&EPA
United States
Environmental Protection
Agency
Office of
Research and Development
Washington, DC 20460
EPA/600/6-91/003
February 1991
Methods for Aquatic
Toxicity Identification
Evaluations
Phase I Toxicity
Characterization Procedures
Second Edition
-------
EPA/600/6-91/003
February 1991
Methods for Aquatic
Toxicity Identification Evaluations
Phase I Toxicity Characterization Procedures
(Second Edition)
Edited by
T.J. Norberg-King
D.I. Mount
E.J. Durhan
G.T. An kley
L.P. Burkhard
Environmental Research Laboratory
Duluth, MN 55804
J.R. Amato
M.T. Lukasewycz
M.K. Schubauer-Berigan
AScI Corporation
6201 Congdon Boulevard
Duluth, MN 55804
L. Anderson-Carnahan
Region IV - Policy Planning & Evaluation Branch
Atlanta, GA 30365
National Effluent Toxicity
Assessment Center
Technical Report 18-90
Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Duluth, MN 55804
Printed on Recycled Paper
-------
Notice
This document has been reviewed in accordance with U.S. Environmental Protection
Agency Policy and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
-------
Foreword
This document is one in a series of guidance documents intended to aid dischargers and
their consultants in conducting aquatic organism Toxicity Identification Evaluations (TIEs)
as part of Toxicity Reduction Evaluations (TREs). Such effluent evaluations may be required
as the result of an enforcement action or as a condition of a National Pollutant Discharge
Elimination System (NPDES) permit. This document will also help to provide U.S. Environ-
mental Protection Agency (EPA) and State Pollution Control Agency staff with the back-
ground necessary to oversee and determine the adequacy of effluent TIEs proposed and
performed by NPDES permittees. While this TIE approach was developed for effluents, the
methods and techniques have direct applicability to other types of aqueous samples, such
as ambient waters, sediment pore waters, sediment elutriates, and hazardous waste
leachates.
The TIE approach is divided into three phases. Phase I (this document) contains
methods to characterize the physical/chemical nature of the constituents which cause
toxicity. Such characteristics as solubility, volatility and filterability are determined without
specifically identifying the toxicants. Phase I results are intended as a first step in specifically
identifying the toxicants but the data generated can also be used to develop treatment
methods to removetoxicity without specific identification of the toxicants. Two EPA TRE
manuals (EPA, 1989A; 1989B) use parts of Phase I in developing those approaches.
Phase II (EPA, 1989C) describes methods to specifically identify toxicants if they are
non-polar organics, ammonia, or metals. This Phase is incomplete because methods for
other specific groups, such as polar organics, have not yet been developed. As additional
methods are developed, they will be added.
Phase III (EPA, 1989D) describes methods to confirm the suspected toxicants. It is
applicable whether or not the identification of the toxicants was made using Phases I and
II. Complete Phase III confirmations have been limited to date, but avoiding Phase III may
invite disaster because the suspected toxicant(s) was not the actual toxicant(
Phases I and II are intended for acutely toxic effluents. However, that limitation does not
mean that effluents having chronic limits cannot be evaluated using these methods. TIE
methods to evaluate the cause of chronic toxicity in effluents are being developed (EPA,
1991 A).
These methods are not mandatory but are intended to aid those who need to character-
ize, identify or confirm the cause of toxicity in effluents or other aqueous samples such as
ambient waters, sediments, and leachates. Where we lack experience, we have indicated
this and have suggested avenues to follow. All tests need not be done on every sample; the
tests are, in general, independent. However, experience has taught us that skipping tests
may result in wasted time, especially during the early stages of Phase I. An exception to this
is when one wants to know if only a specific substance, for example ammonia, is causing
the toxicity or if toxicants other than ammonia are involved. Otherwise, we urge the use of
the whole battery of tests.
We welcome comments from users of these manuals so that future editions can be
improved. Comments can be sent to NETAC, ERL-Duluth, 6201 Congdon Boulevard,
Duluth, MN 55804.
-------
Abstract
In 1988, the first edition "Methods for Aquatic Toxicity Identification Evaluations:
Phase I Toxicity Characterization Procedures" was published (EPA, 1988A). This second
edition provides more details and more insight into the techniques described in the 1988
document. The manual describes procedures for characterizing the physical/chemical
nature of toxicants in acutely toxic effluent samples, with applications to other types of
samples such as receiving water samples, sediment pore water or elutriate samples, and
hazardous wastes. The presence and the potency of the toxicants in the samples are
detected by performing various manipulations on the sample and by using aquatic organ-
isms to track the changes in the toxicity. This toxicity tracking step is the basis of the toxicity
identification evaluation (TIE). The final step is to separate the toxicants from the other
constituents in the sample in order to simplify the analytical process. Many toxicants must
be concentrated for analysis.
The Phase I manipulations include pH changes along with aeration, filtration, sparging,
solid phase extraction, and the addition of chelating (i.e., ethylenediaminetetraacetate
ligand (EDTA)) and reducing (i.e., sodium thiosulfate) agents. The physical/chemical
characteristics of the toxicants are indicated by the results of the toxicity tests conducted on
the manipulated samples.
Since the first document was developed, additional options or new procedures have
been developed. For example, additional options are provided in the EDTA and sodium
thiosulfate addition tests, and in the graduated pH test. Also a discussion has been added
for testing the effluent sample over time (weekly) to measure the rate of decay of toxicity
which is used to detect the presence of degradable substances, particularly chlorine or
surfactants. Guidance for characterizing whether a toxicant(s) removed by aeration is
sublatable is described, and techniques for characterizing filterable toxicity and adiscussion
of C18 solid phase extraction elutable toxicity has been added. Use of multiple manipulations
is dtscussed and example interpretations of the results of the Phase I manipulations are
provided.
Additional manuals describe the methods used to specifically identify the toxicants
(EPA, 19896) and to confirm whether or not the suspect toxicant(s) is the actual toxicant(s)
(EPA, 19890).
-------
Contents
Page
Foreword in
Abstract iv
Contents v
Figures vii
Tables viii
Acknowledgments ix
1. Introduction I-I
1.1 Background I-I
1.2 Conventional Approach to TIEs l-l
1.3 Toxicity Based Approach I-3
2. Health and Safety 2-I
3. Quality Assurance 3-I
3.1 TIE Quality Control Plans 3-I
3.2 Cost Considerations/Concessions 3-I
3.3 Variability 3-2
3.4 Intra-Laboratory Communication 3-2
3.5 Record Keeping 3-2
3.6 Phase I Considerations 3-2
3.7 Phase II Considerations 3-3
3.8 Phase III Considerations ._. 3-3
4. Facilities and Equipment 4-I
5. Dilution Water 5-I
6. Effluent Sampling and Handling 6-I
6.1 Sample Shipment and Collection in Plastic versus Glass 6-3
7. Toxicity Tests 7-I
7.1 Principles 7-I
7.2 Test Species 7-I
7.3 Toxicity Test Procedures 7-2
7.4 Test Endpoints 7-3
7.5 Feeding 7-5
7.6 Multiple Species 7-5
8. Phase I Toxicity Characterization Tests 8-I
8.1 Initial Effluent Toxicity Test 8-4
8.2 Baseline Effluent Toxicity Test 8-5
8.3 pH Adjustment Test 8-8
8.4 pH Adjustment/Filtration Test 8-15
8.5 pH Adjustment/Aeration Test 8-21
-------
Contents (continued)
Page
8.6 pH Adjustment/C,, Solid Phase Extraction Test 8-27
8.7 Oxidant Reduction Test 8-33
8.8 EDTA Chelation Test 8-38
8.9 Graduated pH Test 8-44
9. Time Frame and Additional Tests : 9-1
9.1 Time Frame for Phase I Studies 9-1
9.2 When Phase I Tests are Inadequate 9-1
9.3 Interpreting Phase I Results 9-2
9.4 Interpretation Examples 9-3
10. References 10-1
VI
-------
Figures
Number Page
l-l. Conventional approach to TIEs 1-2
1-2. Flow chart for toxicity reduction evaluations I-5
6-1. Example data sheet for logging in samples 6-2
7-I. Schematic for preparing effluent test concentrations using
simple dilution techniques 7-4
8-I. Overview of Phase I effluent characterization tests 8-2
8-2. Example data sheet for initial effluent toxicity test 8-6
8-3. Example data sheet for baseline effluent toxicity test 8-7
8-4. p£ -pH diagrams for the CO,, H20, and Mn-CO, systems (25°C) 8-9
8-5. Flow chart for pH adjustment tests 8-11
8-6. Example data sheet for pH adjustment test 8-13
8-7. Overview of steps needed in preparing the filter and dilution water blanks
for the filtration and/or the C18 SPE column tests 8-17
8-8. Overview of steps needed in preparing the effluent for the filtration and/or
C,8 SPE column tests 8-18
8-9. Example data sheet for filtration test 8-20
8-10. Diagram for preparing pH adjustment/aeration test samples 8-23
8-11. Example data sheet for aeration test 8-24
8-12. Closed loop schematic for volatile chemicals 8-26
8-13. Step-wise diagram for preparing the C18 SPE column samples 8-29
8-14. Example data sheet for effluent SPE test with and without pH adjustment 8-3 1
8-15. Example data sheet for the oxidant reduction test when using a
gradient of sodium thiosulfate concentrations 8-36
8-I 6. Example data sheet for the oxidant reduction test when effluent
dilutions are used 8-37
8-I 7. Example data sheet for EDTA chelation test when using a gradient of
EDTA concentrations 8-41
8-I 8. Example data sheet for the EDTA chelation test when
effluent dilutions are used 8-42
8-I 9. Example of data sheet for the graduated pH test when
effluent dilutions are used 8-46
VII
-------
Tables
Number Page
6-1. Volumes needed for Phase I tests 6-3
8-1. Outline of Phase I effluent manipulations ' 8-3
8-2. Acute toxicity of sodium chloride to selected aquatic organisms 8-14
8-3. Toxicity of methanol to several freshwater species * 8-32
8-4. Toxicity of sodium thiosulfate to Ceriodaphnia dubia, Daphnia magna, and
fathead minnows 8-34
8-5. Toxicity of EDTA to Ceriodaphnia dubia and fathead minnows in
water of various hardnesses and salinities 8-39
8-6. The toxicity of the Mes, Mops, and Popso buffers to Ceriodaphnia dubia and
fathead minnows 8-48
VIII
-------
Acknowledgments
We wish to acknowledge the assistance of many people from the Environmental
Research Laboratory-Duluth (ERL-D) for their help and advice in preparing the manual. The
experience referred to throughout the document is the collective experience of the individu-
als in the National Effluent Toxicity Assessment Center (NETAC). During the development
of this document and the first edition, that group has consisted of Don Mount, Teresa
Norberg-King, Larry Burkhard, Liz Durhan, Gary Ankley, Shaneen Murphy (all ERL-D staff),
Linda Anderson-Carnahan, (EPA, Region IV), Joe Amato, Marta Lukasewycz, Greg
Peterson, Jim Taraldsen, Jim Jenson, Mary Schubauer-Berigan, Art Fenstad, Doug Jensen,
Steve Baker, Liz Makynen, Jo Thompson, Correne Jenson, Lara Andersen, Linda
Eisenschenk, Nola Englehorn, and Eric Robert, (all currently or formerly with AScI Corpo-
ration, Duluth).
For this revision, several individuals were assigned sections to write or re-write. This
group consisted of Don Mount, Teresa Norberg-King, Larry Burkhard, Liz Durhan, Gary
Ankley, Marta Lukasewycz, Joe Amato, and Mary Schubauer-Berigan. Teresa synthesized
all the rewrites into similar styles, added additional sections, and updated the entire
document. The assistance Debra Williams (AScI) provided to produce the graphics, to
prepare the document, and assist in all aspects is greatly appreciated. Without her input the
production of the document would have been slowed tremendously. Much of the data have
been developed by a few people: Joe Amato generated much of the laboratory data that is
discussed in this document, and others also generated the data for the tables (Jim Jenson,
Doug Jensen, Shaneen Murphy, Greg Peterson, Gary Ankley, Mary Schubauer-Berigan,
and Jo Thompson).
We also want to express our appreciation to Russ Hockett and Dave Mount (ENSR, Fort
Collins, CO) for their data on sodium thiosulfate and W. Tom Waller (University of North
Texas, Denton, TX) for his data and his technique for sulfur dioxide dechlorination.
As stated in the first edition of the TIE characterization document, the effluent group
would not have been able to complete the work that is summarized in this report without the
support and backing of Nelson Thomas, Senior Advisor for National Programs (ERL-D). In
addition, Rick Brandes (EPA, Permits Division, Washington, D.C.) has been a strong voice
in support of all the work upon which the manual is based. The support provided from the
Off ice of Water through his impetus has enabled NETAC to become a well-established and
well-staffed center.
This manual is truly the result of the effort of many people. We welcome your suggestions
for improvement so that any future revision can make the methods more useful.
-------
Section 1
Introduction
1. I Background
The Clean Water Act (CWA, 1972) provides the
basis for control of toxic substances discharged to
waters of the United States. The Declaration of Goals
and Policy of the Federal Water Pollution Control Act of
1972 states that ". ..it is the national policy that the
discharge of toxic pollutants in toxic amounts be pro-
hibited." This policy statement has been maintained in
all subsequent versions of the CWA.
It is the goal of the CWA that zero discharge of
pollutants to waters of the U.S. be achieved. Because
this goal is not immediately attainable, the CWA allows
for National Pollutant Discharge Elimination System
(NPDES) permits for wastewater discharges. The five
year NPDES permits contain technology-based effluent
limits reflecting the best controls available. Where these
technology-based permit limits do not protect water
quality, additional water quality-based limits are included
in the NPDES permit in order to meet the CWA policy
of "no toxic pollutants in toxic amounts.*' State narrative
and numerical water quality standards are used in con-
junction with EPA criteria and other toxicity databases
to determine the adequacy of technology-based permit
limits and the need for additional water quality-based
controls.
To insure that the CWA's prohibitions on toxic dis-
charges are met, EPA has issued a "Policy for the
Development of Water Quality-Based Permit Limita-
tions for Toxic Pollutants" (Federal Register, 1984).
This national policy recommends an integrated approach
for controlling toxic pollutants that uses whole effluent
toxicity testing to complement chemical-specific analy-
ses. The use of whole effluent toxicity testing is neces-
sitated by several factors including a) the limitations
presented by chemical analysis methods, b) inadequate
chemical-specific aquatic toxicity data, and c) inability
to predict the aggregate toxicity of chemicals in an
effluent.
To determine the toxicity of effluents to aquatic life,
standardized methods for measuring acute and chronic
toxicity have been developed by EPA (EPA, 1985A;
EPA, 1988B; EPA, 1989E). These cost-effective meth-
ods facilitate the inclusion of whole effluent toxicity
limits and biomonitoring conditions in NPDES permits
for facilities suspected of causing violations of state
water quality toxicity standards.
As a result of the increasing use of aquatic organ-
ism toxicity limits and biomonitoring conditions in per-
mits, a substantial number of unacceptably toxic efflu-
ents have been and continue to be identified. To rectify
these problems, permittees are being required, through
permit conditions and administrative orders or other
enforcement actions, to perform effluent toxicity reduc-
fion evaluations (TREs). The object of the TRE is to
determine which measures are necessary to maintain
the effluent's toxicity at acceptable levels. Such evalua-
tions, however, have often proven to be very compli-
cated.
The goal of the TRE wilt be set by either the state
regulatory agency or EPA and will be dependent on
state standards that define acceptable levels of toxicity
in the receiving water and effluent. Because of trfts, and
because specific TRE actions may also be required,
communication between the regulators and TRE inves-
tigators is crucial.
