herbicides and
their metabolites in rainfall: origin, transport, and deposition patterns across the midwestern and
northeastern United States, 1990-1991. Environmental Science and Technology. 31: 1325-1333.
Hayes, T., Hasten, K., Tsui, M., Hoang, A., Haeffele, C., Vonk, A., 2002. Feminization of male frogs in
the wild. Nature. 419,895-896.
Kolpin, D.W. and S. J. Kalkhoff. 1993. Atrazine degradation in a small stream in Iowa. Environmental
Science and Technology. 27:134-139.
Lemieux, C., B. Quemerais, and K.R. Lum. 1995. Seasonal patterns of atrazine loading for the St.
Lawrence River (Canada) and its tributaries. Water Research. 29:1491-1504.
Majewski, M.S., W.T. Foreman, and D.A. Goolsby. 2000. Pesticides in the atmosphere of the
Mississippi River Valley, part I - rain. The Science of the Total Environment. 248: 201-212.
Miller, S.M., C.W. Sweet, J.V. Depinto, and K.C. Hornbuckle. 2000. Atrazine and nutrients in
precipitation: results from the Lake Michigan Mass Balance Study. Environmental Science and
Technology. 34:55-61.
Muller, S.R., M. Berg, M.M. Ulrich, and R.P. Schwarzenbach. 1997. Atrazine and its primary
metabolites in Swiss lakes: input characteristics and long-term behavior in the water column.
Environmental Science and Technology. 31: 2104-2113.
R-1
-------
Results of the LMMB Study: Atrazine Data Report
Nations, B. K. and G.R. Hallberg. 1992. Pesticides in Iowa precipitation. Journal Environmental
Quality. 21:486-492.
Pereira, W.E. and F.D. Hostettler. 1993. Nonpoint source contamination of the Mississippi River and its
tributaries by herbicides. Environmental Science and Technology. 27: 1542-1552.
Ribaudo, M.O. and A. Bouzaher. 1994. Atrazine: Environmental Characteristics and Economics of
Management. U.S. Department of Agriculture, Washington, B.C.
Richards, R.P. and D.B. Baker. 1993. Pesticide concentration patterns in agricultural drainage networks
in the Lake Erie Basin. Environmental Toxicology and Chemistry. 12: 13-26.
Richards, R.P., D.B. Baker, J.W. Kramer, and D.E. Ewing. 1996. Annual loads of herbicides in Lake
Erie tributaries of Michigan and Ohio. Journal of Great Lakes Research. 22(2): 414-428.
Rygwelski, K.R., W.L. Richardson, and D.D. Endicott. 1999. A screening-level model evaluation of
atrazine in the Lake Michigan basin. Journal of Great Lakes Research. 25: 94-106.
Schottler, S.P. and S.J. Eisenreich. 1994. Herbicides in the Great Lakes. Environmental Science and
Technology. 28: 2228-2232.
Schottler, S.P. and S.J. Eisenreich. 1997. Mass balance model to quantify atrazine sources,
transformation rates, and trends in the Great Lakes. Environmental Science and Technology. 31:
2616-2625.
Schottler, S.P., S.J. Eisenreich, and P.D. Capel. 1994. Atrazine, alachlor, and cyanizine in a large
agricultural river system. Environmental Science and Technology. 28: 1079-1089.
Spalding, R.F., D.D. Snow, D.A. Cassada, and M.E. Burbach. 1994. Study of pesticide occurrence in
two closely spaced lakes in northeastern Nebraska. Journal of Environmental Quality. 23: 571-
578.
Stamer, J.K., R.B. Swanson, and P.R. Jordan. 1994. Atrazine in spring runoff as related to
environmental setting in Nebraska, 1992. Water Resources Bulletin. 30(5): 823-831.
Thurman, E.M and A.E. Cromwell. 2000. Atmospheric transport, deposition, and fate of triazine
herbicides and their metabolites in pristine areas at Isle Royale National Park. Environmental
Science and Technology. 34: 3079-3085.
Thurman, E.M., D.A. Goolsby, M.T. Meyer, M.S. Mills, M.L. Pomes, and D.W. Kolpin. 1992. A
reconnaissance study of herbicides and their metabolites in surface water of the midwestern
United States using immunoassay and gas chromatography/mass sprectrometry. Environmental
Science and Technology. 26: 2440-2447.
Tierney, D.P., P.A. Nelson, B.R. Cristensen, and S.M.K. Watson. 1999. Predicted atrazine
concentrations in the Great Lakes: implications for biological effects. Journal of Great Lakes
Research. 25(3): 455-467.
R-2
-------
References
USEPA. 1995a. National Primary Drinking Water Regulations, Contaminant Specific Fact Sheets,
Inorganic Chemicals, Technical Version. EPA 811/F-95/002-T. U.S. Environmental Protection
Agency, Office of Water, Washington, B.C.