This document provides NPDES permittees with
procedures to assess the nature of effluent toxicity to
aquatic organisms. It is intended for use by those
permittees having difficulty meeting their permit for whole
effluent aquatic organism toxicity limits or permittees
required, through special conditions, to reduce or elimi-
nate effluent toxicity. This document does not address
human health toxicity concerns such as those from
bioconcentration, water supplies and recreational uses.
The methods are applicable to identifying the cause of
toxicity for samples other than effluents which display
acute toxicity, such as ambient water samples, elutriates
and pore waters from sediments, and possibly leachates.
While we generally refer to effluents, the application of
the techniques for any aqueous sample is implied.
These methods may have applicability to effluents and
other types of samples that exhibit chronic toxicity as
well.
1.2 Conventional Approach to TIEs
In order to appreciate the complexities involved in a
typical effluent toxicity identification evaluation (TIE),
one must first understand the drawbacks in what can
be considered the conventional approach to the prob-
lem of controlling toxics. The following discussion is
meant, to exemplify the need for a logical approach
which builds on the effluent data as they are being
collected.
Traditionally, when an effluent has been identified
as toxic or is suspected of being toxic to aquatic organ-
isms, a sample of the wastewater is analyzed for the
126 Apriority pollutants." The concentration of each pri-
1-1
-------
Toxic Eff luent
Generate
Eff luent LC50
Conduct
Priority
Pollutant
Analysis
Mass Balance
and
Comparison
t
Evaluate Effluent
Constituents and
Their
Concentrations
Search Literature
for Aquatic
Toxicity
Data on Effluent
Constituents
Figure l-l. Conventional approach to TIEs.
ority pollutant present in the sample is subsequently
compared to literature toxicity data for the pollutant, or
is compared to EPA's Ambient Water Quality Criteria or
state standards for aquatic life protection for that com-
pound. The goal of this exercise is to determine which
pollutants in the wastewater sample are responsible for
effluent toxicity (Figure l-l). Unfortunately, determining
the source of an effluent's toxicity is rarely this straight-
forward.
The first problem encountered in this course is one
of effluent variability. Because toxicity is a generic re-
sponse, there is no way .to determine whether the
toxicity observed over time is consistently caused by a
single constituent or a combination of constituents or a
number of different constituents, each acting periodi-
cally to cause effluent toxicity. Experience has shown
that the latter may be a frequent occurrence especially
in publicly owned treatment works (POTW) effluents.
To further complicate the problem, the variability in
conventional effluent monitoring parameters may not
coincide with variability in the effluent toxicant( Moni-
toring methods for conventional parameters such as
biological oxygen demand (BOD) frequently are not
responsive to shifts in the toxicants because they are at
relatively low concentrations in the effluent or simply
1-2
-------
because the toxicants are not amenable to analysis by
these procedures. For the conventional TIE approach
to be successful, it is crucial that the same sample be
analyzed using both chemical and biological techniques,
and that a number of samples over time be studied to
assess the variability in the toxicant(
A second problem with the conventional approach
involves the focus on the priority pollutants. These
have become known as the "toxic pollutants," convey-
ing an implication that they constitute the universe of
toxic chemicals but the priority pollutants are only a tiny
fraction of all chemicals. Limiting the search to these
126 compounds will result in failure to identify the
cause of toxicity in most cases.
On the surface, solving this difficulty may seem
inconsequential; the effluent analysis must include moni-
toring techniques for "non-priority" as well as priority
pollutants. To analyze an effluent for every chemical
would cost tens of thousands of dollars and there
would be no assurance that the detection levels would
be low enough. Determination of the composition of an
effluent is limited to the analyses used. For instance
gas chromatography/mass spectrometry (GC/MS) will
not identify cadmium and Inductively Coupled Emission
Spectroscopy (ICP) may not detect it when the concen-
tration is low. The absence of a measurable quantity of
any substance at the method detection level is often
interpreted as meaning that it is not present in the
effluent at all or not at toxic levels.
The toxicants may be present at low concentrations
because only small concentrations of highly toxic chemi-
cals are needed to produce toxicity. If this is true, then
low concentrations must be measured. Such chemicals
are not easily found by examining system loadings. For
example, if a chemical has an LC50 of 1 \ig/L, 380 g
(less than a pound per day) of the compound must be
present to cause lethality in the effluent of a 100 million
gallons per day (mgd) treatment plant. With a removal
efficiency of 99%, a loading of only 100 pounds per day
would be needed to produce a toxic effluent. Clearly
then, large loadings cannot be used to guide selection
of analytical techniques, and loads of a few pounds in a
collection system producing 100 mgd may be next to
impossible to identify by the usual methods of estab-
lishing loadings.
Many analytical methods are relatively limited in
their applicability. Even GC/MS, an instrument heavily
relied upon in typical wastewater analyses, is incapable
of detecting about 80% of all synthetic organic com-
pounds (G. Veith, personal communication, ERL-Duluth).
This limitation is related to selection and efficiency of
solvent extraction techniques, analyte volatility and ther-
mal stability, detector specificity and sensitivity, and
analytical interferences and artifacts. The percentage
of organics detected can be improved by derivatization
but the results are much more difficult to interpret. In
general, the broader spectrum methods are less sensi-
tive and require higher concentrations of analytes for
detection and are costly. To detect lower concentra-
tions, more specific methods are usually more sensi-
tive. To choose specific methods one must have knowl-
edge of the toxicants-knowledge which does not exist,
since that is the purpose of the analyses.
Surprisingly, even with these limitations, one usu-
ally sees lengthy lists of effluent constituents when
analyses are performed on wastewater. In the case of
GC/MS chromatograms, large peaks of non-toxic efflu-
ent constituents can overlap and hide smaller peaks
that may represent the toxicants of concern. When
many chemicals are present, the number of peaks that
can be identified may be small. Failure to identify a
component does not mean that the chemical is not
toxic. By using reference spectra, many peaks may be
tentatively identified as several different compounds
which serves only to increase, not decrease, the num-
ber of possibilities. No aquatic toxicity data will be
available for most of these compounds, so toxicity data
must be generated during the study. Compounds may
need to be synthesized in order to test them because
they are not available commercially. For those com-
pounds for which aquatic toxicity data are available, the
data may not include the species used for the TIE.
Even if all this work is done, trying to pinpoint the cause
of toxicity in such a complex mixture is likelyjo fail
because this approach does not include matrix effects
and toxicant bioavailability. For example, several met-
als may be present in an effluent sample at concentra-
tions well above the toxic threshold. These metals may
not be the source of the effluent's toxicity, however,
because they are not biologically available. Character-
istics such as total organic carbon (TOC), total sus-
pended solids (TSS), ionic strength, pH, hardness and
alkalinity can change toxicity. The inability to quantitate
the effects these parameters have on toxicity further
decreases the chances for a successful TIE.
1.3 Toxicity Based Approach
The approach described in this manual uses the
responses of organisms to detect the presence of the
toxicant during the first stages of the TIE. In this way,
the number of constituents associated with the toxi-
cants can be reduced before analyses begin and some
knowledge of physical/chemical characteristics is gained.
This approach simplifies the analytical problems and
reduces cost. Some of the problems limiting the con-
ventional approach can be used to enhance the suc-
cess of this alternate approach.
There are two main objectives in the first step of
this approach. First, characteristics of the toxicants
(e.g., solubility, volatility) must be established. This
allows them to be separated from other non-toxic con-
stituents to simplify analyses and enhance interpreta-
tion of analytical data. Secondly, throughout the TIE,
one must establish whether or not the toxicity is consis-
tently caused by the same substances. Failing to es-
tablish the variability related to the toxicants could lead
to control choices that do not correct the problem.
Knowledge of the physical/chemical characteristics
of the toxicants aids in choosing the appropriate ana-
1-3
-------
lytical method. Such information also may be useful in
selecting an effluent treatment method.
Figure 1-2 is a flowchart representation of a TRE.
This document details the toxicity characterization pro-
cedures (Phase I). Phase II (toxicant identification) and
Phase III (toxicant confirmation) usually follow Phase I.
Two other EPA manuals (EPA, 1989A; 19898) can be
consulted for more information on bench scale and pilot
plant effluent toxicity treatability studies and source
control options.
Phase I tests characterize the physical/chemical
properties of the effluent toxicant(s) using effluent ma-
nipulations and accompanying toxicity tests. Each char-
acterization test in the Phase I series is designed to
alter or render biologically unavailable a group of toxi-
cants such as oxidants, cationic metals, volatiles, non-
polar organics or chelatable metals. Aquatic toxicity
tests, performed on the effluent before and after the
individual characterization treatment, indicate the effec-
tiveness of the treatment and provide information on
the nature of the toxicant( By repeating the toxicity
characterization tests using samples of a particular
effluent collected over time, these screening tests will
provide information on whether the characteristics of
the compounds causing toxicity remain consistent. These
tests will not provide information on the variability of
toxicants within a characterization group. Knowing that
the toxicants have similar physical/chemical properties
means that they can be identified in Phase II using
similar techniques. With successful completion of Phase
I, the toxicants can be tentatively categorized as cat-
ionic metals, non-polar organics, oxidants, substances
whose toxicity is pH dependent, and others. Informa-
tion on physical/chemical characteristics of the toxi-
cants will indicate filterability, degradability, volatility,
and solubility. Either of two choices is available in the
second phase of testing, i.e., toxicant treatability or
toxicant identification studies.
Toxicant identification is described in Phase II (EPA,
1989C). Phase II involves several steps, all of which
rely on tracking the toxicity of the effluent throughout
the analytical procedure. Although effluent toxicants
are partially isolated in the first phase of the study,
further separation from other compounds present in the
effluent is usually necessary. Techniques are available
to reduce the number of compounds associated with
the toxicants. Unlike Phase I procedures, Phase II
methods will be toxicant-specific. Currently available
techniques in Phase II are for identifying non-polar
organics, EDTA chelatable metals, and ammonia.
Enough information exists now to add a section for
surfactants. Additional procedures for other toxicants
will be added as they are developed. Once the toxi-
cants have been adequately isolated from other com-
pounds in the effluent and tentatively identified as the
causative agents, final confirmation (Phase III) can be-
gin.
Like Phase I, Phase III (EPA, 19890) contains
methods generic to all toxicants. No single test pro-
vides irrefutable proof that a certain chemical is caus-
ing effluent toxicity. Rather, the combined results of the
confirmation tests are used to provide the "weight of
evidence" that the toxicant has been identified.
Once the toxicant has been identified, it can be
tracked through the process collection system using
chemical analyses. Toxicity cannot be used to find the
source for untreated wastes because toxicity from other
constituents that are toxic in untreated waste but re-
moved by treatment, will confuse the results. Of course,
using bench- or pilot-scale systems and measuring
toxicity on treated waste, is feasible.
TIEs require that toxicity be present frequently
enough and endure storage (that is, the toxicity is not
rapidly degrading) so that repeated testing can chaec-
terize and subsequently identify and confirm the toxi-
cants in Phases II and III. Therefore, enough testing
stiould be done to assure consistent presence of toxic-
ity before TIEs are initiated. This is done not to validate
a given test but to establish the sufficient and frequent
presence of toxicity.
The methods described herein are applicable pri-
marily to acute toxicity. Chronic toxicity identification
methods are being developed (EPA, 1991 A). In some
special cases in which toxicity can be concentrated (as
in the non-polar organic section of Phase II) one may
be able to "convert" chronic toxicity to acute toxicity by
concentration and successfully identify what is causing
the chronic toxicity.
To be successful, TIEs must be conducted by
multidisciplinary teams whose members must interact
daily so that toxicologists and chemists are aware of
the many concerns that affect test results. Speed is
usually important because effluents may decay during
storage. Often subsequent tests cannot be designed
until the results of the previous ones are known. Obvi-
ously then, waiting a week for analytical or toxicological
results may preclude more work while the effluent
sample undergoes changes during the waiting period. If
this happens, one must begin again on a new sample
in which case resources are not being used effectively.
1-4
-------
Effluent Sample
IDENTIFY TOXICANT(S)
Phase
Toxicant Characterization Tests
TREATABILITY
Treatabihty Approach
or Identify Toxicant
APPROACH
Phase
Toxicant Identification Analyses
Phase
Toxicant Confirmation Procedures
Based on Site Specific
Considerations
Source
Investigation
Toxicity Treatability
Evaluations
Control Method Selection
and Implementation
i
Post Control Monitoring
Figure 1-2. Flow chart for toxicity reduction evaluations.
1-5
-------
Section 2
Health and Safety
Working with effluents of unknown composition is
the nature of toxicity identification evaluations. There-
fore safety measures must be adequate for a wide
spectrum of chemicals as well as biological agents.
From the type of treatment used one may be able to
judge probable concerns. For example, extended aera-
tion is likely to minimize the presence of volatile chemi-
cals and chlorinated effluents are less likely to contain
viable pathogens.
Exposure to the wastewater during collection and
its use in the laboratory should be kept at a minimum.
Inhalation and dermal adsorption can be reduced by
wearing rubber gloves, laboratory aprons or coats, safety.
glasses, and respirators, and by using laboratory hoods.
Further guidance on health and safety for toxicity test-
ing is described in Walters and Jameson (1984).
In addition to taking precautions with effluent
samples, a number of the reagents that might be used
during Phase II toxicant identification and Phase III
toxicant confirmation studies are known or suspected
to be very toxie to humans. Analysts should familiarize
themselves with safe handling procedures for these
chemicals (DHEW, 1977; OSHA, 1976). Use of these
compounds may also necessitate specific waste dis-
posal practices.
2-1
-------
Section 3
Quality Assurance
Quality assurance is composed of two aspects,
quality verification and quality control. Quality verifica-
tion entails a demonstration that the proposed study
plan was followed as detailed and that work carried out
was properly documented. Some of the aspects of
quality verification include chain of custody procedures,
statements on the objective of the study and what is
known about the problem at its outset, instrumental log
books, and work assignments. This aspect of quality
assurance ensures that a "paper trail" is created to
prove that the work plan has been covered completely.
The quality control aspect of quality assurance involves
the procedures which take place such as the number of
samples to be taken and the mode of collection, stan-
dard operating procedures for analyses, and spiking
protocols.
No set quality assurance program can be dictated
for a TIE; the formula to a successful study will be
unique to each situation. However, adherence to some
general guidelines in formulating a Quality Assurance
Plan (GAP) may increase the probability of success.
In preparing a GAP, enough detail should be in-
cluded so that any investigator with an appropriate
background could take over the study at any time.
Cross checking of results and procedures should be
built into the program to the extent possible. Records
should be of a quality that can be offered as evidence
in court. Generally, the GAP should be provided in a
narrative form that encourages users to think about
quality assurance. To be effective, the GAP must be
more than a paper exercise simply restating standard
operating procedures (SOPs). It must increase commu-
nication between clients, program planners, field and
laboratory personnel and data analysts. The GAP must
make clear the specific responsibilities of each indi-
vidual. The larger the staff, the more important this
becomes. While QAPs may seem to be 'an inconve-
nience, the amount of effort they require is commensu-
rate with the benefits derived.
3.1 TIE Quality Control Plans
A successful TIE is dependent upon a strong qual-
ity control program. Obtaining quality TIE data is more
difficult because the constituents are unknown in con-
trast to quality control procedures for a standard ana-
lytical method for a specific chemical. In such an analy-
sis, one knows the characteristics of the analyte and
the implications of the analytical procedure being uti-
lized. Without knowledge of the physical/chemical char-
acteristics of the analyte, however, the impact of vari-
ous analytical procedures on the compound in question
is not known. Further, quality control procedures are
specific to each compound; quality control procedures
appropriate to one analyte may be completely inappro-
priate to another.
The problem of quality control is further exagger-
ated because quality control procedures for aquatic
toxicity tests may be radically different from those re-
quired for individual chemical analyses. This additional
dimension to quality control requires a unique frame-
work of checks and controls to be successful. The
impacts of chemical analytical procedures on sample
toxicity must be included. Likewise, procedures used to
insure quality toxicity test results should not impact
chemical analyses. For example, in performing stan-
dard aquatic toxicity tests, samples with low dissolved
oxygen (DO) are usually aerated. This practice may,
however, result in a loss of toxicity if the toxicant is
volatile or subject to oxidation.
3.2 Cosf Considerations/Concessions
The quality control practices required in any given
experiment must be weighed against the importance of
the data and decisions to be based upon it. The crucial
nature of certain data will demand stringent controls,
while quality control can be lessened in other experi-
ments having less impact on the overall outcome.
Effluent toxicant identification evaluations require a
large number of aquatic toxicity tests. The decision to
use the standard toxicity test methods described in
EPA (1985A; 19918) (involving a relatively high degree
of quality control), must be weighed against the degree
of complexity involved, the time required and the num-
ber of tests performed; all of these affect the cost of
testing. For this reason, toxicity tests used in the early
phases of the evaluation generally do not follow this
protocol, nor do they require exacting quality controls
because the data are only preliminary. Phase I, and to
a lesser extent, Phase II results are more tentative in
nature as compared to the tests performed for the
confirmation of the effluent toxicant(s) in Phase III.
The progression towards increasingly definitive re-
sults is also reflected in the use of a single species in
the initial evaluation studies and multiple species in the
later stages. The use of several species of aquatic
organisms to assure that effluent toxicity has been
reduced to acceptable levels is necessary because
species have different sensitivities to the same pollut-
-------
ant. Quality control must relate to the ultimate goal of
attaining and maintaining the designated uses of the
receiving water. For this reason, final effluent test re-
sults must be of sufficient quality to ensure ecosystem
protection. The use of dilution water for the toxicity
tests which mimics receiving water characteristics (in
hardness and pH) will help to ensure that the effluent
will remain non-toxic after being discharged into the
environment. In the instances where the effluent domi-
nates the receiving water, the dilution water should
mimic the water chemistry characteristics of the efflu-
ent. This is discussed in Section 5, Dilution Water. In
addition, it is essential that the variability in the cause
of effluent toxicity be defined during the course of the
TIE so that appropriate control actions provide a final
effluent safe for discharge.
3.3 Variability
The opportunities to retest any effluent to confirm
the quality of initial TIE results will be limited at best. In
addition to the shifting chemical and toxicological na-
ture of the discharge over time, individual effluent
samples stored in the laboratory change. Effluent con-
stituents degrade at unknown rates, as each compound
has its own rate of change. The change in a sample's
toxicity over time represents the cumulative change in
all of the constituents, plus that variation resulting from
experimental error. Some guidelines for assessing and
minimizing changes in sample chemistry and toxicity
are discussed in Sections 6 and 8. Regardless of the
precautions taken to minimize sample changes, a
sample cannot be retested with certainty that it has not
changed.
3.4 Intra-Laboratory Communication
Quality control procedures in chemistry and biology
can be quite different. For example, phthalates are a
frequent analytical contaminant requiring special pre-
cautions that are not of toxicological concern. The toxi-
cological problem presented by the zinc levels typically
associated with new glassware are of no concern to
those performing organic analyses. The difference in
glassware cleanup procedures is an example of many
differences that must be resolved. Cleaning procedures
must be established to cover the requirements of both.
Time schedules for analyses must be detailed in ad-
vance. One cannot assume compound stability; there-
fore, time delays between the biological and chemical
analysis of a sample cannot be tolerated.
3.5 Record Keeping
Throughout the TIE, record keeping is an important
aspect of quality verification. All observations, including
organism symptoms, should be documented. Details
that may seem unimportant during testing may be cru-
cial in later stages of the evaluation. Investigators must
record test results in a manner such that preconceived
notions about the effluent toxicants are not unintention-
ally reflected in the data. TIEs required by state or
federal pollution control agencies may require that some
or all records be reviewed.
3.6 Phase I Considerations
Effluent toxicity is "tracked" through Phases 1,11
and III using aquatic organisms. Such tracking is the
only way to detect where the toxicants are until their
identity is known. The organism's response must be
considered as the foundation and therefore, the toxicity
test results must be dependable. System blanks (blank
samples carried through procedures and analyses iden-
tical to those performed on the effluent sample) are
used extensively throughout the TIE in order to detect
toxic artifacts added during the effluent characterization
manipulations. With the exception of tests intended to
make the effluent more toxic, or situations in which a
known amount of artifactual toxicity has been intention-
ally added, sample manipulation should not cause the
effluent toxicity to change.
There are many sources of toxicity artifacts in Phase
I. These include: excessive ionic strength resulting from
the addition of acid and base during pH adjustment,
formation of toxic products by acids and bases, con-
taminated air or nitrogen sources, inadequate mixing of
test solutions, contaminants leached from filters, pH
probes, solid phase extraction (SPE) columns, and the
reagents added and their contaminants. The appropri-
ate toxicity data for the reagent chemicals used in
Phase I and common aquatic test organisms are pro-
vided as needed in subsequent sections of this docu-
ment.
Frequently toxic artifacts are unknowingly intro-
duced. For example, pH meters with refillable elec-
trodes can act as a source of silver which can reach
toxic levels in the solutions being measured for pH.
This is especially a problem where there is a need to
carefully maintain or track solution pH. Using pH elec-
trodes without membranes avoids the silver problem
(which can only be detected by profuse use of blanks).
Oil in air lines or from compressors is a source of
contamination. Simple aeration devices, such as those
sold for use with aquaria are better as long as caution
is taken to prevent contamination of the laboratory air
which is taken in by the pump.
Worst case blanks should be used to better ensure
that toxicity artifacts will be recognized. Test chambers
should be covered to prevent contamination by dust
and to minimize evaporation. Since small volumes are
often used, evaporation must be controlled. Plastic dis-
posable test chambers are recommended to avoid prob-
lems related to the reuse of test chambers. Cups from
the same lot should be spot-checked for toxicity.
Glassware used in various tests and analyses must
be cleaned not only for the chemical analyses but so
that toxicity is not introduced either by other contami-
nants or by residues of cleaning agents. Since the
organisms are sensitive to all chemicals at some con-
centrations, all toxic concentrations must be removed
and not just those for which analyses are being made.
Randomization techniques, careful observance of
organism exposure times and the use of organisms of
3-2
-------
approximately the same age ensure quality data. Stan-
dard reference toxicant tests should be performed with
the aquatic test species on a regular basis and control
charts should be developed (EPA, 1985A; 1991B). Dur-
ing Phase I it will not be known how much the toxicity
of the reference toxicant varies over time compared to
the toxicant( When the toxicants are known, they
should be used as the reference toxicant. Reference
toxicant tests should be performed to coincide with the
TIE testing schedule.
3.7 Phase II Considerations
In Phase II, a more detailed quality control program
is required. Interferences in toxicant analysis are for the
most part unknown initially but as toxicant identifica-
tions are made, interferences can be determined. Like-
wise instrumental response, degree of toxicant separa-
tion, and detector sensitivity can be determined as
identifications proceed.
3.8 Phase 111 Considerations
In Phase III of a TIE, the detail paid to quality
control and verification is at the maximum. This phase
of the study responds to the compromises made to
data quality in Phases I and II. For this reason, confi-
dence intervals for toxicity and chemical measurements
must be calculated. These measurements allow the
correlation between the concentration of the toxicants
and effluent toxicity to be checked for significance based
on test variability. Effluent manipulations prior to chemi-
cal analysis and toxicity testing are minimized in this
phase in an effort to decrease the chance for produc-
tion of artifacts. Field replicates to validate the precision
of the sampling techniques and laboratory replicates to
validate the precision of analyses must be included in
the Phase III quality control program. System blanks
must be provided. Calibration standards and spiked
samples must also be included in the laboratory quality
control program. Because an attempt will be made to
correlate effluent toxicity to toxicant concentration, spik-
ing experiments are important in determining recovery
for the toxicant( These procedures are feasible in
this phase of the study because the identities of the
substances being measured are known.
The toxicants being analyzed can be tested using
pure compounds, thereby alleviating the need for a
general reference toxicant. Because the test organism
also acts as an analytical detector in the correlation of
effluent toxicity with toxicant(s) concentration, changes
in the sensitivity of the test organisms must be known.
This is best achieved by using the same chemicals
identified for the reference toxicants.
3-3
-------
Section 4
Facilities and Equipment
The facilities, equipment and reagents needed to
perform an effluent TIE will depend on the phase of the
study and the characteristics of the toxicant( The
equipment required for Phase I characterization tests is
described throughout Section 8. The facility and equip-
ment needs in Phase II of the TIE will be site-specific
and will depend both on the physical/chemical charac-
teristics of the toxicants and on the choice of the Phase
II approach.
Phase I requires only basic analytical and toxicity
testing equipment which would be available in most
laboratories where toxicity tests and the usual water
chemistry analyses are performed. Phase III require-
ments are largely limited by equipment found in a
typical toxicity testing lab and equipment necessary for
the analysis of the toxicant(
Because of the equipment needs and time required
to conduct the evaluations, complete on-site effluent
TIEs using a mobile laboratory are generally not fea-
sible. Measurement of the loss of toxicity over time in
several effluent samples will provide information upon
which to base acceptable storage times. Usually, with
modern rapid sample shipment methods, off-site work
is practical. The cost of shipment is usually far less
than the cost of on-site work.
Large numbers of organisms and many tests are
needed for TIEs. Ready availability of test organisms is
important because often the test(s) needed are not
predictable. Only after the results of one experiment
are known can the next test be planned. It is probably
more economical to culture many of the test species
that might be used in TIEs than it is to purchasehhem.
A delay in testing caused by shipment time or lack of
availability of test organisms could cost far more in
work loss than it would cost to maintain cultures for
many weeks.
4-1
-------
Section 5
Dilution Water
The choice of dilution water will change with the
purpose of the tests and therefore the choice will often
be more varied in Phases I and II than in Phase III.
Particularly for some toxicant groups in Phase II, some
very unusual dilution water is recommended in order to
achieve the desired chemical conditions. Sometimes
the water may in itself be toxic! Such concepts are
foreign to conventional toxicology and rightly so, but
this is not conventional toxicology.
Much of Phase I and parts of Phase II utilize
organism tolerance and relative toxicity to accomplish
the objectives of the study. Methanol, hydrogen ion
concentration, and osmotic pressure may sometimes
be near lethal levels in order to test necessary condi-
tions. In some cases, the dilution medium may cause
complete mortality in 48 h, but the point of interest is
whether treatment causes more rapid mortality. If so,
one can say that one condition is more toxic than
another and obtain important information from the test.
The key is to run sufficient numbers of system blanks
so that the relative contribution to mortality is known
and toxicity is not attributed to an incorrect cause.
These are examples of the previous statement that
these methods "utilize tolerance and relative toxicity.*
In reality, this approach is very much like the compari-
son of the toxicity of two chemicals, A and B. If one
determines LCSOs for A and B and concludes that A is
twice as toxic as B, lethal conditions are being com-
pared in order to say this. Controls are not involved in
the LC50 calculation and high control survival does not
change the data interpretation, The same concept of
relative toxicity is used here. Chemical "A,, is the blank
and chemical "B" is the treated sample and the ques-
tion is, "which is more toxic?".
As these methods are built on tolerance (i.e., sur-
vival), chronic toxicity endpoints cannot be used and
that is why these methods are primarily intended for
acute toxicity. Obviously, if one wants to measure chronic
effects, the test organisms must be able to live long
enough to display chronic effects. Many of the pH
changes and other manipulations used in these meth-
ods do not allow sufficient survival time or health for
reproduction or growth. For chronic TIEs, more atten-
tion has to be given to acclimation, feeding and general
living conditions (EPA, 1991 A).
Many of the additives used in the Phase I manipu-
lations change the mixture of the effluent much more
than the dilution water. In general, for Phase I, any
water which is of a consistent quality and which will
support growth and reproduction of the test species is
suitable. We have found the use of a dilution water that
has a hardness similar to that of the effluent or the
receiving water to be beneficial. A variety of dilution
water choices are provided by EPA (1985A; 1989E)
and any of these may be used for TIEs.
In Phase III, where the objective is to confirm the
true cause of toxicity, where artifacts are to be ex-
cluded to the extent possible and where absolute toxic-
ity is more important than relative toxicity, practices
including choice of dilution water, must follow conven-
tional toxicological methodology. Tolerance to additives
must not be necessary in order to provide the desired
response. Attention must be given to simulation of the
dilution water into which the effluent is discharged.
Some toxicant dose response relationships may be
totally different as the water quality characteristics
change. These factors must be incorporated into Phase
III where absolute toxicity is of the utmost concern. In
Phases I and II, only relative differences are being
considered.
Perhaps a cautionary note is warranted regarding
the effects of dilution water on effluent toxicity. If high
concentrations of effluent are being tested (e.g., 80%)
the physical/chemical characteristics will resemble those
of the effluent. If low concentrations are tested (e.g.,
5%) then characteristics will resemble those of the
dilution water.
Little specific information can be given about the
selection of dilution water in Phases I and II except that
the desired tested conditions will often dictate its char-
acteristics. For example, in Section 8.6, the same col-
umn used for the blank may not be usable for the
effluent sample if receiving water is used as the dilution
water. Secondly, sufficient numbers of blanks must be
included to interpret the results. In Phase III, the choice
of the appropriate dilution water should be based on
the characteristics of the receiving water where the
discharge occurs.
5-1
-------
Section 6
Effluent Sampling and Handling
A wastewater sample may be representative only
of the discharge at the time of sampling. In effect, each
sample is a "snapshot" of the effluent's toxicological
and chemical quality over time. To determine whether
any effluent sample is typical of the wastewater may
require the collection of a large population of samples.
Further, what constitutes a "representative" sample is a
function of the parameter of concern. Because effluents
vary in composition, sampling must be extensive enough
that one is confident that the groups of samples repre-
sent the discharge over time. Guidelines for determin-
ing the number and frequency of samples required to
represent effluent quality are contained in the "Hand-
book for Sampling and Sample Preservation of Water
and Wastewater" (Berg, 1982). However, since this
guidance is not based on toxicity, it should be used
with caution.
Both quantitative (change in concentration) and
qualitative (change in toxicants) variability commonly
occur in effluents and both may affect toxicity. Changes
in effluent toxicity are the result of varying concentra-
tions of individual toxicants, different toxicants, chang-
ing water quality characteristics (affecting compound
toxicity) and analytical and toxicological error. Even if
the toxicity of an effluent to an aquatic organism is
relatively stable, this does not mean that there is only a
single toxicant causing toxicity in any given sample or
among several samples.
Determining whether a sample is typically toxic is
not as simple as comparing the conventional pollutants
of the sample to long-term effluent averages. Effluent
toxicants often do not follow the same trends as BOD,
TOC and TSS. The toxicant(s) may be present at such
a low level that it does not significantly affect the quan-
tity of the conventional pollutant, even though it is
present in toxic concentrations.
Conventional parameters, BOD, TSS, and other
pollutants limited in the facility's NPDES permit, will
provide an indication of the operational status of the
treatment system on the day of sampling. For industrial
discharges, information on production levels and types
of operating processes may be helpful. The condition of
the facility's treatment system at the time of sampling
should be determined by the individual collecting the
sample. The type of sample, time of collection, and
other general information on the facility should be re-
corded. An example of a page of a log book is given in
Figure 6-1.
Upon the arrival of the sample in the laboratory,
temperature, pH, toxicity, hardness, conductivity, total
residual chlorine (TRC), total ammonia, alkalinity, and
DO should be measured. Toxicity should be measured
periodically during storage to document any changes
(cf. Section 8).