USEPA. 1995b. National Primary Drinking Water Regulations, Contaminant Specific Fact Sheets,
Synthetic Organic Chemicals, Technical Version. EPA 811/F-95/003-T. U.S. Environmental
Protection Agency, Office of Water, Washington, B.C.
USEPA. 1997a. Lake Michigan Mass Balance Project (LMMB) Methods Compendium, Volume 1:
Sample Collection Techniques. EPA 905/R-97/012a. U.S. Environmental Protection Agency,
Great Lakes National Program Office, Chicago, IL.
USEPA. 1997b. Lake Michigan Mass Balance Project (LMMB) Methods Compendium, Volume 2:
Organic and Mercury Sample Analysis Techniques. EPA 905/R-97/012b. U.S. Environmental
Protection Agency, Great Lakes National Program Office, Chicago, IL.
USEPA. 1997c. Lake Michigan Mass Budget/Mass Balance Work Plan. EPA 905/R-97/018. U.S.
Environmental Protection Agency, Great Lakes National Program Office, Chicago, IL.
USEPA. 1997d. The Enhanced Monitoring Program Quality Assurance Program Plan. EPA
905/R-97/017. U.S. Environmental Protection Agency, Great Lakes National Program Office,
Chicago, IL.
USEPA. 1997e. Mercury Study Report to Congress. EPA 452/R-97/003. U.S. Environmental
Protection Agency, Office of Air Quality Planning and Standards and Office of Research and
Bevelopment, Research Triangle Park, NC.
USEPA. 1998. The Lake Michigan Mass Balance Project Quality Assurance Plan for Mathematical
Modeling, Version 3.0. U.S. Environmental Protection Agency, Great Lakes National Program
Office, Chicago, IL.
USEPA. 1999. National Recommended Water Quality Criteria-Correction. EPA 822/Z-99/001. U.S.
Environmental Protection Agency, Office of Water, Washington, B.C.
USEPA. 2001a. Ambient Aquatic Life Water Quality Criteria for Atrazine. U.S. Environmental
Protection Agency, Office of Water, Washington, B.C.
USEPA. 2001b. Atrazine; availability of preliminary risk assessment. Federal Register. 66: 10287-
10289.
USEPA. 200 Ic. Drinking Water Exposure Assessment for Atrazine and Various Chloro-triazine and
hydroxy-triazine Degradates. U.S. Environmental Protection Agency, Office of Pesticide
Programs, Washington, B.C.
USEPA. 200 Id. Notice of availability of draft aquatic life criteria document for atrazine. Federal
Register. 66: 49186-49188.
USEPA. 2001e. Reregistration Eligibility Science Chapter for Atrazine: Fate and Environmental Risk
Assessment Chapter. January, 2001. U.S. Environmental Protection Agency, Office of Pesticide
Programs, Washington, B.C.
R-3
-------
Results of the LMMB Study: Atrazine Data Report
USEPA. 2001f. Results of the Lake Michigan Mass Balance Study: Poly chlorinated Biphenyls and
trans-Nonachlor Data Report. EPA 905/R-01/011. U.S. Environmental Protection Agency,
Great Lakes National Program Office, Chicago, IL.
USEPA. 200 Ig. Results of the Lake Michigan Mass Balance Study: Mercury Data Report. EPA 905/R-
01/012. U.S. Environmental Protection Agency, Great Lakes National Program Office, Chicago,
IL.
USEPA. 2001h. Revised Preliminary Human Health Risk Assessment: Atrazine. January, 2001. U.S.
Environmental Protection Agency, Office of Pesticide Programs, Washington, B.C.
USEPA. 2001L The Lake Michigan Mass Balance Study Quality Assurance Report. EPA 905/R-01/013.
U.S. Environmental Protection Agency, Great Lakes National Program Office, Chicago, IL.
USGS. 1998. Annual Use Maps. U.S. Geological Survey, National Water Quality Assessment, Pesticide
National Synthesis Project, http://ca.water.usgs.gov/pnsp/use92.
Van Dijk, H.F.G. and R. Guicherit. 1999. Atmospheric dispersion of current-use pesticides: a review of
the evidence from monitoring studies. Water, Air, and Soil Pollution. 115: 21-70.
Williams, A.L., C.W. Sweet, and C. Peters. 1992. Herbicide concentrations in air and rain. In
Proceedings of the 85th Annual Meeting of the Air and Waste Management Association, Paper 92-
62.16. Air and Waste Management Association, Pittsburgh, PA.
Wisconsin Department of Agriculture. 1994. Wisconsin Crop Weather, Vol. 17(11). Wisconsin
Agricultural Statistics Service, Wisconsin Department of Agriculture, Madison, WI.
Worthing, C.R., and R.J. Hance. 1991. The Pesticide Manual. Ninth ed. British Crop Protection
Council, Surrey, UK.
R-4
------- |