Investigators should not be surprised to find that
well operated municipal and industrial treatment sys-
tems discharge unacceptably toxic wastewaters. Eff lu-
ent guideline-based limits which reflect best achievable
technology, do not prescribe limits for more than a few
chemicals. Many compounds present in effluents are
not regulated because the discharger is not required to
report their presence in permit applications or they
cannot be detected using typical methods for wastewa-
ter analysis.
For chlorinated effluents, whether sampling should
be done before chlorination depends on the question to
be answered. Sometimes the question may be whether
or not there are toxicants other than chlorine present.
Dechlorination prior to toxicity characterization may be
needed in order to distinguish toxicity from causes
other than chlorine. Usual methods of dechlorination
may remove more than toxicity from chlorine alone and
careful data interpretation is needed to understand the
results. Toxicity from more than one cause is often not
additive in effluents, so relative contributions from two
or more causes can be very hard to decipher.
The choice of grab or composite samples will de-
pend on the specific discharge situation, (e.g., plant
retention time) questions to be answered by the TIE
and the stage of the TIE. In Phase I testing, samples
that are very different from one another give results
that are difficult to interpret; therefore composite samples
are more similar and are easier to. use. In Phase III,
effluent variability is used to advantage; therefore, grab
samples are often best. If toxicity is low or intermittently
present, grab samples may be best during all phases.
The additional difficulty of getting flow proportional
samples should be balanced against their advantage in
each situation. While grab sampling may provide maxi-
mum effluent toxicity, it is more difficult to catch peaks
in toxicity and Phase I sampling may require more time.
EPA (1985A; 1991 B) discusses the advantages and
disadvantages of grab and composite sampling and
have also detailed methods for sampling intermittent
discharges.
6-1
-------
Figure 6-1. Example data sheet for logging in samples.
Sample Log No.:.
Date of Arrival:_
Date and Time
of Sample Collection:.
Facility:
Location:
NPDES No:.
Contact:
Phone Number.
Sampler:
Sample Type: Q Grab Q Composite
Q Glass 0 Plastic
Q Prechlorinated
Q Chlorinated
cl Dechlorinated
Sample Conditions Upon Arrival:
Temperature
pH
Total Alkalinity
Total Hardness
Conductivity/Salinity
Total Residual Chlorine.
Total Ammonia
Condition of treatment system at time of sampling:
Status of process operations/production (if applicable):
Comments:
6-2
-------
If the TIE analyses are not conducted on-site,
samples must be shipped on ice to the testing location.
Effluent samples should not be filtered prior to testing
unless it is necessary to remove other organisms.
Sample filtration could affect the results of the charac-
terization tests, one of which entails filtering the efflu-
ent. Sample aeration should also be minimized during
collection and transfer. Initial sample analysis should
begin as soon as practical after effluent sampling. Phase
II and especially Phase III may require specific types of
sample containers or the addition of preservative to
aliquots of sample designated for chemical analyses.
For a single Phase I test series, 3 L of effluent are
needed for analysis if test organisms such as daphnids
or newly hatched fathead minnows are used. The exact
volume required depends on the toxicity of the effluent
and to a lesser extent, the test options chosen (cf.,
Section 8). For other species different volumes may be
necessary. Volumes frequently used in each character-
ization test are supplied in Table 6-1.
The extent of the analyses carried out on any
individual sample must be weighed against the cost of
Table 6-1. Volumes needed for Phase I tests
Characterization Step
Volume for Total
Each Step1 Volumes2 (m L)
Chemical analyses3
pH 3 Adjustment
filtration
solid phase extraction
aeration
pH 11 Adjustment
filtration
solid phase extraction'
aeration
--
30
235
200
35
30
235
200
35
<500
-300
-300
Unadjusted pH effluent (pH()5
initial test 40
baseline test 80
filtration 235
solid phase extraction 200
aeration 35
EDTA additions 100
sodium thiosulfate additions 100
-590
Graduated
pH6
pH7
pHs
pH,
40-500
40-500
40-500
-120-1000
Amount is dependent on effluent characteristics.
Total volume is -3 L; this is maximum needed, does not include
subsequent testing.
These include temperature, pH, hardness, conductivity, TRC,
total ammonia, alkalinity, and DO.
The pH is readjusted to pH 9 before it is put through the C]8 SPE
column.
The pH/of the effluent is the initial pH of the effluent sample. It
may be important to know the pH at the point of discharge as
well as the receiving water pH and to know the pH of the
effluent at air equilibrium.
additional sampling, the stability of the sample, sample
representativeness and the need to have samples of
different toxicity. Clearly, the resources required for
such TIEs are too great to expend on a single sample
or on a few samples which do not represent the efflu-
ent. Likewise, there is not a set number of samples
which should be analyzed in Phases I, II or III before
going on to subsequent phases of the study or taking
final measures to control effluent toxicity. The number
of samples analyzed in each phase will be a function of
the apparent variability in the effluent, the number of
toxicants, how persuasive the data are, the cost of the
remedial action, regulatory deadlines and finally, the
success of each study phase.
6.7 Sample Shipment and Collection in
Plastic versus Glass
Effluent samples often have been collected, shipped
and stored in various types of plastic (e.g., polyethyl-
ene) containers rather than glass. However, with a few
effluents, we have noted that samples shipped and
stored in glass were more toxic and retained their
toxicity longer than split samples shipped and stored in
plastic. This effect appeared to be due to adsorption of
certain types of toxicants (e.g., surfactants) to thejolas-
tic. For these instances the samples in glass were
more representative of the effluent, and thus for TIE
purposes were preferable to the samples in plastic.
An easy way to check whether or not there is a
difference in the toxicity of samples shipped and stored
in glass containers versus those shipped and stored in
plastic containers, is to test two or three sets of effluent
samples. Effluent should be collected in glass or stain-
less steel, then a portion shipped in glass and another
portion shipped in plastic. Baseline toxicity tests (cf.,
Section 8) are conducted on each, perhaps on days 4
and 7 after receipt. If the initial toxicity of the sample is
similar for both the plastic and the glass containers,
and the toxicity for samples from the two containers is
similar over time (i.e., over storage time), it is appropri-
ate to have the effluent samples shipped and stored in
plastic containers. However, if effluent shipped and
stored in glass appears to be more toxic, and retains
the toxicity longer than the effluent sample shipped and
stored in plastic, glass containers should be used for all
shipments and storage for that particular effluent. These
same considerations also apply to the sampling/collect-
ing equipment. Collection, shipment, and storage of
effluent samples in glass may involve more effort than
plastic containers. The use of glass containers for
samples that retain their toxicity longer might result in
more rapid and cost effective progress through the TIE
because fewer samples might be required for identifica-
tion of effluent(s). Since only certain classes of com-
pounds are expected to adsorb to plastic containers
(e.g., surfactants), if the effluent is more toxic in glass,
this can be a useful piece of information for characteriz-
ing the toxicants.
6-3
-------
Section 7
Toxicity Tests
7.1 Principles
Acute lethality tests with aquatic organisms are
utilized throughout the toxicity characterization proce-
dures described in this manual as well as in Phases II
and III. Using toxicity for such evaluations is logical
since toxicity triggers the TIE requirement. In these
tests the organism acts as the "detector" for chemicals
causing effluent toxicity. As such, they provide the true
response regardless of the outcome of other analyses.
The toxicity test is the only analytical procedure that
can be used to measure toxicity. Until the cause of
toxicity is known, chemical methods cannot be used to
identify and quantify the toxicants.
There are a number of consequences associated
with this reliance on toxicity. The organism responds to
every constituent, provided that it is present above a
threshold level either individually or collectively if the
constituents are additive. While this general response
to any compound presents an advantage as a broad
spectrum test for toxicants, it requires considerable
effort to determine the primary cause of toxicity be-
cause it is not specific. This non-specific response
necessitates a generic chemical/physical characteriza-
tion of toxicants during Phase I testing before Phase II
identification is begun.
A further repercussion of this universal response is
the probability of artifactual toxicity. Because the ana-
lyst is reliant upon the organism's ability to track toxicity
throughout the effluent characterization steps, sample
manipulations are constrained. While characterizing the
effluent, no manipulation should change the toxicity of
the sample in an unpredictable manner. "Toxicity-blanks
and controls" are helpful but the difficulties associated
with them are far greater than those connected with
chemical analyses because of their non-specificity. As
a result many more blanks are employed in TIE testing
than in chemical analyses or standard toxicity testing.
Negative blank toxicity cannot be assumed regardless
of past results. Quite unexpected sources of artifactual
toxicity will occur in the course of conducting an evalu-
ation.
For some Phase I tests the corresponding blanks
(treatments on the dilution water) do not provide com-
pletely relevant information concerning the effect of the
manipulation on the effluent. For example, blanks of
the graduated pH test (Section 8.9) are not particularly
useful whether the pH is adjusted with acids, bases, or
CO,. The amount of the acid or CO, used to adjust and
maintain the same pH for an effluent sample and a
blank are often radically different due to the differences
in the buffering capacity of each of the solutions. Since
the matrix of the effluent and dilution water are differ-
ent, the pH in each solution will change at different
rates during the toxicity tests. Therefore the blanks are
not representative of what is occurring in the effluent
test and the controls exposure does not provide infor-
mation on the manipulation effect on the test organ-
isms. The use of blanks for the other manipulation
steps is relevant, and they provide information on clean-
liness of the acids and bases added, the air system,
filter apparatus, and SPE columns.
7.2 Test Species
Just as different analytical methods have different
detection levels for the same chemical, different spe-
cies have different sensitivities to the same toxicants.
The major difference is that the toxicity measurement is
non-specific to chemicals and so for an unknown mix-
ture (effluent, sediment pore water) one must deter-
mine whether a different toxicity value for the sample is
caused by the organisms different sensitivity to the
same toxicant or to different toxicants.
The choice of species to use for the toxicity test
can change the conclusion reached. In addition to the
obvious need to use species of an appropriate size,
age, availability, and adaptability to test conditions,
there are other important considerations. An effluent
toxic to two species, having equal or different LCSOs
may be toxic because of different toxicants. Differences
of 1,000x in sensitivity are common and differences of
10,000x occur among species exposed to a single
chemical. Anyone involved in identifying the cause of
toxicity of an effluent will be concerned because some-
one has found the effluent toxic to some organism. If
that is not the case, before a TIE is begun, one should
determine to which organisms the toxicity concern is
directed.
Many effluents will be received for TIEs because
they have been found toxic to the cladocerans,
Ceriodaphnia or Daphnia-specles well suited to TIE
methods. TIE test species selection is obvious in these
instances. Where toxicity concern is based on species
(trout or mysid shrimp), that are not going to be the TIE
test species, one must demonstrate that the toxicity of
concern has the same cause as the toxicity manifested
by the species to be used in the TIE. The difficulty
depends on the effluent characteristics (especially tox-
7-1
-------
icity variability), the number of TIE steps which affect
toxicity and the difference in sensitivity between the
species being compared. Since this problem has not
been one we have experienced frequently, our sugges-
tions are certainly not all-inclusive. The final confirma-
tion (Phase III) methods are designed to show whether
the wrong toxicant was identified. However, many re-
sources may be consumed before reaching that stage
and earlier assurances should be obtained if reason-
able, to save time and cost.
One approach is to compare the LC50 values of
whole, unaltered effluent samples for the species origi-
nally raising the toxicity concern and the selected spe-
cies for the TIE. If the acute toxicity varies similarly for
each species among samples then there is evidence
that the two species are responding to the same
toxicant( If the LC50 values vary differently for the
two species, there is evidence that the toxicants are
different. If the LC50 values among samples do not
vary more than the precision of the test method this
approach is useless for that effluent. Successful appli-
cation of this approach does not require equal sensitiv-
ity of the two species or the greatest toxicity of the TIE
organism, but rather sensitivity of the species to the
same toxicant.
If in Phase I, several steps (e.g., pH decrease,
aeration, solid phase extraction) all changed toxicity,
and if the direction and relative magnitude of change
was the same for both test species, then there is
evidence that both are sensitive to the same toxicant. If
one or more parameters are different, the evidence is
strong that the toxicants are different. This is not to say
that if a Phase I technique completely removes toxicity
to one species, it will remove it to the same extent for
the other species. Because different species have dis-
similar sensitivities to the same chemical, removal of
90% of a compound in an effluent sample may lead to
a non-toxic concentration to one species while only
reducing the toxicity to another species. If the Phase I
procedures that successfully remove or reduce effluent
toxicity differ by test species, it is unlikely that toxicity is
caused by the same chemical(s).
Symptom comparisons are useful, especially if one
is comparing similar organisms. Comparing fish symp-
toms to Daphnia symptoms could be very misleading
but comparing symptoms of Daphnia magna to those of
Ceriodaphnia dubia should be relatively safe. If one
finds comparable symptoms, the evidence is not con-
vincing because many toxicants cause specific symp-
toms but if symptoms are distinctly different, the evi-
dence is strong that the toxicants are different. This is
true only when symptoms are compared at effluent
concentrations that are the same multiple of the LC50
for each species. For example, if two species have
LC50 values of 10% and 90%, comparing symptoms at
100% concentrations could be misleading. At 100%
effluent, the species with an LC50 of 10% might experi-
ence the symptoms so fast that their sensitivity would
appear completely different from those of the less sen-
sitive species. Experience will reveal additional tech-
niques that can be used.
Freshwater discharges to saline receiving water
require separate considerations. Sea salts can be added
to raise the salinity of the effluent (EPA, 1991 B) enough
so that marine species can be used in the TIE. How-
ever, the tolerances of marine organisms to the addi-
tives and effluent manipulations have not been deter-
mined. To do so is costly and time consuming and a
more efficient method may be to use a freshwater
species in Phase I and II. If this is done, data must be
gathered to show that the freshwater species chosen is
sufficiently sensitive and is responding to the same
toxicant(s) as the marine species. The principles of
doing this are the same as described above for differ-
ent freshwater species. When Phase III is reached,
marine species should be used, but in that phase,
manipulations and additives are minimal and little ancil-
lary data are needed in order to use marine species.
For discharges with conductivities comparable to
brackish or marine water, caution is in order. Most
methods for measuring "salinity" (conductivity or refrac-
tion) are non-specific for NaCI, which is the principal
component of sea water. Marine organisms accomplish
osmotic regulation by regulating sodium and chloride. If
salinity of an effluent is not caused by NaCI, marine
species may be stressed as much as freshwater spe-
cies by high concentrations of other dissolved salts.
Unless the "salinity" of an effluent is known to be
caused by NaCI, marine species cannot be used to
avoid the salinity effects.
7.3 Toxicity Test Procedures
The purpose of the toxicity test in Phase I is the
same as that of any analytical method-to measure
(detect) the presence of the toxicants. This use is quite
different than conventional toxicity testing where the
objective is to accurately and quantitatively measure
the sensitivity of the organism to known concentrations
of a chemical or effluent. For this latter purpose, remov-
ing stress (e.g., low DO) or other contaminants, and
lack of space is important because such stresses may
change the sensitivity of the organism to the contami-
nant of concern. In Phase I, relative sensitivity is used;
that is, we compare whether one condition is more or
less toxic than another but both may be toxic. There-
fore, concern of documenting and/or removing other
stresses is not very important. It is important to be sure
that these other stresses are similar for each condition
being compared, each time the manipulation and sub-
sequent toxicity tests are performed.
The reason for this discussion under test methods
is that effort must be made to make the tests used in
Phase I as inexpensive as possible, because for some
effluents, large numbers of tests may be needed. For
example, we have used more than 100 tests on some
effluents in Phase I. If the effort usually expended in
measuring all the required water chemistries for a whole
effluent test (EPA, 1985A) had been done for these
7-2
-------
tests, the cost would have been prohibitive. The reader
may wonder whether data collected from such tests
can be trusted. Confidence in the data hinges on care-
ful assurance that the stresses are similar among com-
parisons. For example, it does not matter if the test
organisms are acclimated to a pH change. It does
matter that stress from lack of acclimation to pH change
occurs in each treatment compared.
Sometimes, in order to achieve desired chemical
conditions, the stress from pH change cannot be made
uniform. In these situations, only gross differences in
response may be dependable. In some cases, errone-
ous conclusions will be reached. While these may cause
wasted effort, the error should be found in Phase III.
That is why, in Phase III, careful quality control must be
exercised and cost saving shortcuts are not acceptable
because one of the purposes of Phase III is to catch
errors or artifacts that may occur in Phases I and II.
One need not use the standard acute methods
(EPA, 1985A; 1991 A) in Phase I for these reasons. The
following mechanics of performing an acute test with
cladocerans and newly hatched fathead minnows have
been found by experience to be very cost effective and
are offered as an aid to those doing Phase I testing.
Specific volumes and sizes are used in this example for
simplicity, but of course, these are varied depending on
each test purpose.
For example, arrange a set of 12 plastic
cups into six pairs. Fill 10 cups with 10 mL
of dilution water using a disposable pipette.
Add 10 mL of effluent to the two empty
cups to make the high concentration, (e.g.,
100%). Add 10 mL of effluent to the next
pair of test cups (duplicates labeled A and
B) already containing 10 mL each of the
dilution water (.Figure 7-1). The resulting
concentration is 50%. From each cup of
the 50% solution, transfer 10 mL to the
third pair of test cups to produce the 25%
concentration. Continue this process until
sufficient exposure concentrations have
been prepared. One pair of cups in the
series contains only dilution water and
serves as the control. Mixing the solutions
prior to the transfer of each aliquot is very
important. This can be accomplished by
drawing the solution into the pipette and
discharging it back into the cup several
times prior to transfer. Additional mixing of
test solutions should be done for
experiments in which reagents such as
sodium thiosulfate (NazSzO) and EDTA
(Phase I), or effluent metnanol eluate
concentrates (Phase I and II) are added to
effluent or dilution water.
The need for duplicates will depend on the accu-
racy and precision required of the test results. Tests
requiring a measure of accuracy in the form of confi-
dence intervals (CIs) should be run in duplicate. Tests
designed to provide only' an indication of positive or
negative toxicity need not be run in duplicate. Beyond
the initial and baseline effluent toxicity tests (Sections
8.1 and 8.2) which are designed to define effluent
toxicity upon arrival in the laboratory and periodically
during the TIE with each effluent sample testing, re-
spectively, Phase I toxicity tests usually do not require
duplicates.
The test organisms of uniform age should be placed
at random in each test cup to insure valid results.
Because the volume of test solution may be small, care
must be taken to minimize the volume added during
test organism transfer. If the volume of water trans-
ferred with the organism is reduced to a drop (50 ^L),
only five organisms are added to the test chamber and
a 10 mL test volume is used, the resulting change in
test solution volume will be 2.5%. Minimizing the change
in volume is more critical as test solution volume is
reduced. This is particularly important in the Phase II
experiments, when limited volumes of effluent fraction
concentrates are available. Care should also be taken
to avoid chemical contamination between concentra-
tions when test animals are being added.
We have stressed a relaxation of the usual water
chemistry requirements in these Phase I tests because
they are not as necessary here as they are in Phdseill.
However, sometimes, in order to maintain the desired
conditions in the test (such as maintaining a specific
pH) frequent specific repetitive measurements of those
items will be necessary. The distinction drawn here is
to avoid measurements you don't need (e.g., sample
hardness) and concentrate on those that are important
(e.g., pH). Effluents are often well buffered and pH
sometimes will change quickly if equilibrium is not al-
ready established. POTW eifluents are not in air equi-
librium when discharged and as soon as they are
exposed to air, the pH will rise. A typical POTW effluent
pH is 7.2-7.4 when discharged but it will equilibrate
after contact with air and may stabilize at 8.2-8.5. If pH
is important to test interpretation, pH must be moni-
tored throughout the test. It will also be important to
decide what the initial pH (pH i) of the effluent is since
the pH at the discharge and/or the initial pH may be
different from the pH of the effluent at air equilibrium.
7.4 Test Endpoints
Little effort should be expended in calculating LC50
values for Phase I toxicity tests. There is no need to
apply sophisticated and complex programs to the test
results. Several methods for estimating the LC50 from
the acute toxicity data are described in EPA (1985A),
however a method which is most easily and quickly
applied to the data should be used. In many cases, the
graphical method entailing interpolation may prove to
be the most convenient. Differences resulting from the
choice of data analysis method should not impair the
outcome of Phase I studies. Phase III tests may require
more sophisticated analyses.
Toxic units (TU) have a special utility in some parts
of a TIE. The TU of whole effluent is 100% divided by
the LC50 of the effluent. For specific chemicals the TU
7-3
-------
Figure 7-1. Schematic for preparing effluent test concentrations using simple dilution techniques. Two replicates are used for initial and baseline whole
effluent toxicity test.
Effluent
• Add 10 mL to cups A and B
for the high concentration.
• Add 10 ml to next A and B cups
for the second high concentration.
Add 10 ml to each
replicate except in
the high concentration
Dilution
Water
Serial Dilutions
Waste
-------
is equal to the concentration of the compound present
in the effluent divided by the LC50 of the compound
(EPA, 1991 B). For example, if the LC50 of an effluent
is 25%, the effluent contains 4 TU (100/25). If the 48-h
LC50 of compound A is 3 mg/L, a solution of 1 mg/L of
this compound contains 0.33 TU. By normalizing the
concentration term (such as the LC50) to a unit of
toxicity, the TU allows the toxicity of effluents and/or
chemicals to be "summed," provided that the test length
and species used are the same in every test. This
cannot be done using LCSOs because chemicals and
effluents each have different toxicity, and different con-
centrations each equal one LC50. Phase III contains
more discussion about adding TUs; however one must
be cautious in summing them. Unless it is known that
the toxicants are strictly additive, simple summation of
TUs will be incorrect.
7.5 Feeding
Most species used in acute tests are not fed during
the test. However, the acute effluent manual (EPA,
1991 B) has modified the effluent tests to allow clado-
cerans to be fed before test initiation. We routinely add
food to all test waters (this includes the 100% effluent)
for all Ceriodaphnia and Daphnia tests but only at the
initiation of each test. This practice is standard in Phases
I, II, and III. However, the decision to feed will be
species specific and dependent on the characteristics
of the effluent. Consistency throughout each phase of
the TIE is most important. All tolerance data for
Ceriodaphnia given in Section 8 are based on tests in
which animals were fed the yeast-cerophyll-trout food
(YCT) mixture (EPA, 1989E; EPA, 1991C). The amount
of YCT added was 66 pL of YCT per 10 ml_ and 5
animals.
7.6 Multiple Species
A useful technique is to test two species together in
the same test chamber (e.g., 1 oz. plastic cup). This is
very beneficial in the initial toxicity test in order to select
the most sensitive species for the Phase I tests or in
situations where two species appear to be responding
to toxicity of the effluent differently. This type of test
also can be useful when conditions in tests with differ-
ent organisms vary independently. For example, testing
C. dubia and fathead minnows <48 h old) together
under the same pH conditions is very, helpful in evaluat-
ing the role of ammonia in an effluent's toxicity. By
testing the species together, the experimental condi-
tions may change but both species experience identical
fluctuations. We have tested the following sets of spe-
cies together: C. dubia and fathead minnows, C. dubia
and D. magna, and C. dubia and D. pulex. "
7-5
-------
Section 8
Phase I Toxicity Characterization Tests
The first phase of a TIE involves characterization of
the toxic effluent. The characterization information gath-
ered in Phase I forms the basis and direction for Phase
II identification of the specific toxicants or may be
useful for treatability evaluations. In Phase I, simple
manipulations for toxicity removal or alteration are per-
formed on the whole effluent. Acute toxicity tests utiliz-
ing aquatic organisms are used to determine whether
the toxic chemicals have certain physical or chemical
characteristics. Two objectives are accomplished dur-
ing the toxicity characterization phase: a) the physical
and chemical characteristics of the toxicant(s) are
broadly defined and b) some information is gathered to
indicate whether the toxicants are similar in effluent
samples taken over time. Several patterns of Phase I
results are indicative of certain toxicants (See Section
9.4) but otherwise Phase I only provides evidence of
characteristics of groups of chemicals that may be the
toxicants. This information can subsequently be used in
the second phase of the study, either in the develop-
ment of bench-scale wastewater treatment processes
(EPA, 1989A; 19898) or in choosing separation and
analytical procedures for toxicant identification as de-
scribed in Phase II.
The tests described in this section are designed
primarily for acutely toxic effluents. Methods for chronic
toxicity are being developed (EPA, 1991 A). The meth-
ods in this section are based on the use of small test
organisms such as daphnids (Ceriodaphnia, Daphnia)
and newly hatched fish (fathead minnows). If larger
species are used, modifications to these methods will
have to be made.
Analysis of samples should begin as soon as prac-
tical following collection. Until experience is gained with
the effluent, there is no way to predict how long samples
can be stored before substantial changes in toxicity
occur. In transit and in the laboratory, the bulk effluent
should be held below 4°C and kept headspace free.
Minimizing the headspace for samples shipped in glass
is not practical. Once in the laboratory, testing on indi-
vidual samples of each effluent may continue indefi-
nitely, provided that whole effluent toxicity stabilizes.
The degree of toxicity can remain similar, while the
cause of toxicity may change with age. Especially in
the early stages of the TIE, fresh samples should be
used regardless of toxicant stability. The degree to
which any single sample is analyzed should be weighed
against the cost of the analyses and the probability that
the sample is an adequate representation of typical
effluent. Obviously, when several samples show that a
single class of compounds is responsible for effluent
toxicity, Phase II procedures should be initiated.
Each of the characterization tests described in Sec-
tion 8 is designed to change the toxicity of groups of
constituents (Figure 8-I). Toxicity before and after the
characterization treatment will indicate for which groups
the toxicity was changed. All but one (initial toxicity
test) of the characterization tests is performed at the
same time in order to minimize confounding effects
resulting from degradation of sample toxicity over time.
While it is not critical that each characterization ma-
nipulation be performed at exactly the same time, the
toxicity tests should be initiated at approximately the
same time. If more than one species is used, the
Phase I results must be interpreted separately for each
because at this stage one cannot tell whether the same
toxicant(s) is involved for all species.
Following receipt of the effluent sample, various
steps to initiate Phase I are done (Table 8-I). Day 1 is
when the sample arrives in the laboratory. On day 1,
initial routine chemical measurements are taken for the
effluent sample and an initial toxicity test is started on
an aliquot of the sample. This LC50 is used to set the
desired exposure concentrations for subsequent Phase
I toxicity tests and is referred to as the "initial" toxicity
test to distinguish it from the "baseline" toxicity test
described below. Other aliquots of the sample are ad-
justed to pH 3 and 11, filtered, aerated and/or
chromatographed using a C18 SPE column. Following
these manipulations, each effluent aliquot is readjusted
to the initial pH(pH/) of the effluent. By pH/, we
generally refer to the pH of the effluent at arrival in the
laboratory, which may or may not be the pH of the
effluent at air equilibrium. These aliquots and the re-
mainder of the effluent are then covered to minimize
evaporation and held at 4°C overnight. However, upon
warming the solutions, supersaturation from dissolved
gases might occur. If the test organism to be used is
sensitive to supersaturation, then the supersaturation
must be removed. Generally, Ceriodaphnia are not
very "sensitive" to such situations, unlike newly hatched
fathead minnows.
Delaying the majority of the toxicity testing until the
next day (day 2) allows the test exposures to be set at
concentrations bracketing the 24-h LC50 of the day 1
initial toxicity test. This procedure also allows pH ad-
justed effluent aliquots more time to stabilize, and addi-
tional pH adjustments can be made as necessary.
-------
Figure 8-1. Overview of Phase I effluent characterization tests. (Note: pH/stands for initial pH.)
Tf
r 1 V.
initial Toxicity Test f
(uayij Baseline Toxicity
Test (Day 2)
Aeration Tests ^.
(Day 2)
*
* * *
Acid pH / Base
)xic Effluent Sa
i
i
t
Filtration Tests
(Day 2)
i
i
i
\
i
mnlp
1 1 I|JIC
t
EDTA
Chelation
Test (Day 2)
GIB Solid Phase
— *- Extraction Tests
(Day 2)
*
i *
Acid pH / I
pH Adjustment Gr
Tests (Day 2) Te
t t
{
Oxidant
Reduction
Test (Day 2)
~~t
3ase
t
aduated pH
jsts (Day 2)
t
f
-------
Table 8-1.
Description
effluent manipulations
Section
6.0
DAY 1 SAMPLE ARRIVAL:
Chemical analyses
.PH
'conductivity
"total residual chlorine (TRC)
"hardness
"temperature
"total ammonia
"dissolved oxygen (DO)
"alkalinity
Initial toxicity test
Sample Manipulations':
• pH adjustment (pH 3, pH /, pH 11)
• pH adjustment/filtration
• pH adjustment/aeration
• pH adjustment/C,, solid phase extraction
DAY 2 TOXICITY TESTING:
Warm effluent samples from day 1 and set-up
toxicity tests'
"baseline toxicity
• pH adjustment samples
"filtration samples
"aeration samples
^ C,, solid phase extraction samples
"sodium thiosulfate addition samples
• EDTA addition samples
"graduated pH samples
Read 24-h mortality on initial toxicity test
DAYS 3 AND 4 MONITORING TESTS:
Read 48 h mortality initial toxicity test
Read 24 h and 48 h mortality on tests from day 2
8.1
8.3-8.9
8.3
8.4
8.5
8.6
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.1
These manipulations and toxicity tests can be performed on day
2 after the presence of toxicity has been confirmed; see text for
details.
However, the manipulations can be performed and the
test initiated the next day (day 2) rather than on sample
arrival. This is useful when the toxicity of the effluent is
unknown, and prevents conducting a Phase I on non-
toxic samples. It is important that sufficient time is
allowed so that the pH adjusted samples can stabilize
at the pH /'. The following sections assume the manipu-
lations were made on day 1.
On the second day the aliquots of whole and
manipulated effluent prepared on day 1 are diluted to
4x-, 2x-, 1x-, and 0.5x the 24-h LC50 of the effluent and
subsequently tested for toxicity. This dilution series is
used so that for highly toxic effluents, smaller changes
in toxicity can be detected than would be the case if
100% effluent was used. (See Section 9 regarding
multiple toxicants and this dilution series.) A whole
effluent toxicity test is begun using unaltered effluent,
now 24 h old. The result of this test (and subsequent
whole effluent tests) is referred to as the "baseline"
effluent LC50. Other toxicity tests involving the addition
of chelating or reducing agents and less severe pH
adjustments are also conducted. For the EDTA addition
test, the time needed for the EDTA to complex any
metals present may be a function of the matrix of the
effluent. Therefore, the addition of EDTA should be
made first on day 2 and the sample held until all other
manipulations are complete before introducing the test
organisms (see Section 8.8 for details on EDTA test).
For one complete "Phase I" of a TIE as described
in this section, there are nine categories of toxicity tests
that are conducted. These are as follows: initial toxicity
test, baseline test, pH adjustment test, pH adjustment/
filtration test, pH adjustment/aeration tests, pH adjust-
ment/C,, SPE test, EDTA addition test, sodium thiosul-
fate addition test, and graduated pH test. Toxicity test
results are read on subsequent testing days and de-
pending on the outcome of the Phase I test series,
additional toxicity tests designed to further define or
confirm the nature of the toxicants are conducted.
For an experienced analyst the amount of time
required to conduct the sample manipulation tasks
scheduled for day 1 is about half of one day. If at 24 h,
less than 50% mortality of test organisms exposed to
the 100% day 1 effluent has occurred, the samp^ can
be discarded and a new sample collected with rela-
tively little loss of resources or time. For this reason,
waiting to perform the manipulations on day 2 is useful.
Alternatively, the test can be continued to 48, 72 or
96 h at which time the effluent may produce an LC50.
In such cases, the baseline toxicity tests prepared on
the second day (day 2) following sample arrival are set
up at exposure levels of 1 00%, 50%, 25%, 12.5%,
6.25% effluent.
For a highly toxic effluent sample with rapidly de-
gradable toxicants, it may be prudent to override the
use of 4 x -24-h LC50 treatment level and opt for con-
ducting the Phase I using 100% effluent. These rapidly
degradable compounds will be discovered only through
periodic testing as the sample ages.
Several Phase I characterization tests are relatively
broad in scope, intended to include more than one
class of toxicant. Therefore, if a significant change in
effluent toxicity is seen following these characterization
procedures, additional tests are needed to further delin-
eate the nature of the toxicity. The amount of testing
beyond the initial characterization of the sample will
depend on the stability of effluent toxicity, the nature of
the toxicity, and previous Phase I results for the effluent
(i.e., observed trends in the nature of the toxicity). A
"significant reduction" in toxicity between aliquots of the
day 2 whole effluent (baseline LC50) and treated efflu-
ent must be decided based upon the laboratory's test
precision. Usually a change in the LC50 equal to one
concentration interval can be considered significant but
when precision is good smaller differences can be
used. This suggestion is arbitrary and should not re-
place good judgement and experience. None of these
tests by themselves are conclusive, so the danger of
type I or type II errors is not great. Experience has
8-3
-------
shown that for many effluents, at least one Phase I
characterization test will be successful in substantially
altering effluent toxicity. If not all toxicity is removed,
other groups of toxicants (not addressed by Phase I
procedures) may be present in the effluent or a single
toxicant may be present in the effluent at such high
concentrations that only partial toxicity removal is
achieved. Additional testing to resolve these findings
involves applying the successful Phase I test at a
higher level (i.e., increased degradation time, increased
aeration, larger C,§ SPE column volume, increased
reagent concentrations).
Another outcome of the Phase I characterization
test series may be that several tests succeed in par-
tially removing effluent toxicity. In this situation, one
may be dealing with several toxicants, each with differ-
ent physical/chemical characteristics, or a single toxi-
cant of such a nature as to be removed by more than
one Phase I test. These results may be resolved by
treating a single aliquot of the sample with all of the
characterization tests that significantly reduced the
baseline toxicity of the effluent. If effluent toxicity re-
moval is enhanced as compared to the reduction pro-
vided by individual characterization tests, the sample
may contain more than one type of toxicant. If the final
toxicity removal at the end of the series of characteriza-
tion tests is approximately the same as that provided
by the most efficient single Phase I test, then it is likely
that all of the test methods involved are successful in
reducing the same toxicant to varying extents. This
outcome is also suggested when one or more Phase I
tests completely remove toxicity while some number of
other tests partially reduce toxicity. Phase I tests over-
lap somewhat in their abilities to remove groups of
toxicants. For example, increasing pH may cause a
metal to precipitate and toxicity removed and EDTA
may also remove toxicity. In any case, results of this
nature are useful in selecting Phase II options. Use of
multiple manipulations (Section 9.2) builds upon these
principles. When several treatments are applied to the
same sample, tests must be designed to ensure that
toxicity does not result from the additives used (e.g.,
acid, base, EDTA) rather than from the effluent's
toxicant(
The assumption must not be made that toxicants
are either additive or synergistic. Our experience shows
that independent action (one or more of multiple toxi-
cants acting independently of the rest, as though the
others were not present) is not uncommon in effluents.
Experience also shows that one should not use se-
lected tests to confirm a suspicion that a certain toxi-
cant is the cause of toxicity. Time and again, this leads
to wasted effort. There are so many possible causes of
toxicity that such guesses are rarely helpful and more
often channel one's thinking and delay the final solu-
tion. On the other hand, if one wants only to know
whether a certain chemical is the toxicant, these tests
can be selected to accomplish that goal. Frequently
one needs to know whether the toxicity is due to am-
monia or whether there are toxicants present other
than salt. These questions are quite different from the
former case where one is playing the "I'll bet you the
toxicant is..." game.
No Phase I characterization test should be dropped
from use on the basis that the toxicants it is designed
to address are not likely to be present in the effluent. In
excluding any Phase I test, the analyst may be limiting
the information that can be gained on effluent toxicants.
The investigator should approach effluent characteriza-
tion without a preconceived notion as to the cause of
toxicity.
There are two types of checks that can be used to
detect artifact toxicity. A "toxicity blank" consists of
performing the same (Phase I) test on dilution water
and measuring to determine whether any toxicity is
added by the test procedure. However, a toxicity blank
does poorly in identifying artifact toxicity if toxicity is
affected by the effluents' matrix (cf., Section 7.1). For
example, the toxicity of the Phase I reagent, EDTA,
may be completely different in dilution water and in
effluent. If so, a toxicity blank is inappropriate for the
chelatipn test. A "toxicity control", for many Phase I
steps involves a comparison of the toxicity of the ma-
nipulated test solution and the baseline effluent tQxicity.
In this case, the comparison must demonstrate that the
manipulated effluent test solution has not become more
toxic than the unaltered effluent (baseline test). If it has,
the test procedure has produced artifactual toxicity.
The "toxicity control" for the pH adjustment/C,, test is
the filtered effluent sample at the respective pH, For
some treatments, valid toxicity blanks or toxicity con-
trols cannot be made. The use of toxicity blanks and
toxicity controls still requires the use of "regular" con-
trols, which are always included to determine the per-
formance of the test organism and dilution water. Dilu-
tion water blanks for the EDTA addition test, sodium
thiosulfate addition test, and the graduated pH test are
not relevant (see Section 7.1 for more information).
No procedure should be assumed to be free of
artifactual toxicity. Many of the Phase I toxicity tests
involve relatively severe or unorthodox effluent manipu-
lations. Toxicity blanks and toxicity controls must be
used consistently and conscientiously to detect the
introduction of toxic artifacts or other changes to the
effluent that increase sample toxicity.
For the following sections, the guidance for the
volumes required, apparatus, and test organisms is
based on test conditions using Ceriodaphnia, Daphnia
and/or larval fathead minnows exposed in 10 ml test
volumes.
8. / Initial Effluent Toxicity Test
Principles/General Discussion:
The major purpose of the "initial" effluent test is to
provide an estimate of the 24-h LC50 for purposes of
setting exposure concentrations in Phase I tests.
Volume Required:
Initial toxicity test is performed in duplicate using 40
ml of effluent.
8-4
-------
Apparatus:
Disposable 1 oz plastic cups or 30 ml glass bea-
species.
Procedure:
1: ml i n d
cate of 1 00%, 50%, 25%, 12.5%, 6.25% effluent, and a
control will suffice for most effluents. Obviously more
toxic effluents will require a lower range. If nothing is
known about the toxicity, more concentrations should
be included. A sample data sheet for the initial test is
shown in Figure 8-2.
8.2 Baseline Effluent Toxicity Test
In order to determine the effects that the various
Phase I characterization tests have on effluent toxicity,
the toxicity of the effluent sample, prior to any treat-
ment in the laboratory, must be determined. The por-
tion of the effluent sample, tested for toxicity the day
after it arrives in the laboratory (day 2), will be referred
LC50
the characterization tests. Such a comparison will dem-
onstrate whether the removal or alteration of various
groups of toxicants changes the effluent toxicity. By
comparing these results, an indication of the physical/
chemical nature of the toxicants can be obtained. If the
LC50 is substantially different
from the toxicity of the effluent when it arrived in the
laboratory (initial toxicity), one must decide whether the
schedule suggested in these methods should be re-
vised to reduce a delay in testing.
When Phase I testing is extended to additional
days, baseline tests must be done each time on suc-
ceeding days, and used for comparison to these addi-
tional manipulation tests.
Volume Required:
The baseline toxicity test is performed in duplicate.
The total volume necessary will depend, on the 24-h
LCSOmL s h o u I
be adequate..
Apparatus:
Disposable 1
oz plastic cups or 30 ml glass bea-
scope (optional).
Test Organisms:
Test organisms, 60 or more, of the same age and
species.
Procedure:
Day 2: Two concentration series will be used in
duplicate for the static acute toxicity test. In preparing
the test solutions for the day 2 baseline test, any
obvious physical changes (e.g., formation of precipi-
tates, odors), which occurred during storage, should be
noted.
The first test series will have exposure levels based
on the 24-h LC50 of the initial (day 1) toxicity and will
include day 2 effluent concentrations at 4x-, 2x-, lx,
Uafidj0..5x- the 24-h LC50. In this case, the method for
making dilutions described earlier may need to be
changed slightly. Most of the toxicity tests with the
characterization solutions will also be performed using
these same exposure concentrations. If the 24-h LC50
of the initial effluent is greater than 25%, the series
obviously begins at 100%, and includes four exposure
concentrations. Of course, if the 24-h LC50 of the day 1
initial effluent is greater than or equal to 25%, the
second series will be unnecessary because this test
fulfills the requirements for comparison to the initial
effluent test and characterization solution toxicity test
results.
The second test series will provide exposures at
effluent dilutions of 1 00%, 50%, 25%, 12.5% and 6,25%
(and lower dilutions as appropriate if the effluent is
more toxic). This series will enable a comparison of the
results of the baseline (day 2) test to the initial effluent
LC50 (cf., Section 8.1).
A sample data sheet is shown in Figure 8-3. In
order to compare the baseline toxicity and the toxicity
of the effluent aliquots subjected to characterization
tests, all of the day 2 toxicity tests should have the test
organisms added to test solutions at approximately the
same time.
The baseline toxicity test (toxicity control) must be
repeated each time additional characterization tests are
performed on the sample after the initial Phase I bat-
tery of tests. The baseline test will serve as the basis
for determining the effects produced by the additional
characterization tests, and will also provide information
on the degradation of sample toxicity. For effluents
whose initial toxicity is low (i.e., LC50~60-70%) and
where the baseline toxicity is greatly changed com-
pared to the initial toxicity of the sample, it may be
advisable to discard the remaining sample and collect a
freshj one.
Interferences/Controls and Blanks:
The control treatment of animals in unaltered dilu-
tion water in this test is used for comparison to several
subsequent tests and provides an important reference
for diluent water and organism acceptability. Mortality
in these controls will negate other work.
Results/Subsequent Tests:
Baseline LCSO's should be generated for as long
as the effluent sample is being used and a baseline
test (toxicity control) should be started every time the
8-5
-------
Figure 8-2. Example data sheet for initial effluent toxicity test.
Test Type: Initial Effluent
Test Initiation (Date & Time):
Investigator:
Sample Log No., Name:.
Date of Collection:
Species/Age:
No. Animals/No. Reps:_
Source of Animals: \
Dilution Water/Control:.
Test Volume:
Other Info:
Cone. 0
(% Effluent)
100
50
25
.12.5
6.25
Control
I
h
IPH
24 h
A B pH DO
Survival Readings:
48 h
A B pH DO
I
I
II II
I
1 LC50 1
| Cl 1
1 LC50
1 ™
72 h I 96 h
A B pH DO 1 A B pH DO
1
LC50
Cl
LC50
Cl
Comments:
8-6
-------
Figure 8-3. Example data sheet for baseline effluent toxiclty test.
Test Type: Baseline Effluent
Test Initiation (Date & Time):
Investiaator:
Sample Loq No., Name:
Date of Collection:
Soecies/Aae:
No. Animals/No. Reps:
Source of Animals:
Dilution Water/Control:
Test Volume:
Other Info:
Cone.
(% Effluent)
4X-LC50I
2x-LC50/
Ix-LCSOl
O.Sx-LCSO/
Control
100
50
25
12.5
6.25
0 h
PH
\
Survival Readings:
24 h
A B pH DO
LC50
c,
48 h
A B pH DO
72 h
A B pH DO
LC50
Cl
96 h
A B pH DO
»»
LC50
Cl
(S0T1M
One)
(Seres
Two)
Comments:
8-7
-------
effluent sample is put through any characterization steps.
(Note: similar procedures should be followed in Phases
II and III.)
8.3 pH Adjustment Test
Principles/General Discussion:
The pH has a substantial effect on the toxicity of
many compounds found in effluents. Therefore pH ad-
justment is used throughout Phase I to provide more
information on the nature of the toxicants. Changes in
pH can affect the solubility, polarity, volatility, stability
and speciation of a compound, thereby affecting its
bioavailability as weil as its toxicity. Before describing
the pH adjustment test, some discussion on the effect
of pH on various groups of compounds is warranted.
Two major groups of compounds significantly im-
pacted by solution pH are acids and bases. To under-
stand how organic and inorganic compounds of this
type are affected by pH changes, one must have a
basic understanding of the thermodynamic equilibrium
acidity constant, Kt, for the proton transfer reaction:
HA + H2O = H3O + A
K, = [A-][H1
[HA]
H*: H30+
HA: protonated acid
K4: thermodynamic equilibrium constant for
the acid
For example:
HCN + H2O = H3O + CN
Ka=[CNJ[HJ_= 6.0 x10'10
[HCN]
The stronger the acid (i.e., the more it tends to
dissociate into its ionic state), the greater the value of
Kt, and the smaller the -log Kt or pK. In effect, the
above reaction is shifted to the right for acidic com-
pounds. For acids in water, when the pH of the solution
equals the pKt of the compound, equivalent amounts of
the compound will exist in the ionized (A) and un-
ionized (HA) forms. At a pH one unit lower than the pKt
of the acid, approximately 90% of the compound will be
in the un-ionized form with the remainder in the ionized
form. A solution pH two units below the acid's pKa will
result in 99% in the un-ionized form and 1% in the
ionized form. Likewise, at one pH unit above the p/<,
90% of the acid will be present in the dissociated
(ionized) form and 10% present in the un-ionized form;
at two pH units above the pKt, 99% of the acid is in the
dissociated form while, 1% is* present in the un-ionized
form. For example, at pH 4.2, the pKt of benzoic acid,
50% of the compound is present as CSHSCOOH and
50% is present at/XJd .COO-.H+. At pH 3.2, this ratio
shifts to roughly 90%C.{H.pOOH:i.a°4CJH..COO-, H*
while at 5.2 the ratio nears 10% C6H5COOH to 90%
C6HSCOO',H*.
This relationship generally holds for diprotic and
triprotic acids (i.e., acids with two and three H atoms,
respectively, that can dissociate from the molecule).
This trend is not followed by multiprotic acids with pK§s
less than three units apart (e.g., H3BO3 with pK'jn=13.8
and pK. 2=12.74). The amount of each dissociated spe-
cies in such cases will not always follow the 90/10,99/1
rule stated above. For example, H,BO,,H BCy, and
HBO,2' will be present at pH 13.5.
Basic compounds function in a similar fashion.
B + H2O = BH' + OH-
Kb = [BHJ[OH_l
[B]
B: unprotonated base
K,: thermodynamic equilibrium constant for the base
For example:
C6H5NH2 + H20 = C6HSNH3* + OH-
Kb= [C6H5NH3*][OH-] = 4.2x1 0''°
[C6H5NHJ
In the above reaction, BH* can be considered the
"conjugate acid of the base", that is, the protonated
form of the base. Thus, the same reaction can be
expressed as follows:
BH++ H2O = H3O+ + B
= 14
[BH1
Note: pK>p
For example:
C6H5NH3
conjugate acid
HO = H3O*
C6H5NH2
Ka= [H1[C6H5NH2] = 2.34 x 10'5
[C.H5NH,1
This convention can be used to simplify dealing
with equilibrium constants for acids and bases.
As with acids, when the solution pH is equal to the
pK, of the conjugate acid of a base, equal amounts of
the* base will exist in the ionized and un-ionized forms.
For example, ammonia in an aqueous solution at pH
9.25 (the pK" of ammonia) will be found as 50% NH,+
and 50% NTH,. At one pH unit above the pK. (i.e.,
10.25) roughly 90% of the ammonia will be in the un-
ionized form (NH,) and the remainder will be in the
NH,+ form. At pH 8.25, one unit below the pK, of
ammonia, approximately 90% of the ammonia will be in
the NH,+ form, and approximately 10% will be in the
NH, form.
8-8
-------
Figure 8-4. p£-pH diagram for the CO,, HO, and Mn-CO, systems (25°C). Solid phases considered: Mn(OH) (a)
(pyrochroite), MnCO,(s) (rhodochrosite), Mna04(s) (hausmannite), y-MnOOH (manganite), y-MnO,
(nsutite). (Reprinted with permission from Stumm & Morgan, 1981.)
The above can be summarized by the following:
Predominant Spprlpg
Organic Inorganic
pH > pK,
acid
base
PH < ptf,
acid
base
RCOO-, RCO-
RNH,
RCOOH, RCOH
RNH/
A
B
HA
BH+
R = aliphatic or aromatic group
The effect of pH on the ratio of the ionized and un-
ionized forms of acids and bases has a number of
impacts on Phase I results. First, compounds may be
more toxic in the unionized form as compared to the
ionized form. For example, un-ionized ammonia (NH,)
is generally recognized as the toxic form of ammonia
while total ammonia (NH,+) is of far less concern (EPA,
19858). A second implication of this effect relates to
toxicant solubility. Unionized forms of acids and bases
can be considered less polar than their ionized forms,
which interact to a greater extent with water molecules.
Consequently, un-ionized forms of acids and bases can
be more easily stripped from water using aeration (Sec-
tion 8.5) or extraction with non-polar solvents or solid
phase column techniques (Section 8.6). Likewise,
changes in compound solubility with pH change may
mediate removal through filtration (Section 8.4).
Another implication of the pH effect involves metal
ion complexes. An example of how pH can alter the
form of a metal in a natural water system is shown in
Figure 8-4. Given a pe (the equilibrium electron activ-
ity-in a simple sense, whether the system is aerobic
or anaerobic), one can see how various forms of man-
ganese are created and eliminated as' pH shifts.
Each of the different forms of a metal will be
manifested differently in aquatic organism effects. Some
forms of the metal will be relatively insoluble; these
forms may not affect toxicity. Likewise, as with acids
and bases, the toxicity of the soluble forms of the metal
will be a function of the actual species present (e.g.,
the LC50 of Mn2* as compared to the LC50 of MnO42-).
The actual species formed will depend, in addition to
pH and pe on the other chemical constituents present
in the water. The hydrolysis rate of organics is greatly
affected by pH, and pH changes may also alter organic
toxicity.
Regardless of the speciation effect on toxicity,
changes in solution pH may affect the toxicity of any
given compound. The pH of the test solution may affect
8-9
-------
membrane permeability at the cell membrane as well
as the chemistry of the toxicant. One might expect that
changing the pH, only to return it to its original pH in a
short time, would not alter toxicity. Experience shows
that this is not the case and that this adjustment some-
times results in reduction, loss or increase in toxicity. If
the kinetics of the pH driven reaction (on return to the
original effluent pH) are slow or irreversible, pH adjust-
ment alone may be effective in evidencing toxicants
affected by pH change. Some organics may also de-
grade due to pH change.
Another purpose of the pH adjustment test is to
provide blanks (with both dilution water and effluent) for
subsequent Phase I pH adjustment tests performed in
combination with other operations. This test will dem-
onstrate whether toxic concentrations of ions have been
reached as a result of the addition of acid and base or
whether the reagent solutions are contaminated.
Comparable results for toxicity blanks are not ob-
tained when the same volumes and same strengths of
acids and bases are added to the effluent. Effluents
already contain substantial concentrations of major an-
ions and cations that are not found in dilution water.
Further, the volumes and strengths of the acid and
base necessary, for example, to lower an effluent with
a pHi of 7.6 to pH 3 and raise it back to pH 7.6, are
not likely to result in the same final pH when added to
dilution water. However, it is necessary to conduct pH
adjustment blanks to determine the cleanliness of the
acid and base solutions used for the pH manipulations.
Volume Required:
To make pH adjustments, 680 mL of whole effluent
is needed to have enough effluent for the four exposure
concentrations at each of the three pH's. A 300 mL
aliquot of the day 1 effluent sample is raised to pH 11,
and the second 300 ml sample is lowered to pH 3. An
aliquot of the pH i effluent (used for the baseline test:
80 mL) is set aside for the duration of the manipulation
(Figure 8-5). Approximately 30 mL will be needed for
the pH adjustment test but the actual amount depends
on the 24-h LC50 of the initial effluent test. The remain-
ing 270 mL of each of these solutions is reserved for
the "pH adjustment/filtration", "pH adjustment/aeration"
and "pH adjustrnent/C18 SPE" Phase I tests.
These pH adjustments must also be done using
dilution water for toxicity blanks for each test. To make
these adjustments, approximately 295 mL of dilution
water will provide enough (and an excess) to test the
blanks with one exposure level and one replicate. One
aliquot of 105 mL is adjusted to pH 3 and another 105
mL is adjusted to pH 11. Only 10 mL is needed for the
pH adjustment blanks but excess (-10 mL) is included.
The, pH i dilution water blank is the control of the
baseline test. The remaining 85 mL of pH adjusted and
85 mL of pH i dilution water are used for the "pH
adjustment/filtration," "pH adjustment/aeration" and "pH
adjustment/C,, SPE" toxicity blanks.
Apparatus:
Six glass stoppered bottles for acid and base solu-
tions, pH meter and probe, 2-500 mL beakers, 2-500 mL
graduated cylinders, 30 mL beakers or 1 oz plastic
cups, stir plate, and stir bars (perfluorocarbon), auto-
matic pipette, disposable pipette tips, eye dropper or
wide bore pipette, light box and/or microscope (op-
tional).
Reagents:
1 .0, 0.1 and 0.01 N NaOH, 1 .0, 0.1 and 0.01 N
HCL (ACS grade in high purity water) and buffers for
pH meter calibration.
Test Organisms:
Test organisms, 40 or more, of the same age and
species.
Procedure:
Day 1: The general procedure for the pH adjust-
ment test is shown in Figure 8-5.
Blank Preparation: The first step is to prepare the
dilution blanks. These blanks are used as the controls
for the other dilution water pH adjustment tests^of
aeration, filtration, and C1g SPE separation as well as to
determine whether the acid or base solutions are con-
taminated. The pH i blank is the control while the pH 3
or pH 11 adjustment blanks are treated in the manner
described below under sample preparation.
Sample Preparation: Stirring constantly, 1 .0 N NaOH
is added dropwise to a 300 mL aliquot day 1 effluent
until the solution pH nears 11. (Note: overshooting
results in the addition of more salts and a volume
change and should be avoided.) In order to minimize
any over-adjustment of the pH, 0.1 N NaOH is added
dropwise in the latter stages to bring the effluent aliquot
to pH 11. The solution should be allowed to equilibrate
after each incremental addition of base. The amount of
time necessary for pH equilibration will depend on the
buffering capacity of the effluent. Caution should be
taken to prevent any solution pH of greater than 11. If
pH 11 is exceeded, 0.1 N HCI must be used to lower
the pH to 11. The goal of the pH adjustment step is to
reach pH 11, while minimizing both the change in
aliquot volume and the increase in ionic strength. Vol-
umes and strengths of base (and any acid added)
should be recorded. A 30 mL volume of effluent and
dilution water is held for the same length of time it
takes to complete other Phase I manipulations with pH
11 effluent.
Once other manipulation work has been completed
with the total volume of the pH 11 effluent, the 30 mL
volume at pH 11 is returned to the initial pH (pH i) of
the day 1 effluent. (The other aliquots of pH 11 effluent
are also returned to pH i at this time.) This is accom-
plished by the slow, dropwise addition of 0.1 N HCI first
and later 0.01 N HCL as the pH of the stirred solution
nears pH i. If pH i is exceeded, the pH must be
appropriately increased with 0.01 N NaOH. Again, the
8-10
-------
(A
t>
«
H adjustment t
w chart for p
2,
u_
i
«
.!»
u.
T-
'••*
o.
13
Q.
CO
C/D
+•*
5=
LU
_J
E
o
8
__
+,
-»
X
Q.
O
.+—1
-T-1
0)
4-*
CO
g^
•^^
^
-I
o
0
CO
•^^^^
'**
Q.
13
_J
E
00
CO
o
*-
T3
0)
175
3
*
*
_J
E
o
o
CO
—
«•••
M^H
•
-^
»
X
n
oa
_i
E
o
CO
X
Q.
13
_i
E
o
IN^^
n*
CM
MMWMM
CO
CL
+*
OJ
_I
E
o
AM
CO
CO
CL
13
_l
E
0
r-.
C\J
•»•*
X
CL
Q B
m~ ' * T^ -
s 1
t
• M*
Q.
-=— 5 £ LU °
0 OJ yj ^-^0- 0
— «*c5 Sw 22^-*- g
5 £• 0) ±± >
OJ
Q
,
*
^
c
C
,
4—1
O I/)
'x 0)
Or-
^°>
.£.-&*-
•5. a <£
JA X *
Sol-
<*^\ I—
CD r~
CVJ
£~
Of/I
x^
0 r-
NJ
>»
CC
ID
ci1*
•o
Q
Q
C
to
CD
E
3
O
CO
CO
_Q
jQ
O)
CO
_CD
^O
W
J£
-1
i £
1 O
X
o
iz
CO
CL
CD
Q.
O
4—1
o3
§ d
> r-
S «
1 §
T3 O)
C U-
0 .E
CD "O
ts s
3 0
•o a
0 -S
8 ^
fi ^
rt to
8. B
CD Q
4— Z
cc £
< C/)
T- CM
8-11
-------
volumes and strengths of acid and any base added
should be recorded.
The pH of the solution should be checked periodi-
cally throughout the remainder of the work day and
readjusted as necessary. Changes in the total volume
of acids and bases added should be recorded.
This procedure is repeated, except the pH is low-
ered to pH 3 using the second 300 ml aliquot of
effluent, and 1 .0 N and 0.1 N HCI. As with the pH 11
effluent, 270 mL of the pH 3 effluent is used for the pH
adjustment/aeration, pH adjustment/filtration, and pH
adjustment/C,, SPE tests. The remaining 30 ml of the
pH 3 effluent is held until all of the work on all of the pH
3 effluent has been completed. At this point, the pH of
the 30 ml volume of pH 3 effluent is readjusted to pH;
by the dropwise addition of 0.1 N and 0.01 N NaOH.
Maintenance of pH; must be assured through check-
ing and readjusting the sample periodically throughout
the work day. All volumes and strengths of acid and
base added should be recorded.
Day 2: At the beginning of the work day (the day
after the arrival of the effluent in the laboratory), the pH
of both of the 30 mL volumes is again checked to
ensure that pH/has been maintained. Any additional
pH adjustments are made and the volumes of the acid
and/or base added are recorded. The acute toxicity of
each pH-adjusted solution is tested at 4x-, 2x-, Ix-,
0.5x-LC50 (the 24-h initial LC50) as described in Sec-
tion 7. Test solution pH should be measured in all
exposure concentrations and recorded at least every
24 h. A sample data sheet is shown in Figure 8-6.
Interferences/Controls and Blanks:
Controls prepared for the baseline toxicity test also
act as a check on the organisms, dilution water, and
test chambers for this test.
The baseline effluent test acts as a control for the
pH adjustment test, indicating whether the addition of
NaCI in the form of the acid and base has increased
effluent toxicity. This pH adjustment test acts as the
control for other Phase I tests entailing pH adjustment.
In addition to serving as a control for other pH adjust-
ment tests, increased toxicity following pH adjustment,
not as a result of NaCI concentration, indicates a pH
effect on toxicity or contamination of acids or bases
(see below).
Results/Subsequent Tests:
If either the pH 3 or pH11 adjustment effluent tests
have significantly greater toxicity than the baseline ef-
fluent test, two possible sources of toxicity exist: 1) the
ions (Na*,CI') added by the acid and base have re-
sulted in a solution with an ionic strength intolerable to
the test organism; or 2) chemical reactions driven by
the pH change have not reversed upon readjusting to
pH;. Neither of these phenomena would be detected
through the use of a blank (dilution water). To help
resolve this situation, the NaCILCBO values for com-
mon test organisms are provided in Table 8-2. The
minimum concentration of NaCI in the test solution (i.e.,
not including the concentration of NaCI originally present
in the effluent) can be calculated from the volumes and
strengths of the acid and base added and final solution
volume. The data in the table can be used only as a
rough guide, however, because the toxicity of sodium
chloride depends on the other anions and cations as
well as the total osmotic pressure exerted by the dis-
solved substances. The toxicity of the added NaCI is
best determined by adding that amount of NaCI directly
to the effluent and to see if the addition increased
effluent toxicity.
If either the pH 3 and/or pH 11 adjustment tests
indicate in a significant decrease in effluent toxicity, it
could result from volume changes by acid and base
additions or from chemical reactions driven by the pH
change that may not have been re-established or are
irreversible. To determine if the addition of acid or base
diluted the sample due to their volume addition, add a
volume of dilution water equivalent to the total volume
of acid and base originally added to the effluent vol-
ume. If a similar loss of toxicity in the diluted wastewa-
ter occurs, the pH adjustment test should be repeated
using more concentrated acid and base solutions.
A reduction or loss of toxicity may also beflhe
result of the degradation of toxicant at the altered pH
values. In some cases, the toxicity could also be in-
creased if the degradation product is more toxic than
the original compound. Both organics and inorganics
can be so changed with a probable loss in toxicity.
Inorganic and organic substances may precipitate dur-
ing the process of pH adjustment. The precipitated
chemical may or may not be the toxicant. The precipi-
tated chemical (which most often forms with the pH11
adjustment) removes from solution via the flocculation
process, suspended solids, microbial growth, and col-
loids, and via the adsorption process, metals and or-
ganics. If the process of precipitation seems to remove
toxicity, it is important to realize that the precipitating
chemical might not be the toxicant, but rather that the
toxicant(s) may have been removed by the flocculation
and/or adsorption processes. In some cases, the pre-
cipitate may dissolve with the adjustment of the effluent
back to pH i. The removal of toxicity when dissolution
of the precipitate occurs should be evaluated carefully
since the toxicant(s) might be unavailable and/or not
completely dissolved.
For most of the Phase I combination pH adjustment
tests (i.e., pH adjustment/filtration), the pH adjustment
test will act as an equivalent or "worst case" toxicity
control for changes in test solution ionic strength and
volume. In effect, most of the operations applied to the
pH adjusted effluent in the following Sections (8.4-8.6)
will either not affect pH or will drive it closer to the pH i.
This may not be the case for the pH adjustment/aera-
tion test, however. Because pH 3 and pH 11 must be
maintained throughout the aeration process and be-
cause the loss of volatiles may result in pH shifts
towards pH;, more acid and/or base may be added to
these test solutions as compared to the pH adjustment
only solutions.
8-12
-------
Figure 8-6. Example data sheet for pH adjustment test.
Test Type: pH Adjustment
Test Initiation (Date & Time):
Investigator:
Sample Log No.. Name:
Date Of Collection:
Species/Age:
No. Animals/No. Reps:
Source of Animals:
Dilution Water/Control:
Test Volume:
Other Info:
pH/Concentration
(%pH adjusted
effluent)
3/4x-LC50
3/2x-LC50
3/1x-LC50
3/0.5x-LC50
3 / blank
pHU blank
11I4X-LC50
11/2X-LC50
11/1X-LC50
11/0.5X-LC50
11! blank
) h
pH
Survival Readings:
24 h
\ApHH DO
1
1
1
1
48\\ I 72\\
A pH DO A pH DO
96\\
A pH DO
II ll 1 1
1 LC50 1 LC50 1 LC50 1 LC50
Note: See baseline data sheet for control data.
Volumes and Strength of Solutions Added:
HCI NaOH
300 ml pH 3
300 mLpH 11
30 ml_pH3
SOmLpH 11
Comments:
8-13
-------
Table 8-2. Acute toxicity of sodium chloride to selected aquatic organisms
Species
Ceriodaphnia dubia
Water
Typ«
SRW12
SRW'2
SRW2
SRW'2
VHRW1-2
Life-
stage
124 h
<24h
<24h
<24h
24 h
4.2
3.3
3.0
2,3
(2.0-2.6)
2.8
LC50 (g/L) (95% Cl)
48 h 72 h 96-K
2.3
(2.0-2.6)
2.7
2.1
2.3
(2.0-2.6)
2.8
Daphnia magna
Pimephales promelas
Lepomis macrochirus
LSW"
RW"
SRW"
SRW24
SRW"
VHRW"
SRW
NR
NR
524 h
524 h
11 wk
S24 h
I-9 g
3.3
(NR)
6.4
(NR)
7.9
(7.0-9.0)
5,3
(4.8-5.8)
7.9
(NR)
8.0
(-)
NR
3.1
(NR)
5.9
(NR)
7.9
(7.0-9.0)
(4 $6)
5.2
(4.2-6.6)
NR
6.9
(5.5-8.7)
4.6
(4.0-5.3)
NR
4.2
(3.3-5.2)
NR
4.6
(3.3-6.2)
3.5
(2.8-4.3)
7.7
(7.4-7.9)
3.7
(2.9-4.7)
12.9
(NR)
' Data generated at ERL-Duluth. C. dubia were S24 h old at test initiation and fed. Water used was soft reconstituted water (diluted mineral
water (DMW)).
1 Static, unmeasured test.
'Dowden and Bennett, 1965.
* Data generated at ERL-Duluth and values represent those from 7-d fathead minnow growth and survival tests and daily renewals.
5 Adelman et al., 1976.
'Patrick et al., 1968.
Note: (—) = Confidence interval cannot be calculated as no partial mortality occurred. NR = Not reported; SRW = soft reconstituted water;
VHRW = very hard reconstituted water; RW = reconstituted water; LSW = Lake Superior water.
There is another factor which must be considered
when carrying out pH adjustment tests. In those ma-
nipulations where the pH is changed to pH 3 or pH 11
and then readjusted to pH /, the pH may tend to drift
over the course of the 48-h or 96-h toxicity tests. The
drift can be very dissimilar among test manipulations.
This is likely to occur even though the starting pH's (of
samples readjusted to pH /} may be similar. This can
lead to confusion in interpreting Phase I results if a
compound whose toxicity is pH dependent is present in
the sample. An example of a manipulation in which this
effect is encountered routinely is the pH 3 adjustment/
aeration test (Section 8.5). For instance, an aliquot of
an effluent with a pH; of 7.5 is adjusted to pH 3 and is
kept at that pH while the other manipulations are con-
ducted. This pH 3 adjusted sample serves as a control
for the pH 3 adjustments. Another portion of the pH 3
adjusted effluent is aerated (Section 8.5). Both aliquots
are then readjusted to pH / (75) prior to toxicity test-
ing. The pH of the pH 3 adjustment/aeration test solu-
tion will probably not behave in a similar manner to the
baseline orpH 3 adjustment test. We have observed
the pH in this test to go unchanged or drift downward
after adjustment up to pH/of 7.5 over the course of the
toxicity test. However, the pH of the effluent in the
baseline test and the pH11 adjustment test may drift
upwards over the course of the toxicity test, from pH;
(7.5) to as high as pH 8.5. By the end of the test, the
analyst may be confronted with interpreting the results
of tests conducted at different pH values. If a com-
pound whose toxicity is dependent upon pH (e.g., am-
monia) is present in the sample (cf., Section 8.9 for a
discussion of the effects of pH on ammonia toxicity),
the fact that pH either did not change, or even drifted
down in the pH 3 adjustment/aeration test sample (rela-
tive to the baseline and/or pH 3 adjustment test), can
complicate interpreting the test results. If ammonia were
present (which is less toxic at a low pH), the sample
would appear to have lost toxicity in the pH 3 adjust-
ment/aeration test, when the loss in toxicity may have
been the result only of the differences in pH drift during
the toxicity tests.
The pH 3 adjustment/aeration test is not the only
manipulation that may cause differential pH drift over
the course of the toxicity test. Virtually all the manipula-
tions in Phase I have the potential to cause this effect.
For example, with some effluents any pH 3 manipula-
tion (Le., pH adjustment, aeration, or filtration) followed
8-I 4
-------
by readjustment to pHi, will cause pH to behave
differently than the pH of the baseline test. Similarly,
the pH 11 manipulations (i.e., pH adjustment, aeration,
or filtration) can cause similar fluctuations, but they do
not seem to occur as frequently as with the pH 3
manipulations. Another manipulation that causes this
differential pH drift is passing effluent over the C18 SPE
column at pH(, pHSorpH 9 (Section 8.6). Although
not as drastic as some of the effects observed with the
pH 3 adjustment/aeration tests, the sample collected
after passing the effluent over the C,, SPE column may
have a slightly lower pH by the end of the toxicity test
than the pH of the baseline test (e.g., pH 8.2 as op-
posed to pH 8.5). A final manipulation that has the
potential to cause acidic pH drift is the addition of
EDTA; details of this pH fluctuation are elaborated in
Section 8.8.
Although ammonia is a commonly encountered
sample toxicant whose toxicity is pH dependent, it is
not the only compound whose toxicity can be affected
by different test pH's (cf., Phase II). We have observed
pH dependence with several metals and the effects of
differential pH drift after various Phase I manipulations
should be considered. pH should always be monitored
and recorded whenever mortality readings are made
(e.g., 2 h, 24 h) as well as at the end of the test. It is
particularly important to record pH of the concentra-
tions that determine the LC50, especially if greater than
5 mg/L of total ammonia is present in the sample.
Differential pH drift after manipulations can be over-
come by closely monitoring the test pH, and adjusting
the pH in the manipulated samples to match the pH of
the baseline toxicity test. These adjustments are done
before animals are introduced.
8.4 pH Adjustment/Filtration lest
Principles/General Discussion:
The filtration experiment provides information on
effluent toxicants associated with filterable material. Toxic
pollutants associated with particles may be less biologi-
cally available. However, aquatic organisms can be
exposed to these pollutants through ingestion of the
particles. This route of exposure may be significant for
cladocerans and other filter feeders ingesting bacterial
cells and other solids with sorbed toxicants. The de-
gree to which any compound exists sorbed or in solu-
tion depends on a number of factors including particle
surface charge (or lack thereof), surface area, com-
pound polarity and charge, solubility and the effluent
matrix. By filtering particles from the effluent, an imme-
diate cause or a sink of toxic chemicals may be re-
moved.
In addition to determining the effect of filtration on
the toxicity of the whole effluent, the effects of pH
adjustment in combination with filtration are also as-
sessed with this manipulation. As discussed in Section
8.3, changes in solution pH can result in the formation
of insoluble complexes of metals (Figure 8-4) . Similarly,
organic acids and bases existing in ionic form can be
transformed into the non-ionic form by pH adjustment.
Shifts in effluent pH can also act to drive dissolved
toxicants onto particles in the effluent (e.g., shifting the
dissolved/sorbed equilibrium away from the free form).
Changes in toxicant polarity resulting from solution pH
change can make some particle/toxicant interactions
stronger. In other cases, the increase in effluent ionic
strength resulting from the shift in pH may force non-
polar organic compounds onto uncharged surfaces to a
greater extent.
By filtering pH adjusted aliquots of effluents, those
compounds typically in solution at unadjusted pH but
insoluble or associated with particles to a greater ex-
tent at more extreme pH's, are removed. By removing
the toxicant-contaminated particles or precipitated com-
pounds prior to readjustment of the sample to pH i,
these toxicants are no longer available for dissolution in
the effluent. The pH change may also destroy or dis-
solve the particles, thereby removing the sorption sur-
faces or driving the dissolved/sorbed equilibrium in the
opposite direction.
Positive pressure filtration is recommended. Use of
a vacuum to draw the effluent sample through the filter
may result in a loss of volatile compounds by degas-
sing the solution during filtration. This problem is ppten-
tially worsened in pH adjusted effluents if toxicants
become more volatile as a result of pH changes. If
vacuum filtration is used and effluent toxicity is re-
duced, subsequent tests must be performed to define
the nature of the toxicity loss. In this filtration step,
whether pressure or vacuum filtering is done, it is
important to avoid stainless steel housings for either
the pH 3 or pH 11 adjustment tests. Teflon, plastic or
glass equipment does not have the associated toxicity
that the stainless steel has under acidic or basic condi-
tions.
The pH adjustment/C,, SPE test (Section 8.6) re-
quires the use of filtered effluent. Without knowledge of
the effect of filtering on the effluent toxicity, it is not
possible to tell whether or not the SPE column or the
filtration removed the toxicity. Filtering may also be
useful in connection with other Phase I tests.
Volume Required:
A 235 ml aliquot of pH i effluent is filtered. Also,
235 ml each of pH 3 and pH 11 effluent aliquots
(Section 8.3) are filtered. The remaining 35 ml_ of each
solution is reserved for the pH adjustment/aeration tests.
A maximum volume of 30 ml_ of each of these three
solutions is needed to perform the filtration toxicity
tests. The exact effluent volume required for the toxicity
test will be a function of the effluent toxicity (Section 7).
Each test (pH 3, pH i, pH 11) requires four exposure
concentrations (10 ml_ each). The remaining filtered
effluent volumes (+200 ml) of pH 3, pH 11, and the
pH i solutions are each reserved for the C,. SPE tests
(Section 8.6). Excess volume has been Included to
cover losses occurring during the filtration operation.
For the blanks, 85 ml_ of dilution water is needed
for each pH test, of which 50 mi_ will be used in the pH
adjustment/SPE test.
8-15
-------
Apparatus:
Six 250 ml graduate cylinders, six 250 mL bea-
kers, six 50 mL beakers, pump with sample reservoir,
teflon tubing, in-line filter housing, ring stands, clamps.
Alternatively, vacuum flask, filter stand, clamp, vacuum
tubing, water aspirator or vacuum pump. Glass-fiber
filters (nominal size of 1 .Q^m, without organic binder),
stainless steel forceps, glass stoppered bottles for acid
and base solutions, pH meter and probe, stir plate,
perfluorocarbon stir bars, automatic pipette, disposable
pipette tips, eye dropper or wide bore pipette, 30 mL
beakers or 1 oz plastic cups, light box and/or micro-
scope (optional).
Reagents:
Solvents and high purity water for cleaning pump
reservoir and filter, 1 .0, 0.1, and 0.01 N NaOH, 1 .0, 0.1,
and 0.01 N HCI (ACS grade in high purity water),
buffers for pH meter calibration.
Test Organisms:
Test organisms, 75 or more, of the same age and
species.
Procedure:
Day 1: First, the filters must be prepared. These
steps are outlined in Figure 8-7. After the filters are
prepared, the dilution water at the appropriate pH is
filtered and collected for the toxicity blanks. Finally the
effluent samples at each of the three pH's are filtered
(Figure 8-8). Use of glass-fiber rather than cellulose-
based filters should minimize the adsorption and loss of
dissolved non-polar organic compounds from the efflu-
ent sample. Adsorption of toxic dissolved compounds
onto the filter or onto particles retained by the filter can
lead to spurious results.
filter Preparation: To prepare the 1 .0 u.m glass-
fiber filter for use, wash two 25 mL volumes of high
purity water through the filter. For the pH 3 effluent
filtration test, the filter must be washed with high purity
water adjusted to pH 3 using HCI. Likewise, the filter
used with the pH 11 effluent sample must first be
washed with high purity water adjusted to pH11 using
a concentrated NaOH solution. Washing the filters with
water adjusted to the same pH as the effluent should
prevent sample contamination with water soluble toxi-
cants contained on the filters.
Blank Preparation: The next step is to prepare filter
blanks using dilution water (Figure 8-7). These blanks
are used to detect the presence of any water soluble
toxicants which may remain on the filter following the
washing process. The pH i filtration blank is simply
prepared by passing 50 mL of dilution water (where the
pH is unadjusted) through a washed filter. The filtered
dilution water is collected and 30 mL of this volume is
reserved for the post-C,, SPE column toxicity blank
(Section 8.6). The remaining 20 mL is used as a filtra-
tion toxicity blank. Again, excess is included to cover
any possible loss during rinses.
To prepare the pH 3 filtration blank, 105 mL of
dilution water is adjusted to pH 3 with HCI, caution
being, taken to minimize the increase in dilution water
ionic strength. Of the pH 3 adjusted dilution water, 20
mL is for the pH adjustment only test, 35 mL is re-
served for use as a toxicity blank in the aeration test
(Section 8.5), and 50 mL of pH 3 dilution water is
passed through a filter previously washed with pH 3
rinse water. The filtered pH 3 dilution water is collected
and 30 mL of this volume is reserved for the pH 3
filtration/C,, SPE toxicity test blank. The remaining 20
mL is readjusted to the initial pH of the dilution water
using NaOH, again taking care not to exceed the initial
pH of the dilution water during the readjustment pro-
cess. This solution is used in a single exposure toxicity
test as the filtered pH adjustment toxicity blank.
The pH 11 toxicity blank sample is prepared in a
similar fashion using 105 mL of dilution water adjusted
to pH 11 with NaOH. Of the pH 11 dilution water, 20
mL is reserved for use in the pH adjustment only test
and 35 mL for the aeration test. The remaining volume
(50 mL) is filtered using the filter previously washed
with pH 11 rinse water and 30 mL of the filtered pH 11
dilution water is collected for use as the pH 11 filtration/
Cie SPE blank. The remaining 20 mL is readjusted to
the initial pH of the dilution water with HCI and used as
the pH 11 filtered toxicity blank using a single expo-
sure.
Sample Preparation: The same filter(s) used to
prepare the pH i (or pH 3 or pH 11) dilution water
filtration blank(s) is now used to filter the pH i forpH 3
orpH 11) effluent. First, a 235 mL aliquot of the pHi
effluent is passed through the pH i prepared filter and
collected; of which 200 mL is reserved for the C1§ SPE
test. The remaining volume (approximately 30 mL) is
held for the pH adjustment/filtration toxicity tests.
Now, using the same filter used to prepare the pH
3 filtration blank, 235 mL of pH 3 effluent (see Section
8.3) is filtered and collected. The filtered pH 3 effluent
is split into two aliquots (200 mL and approximately 30
mL). The 200 mL aliquot is used in the pH 3 adjust-
ment/C,, SPE test. The 35 mL filtered aliquot is read-
justed to pH i using NaOH. Care must be taken to
minimize both an increase in aliquot volume and ionic
strength. The pH readjusted 35 mL aliquot is held for
the toxicity testing.
Finally, the filtration step is repeated using 235 mL
of the pH 11 effluent (Section 8.3) and the filter origi-
nally used to filter pH 11 dilution water. Again, 200 mL
of the pH 11 filtered effluent is used in the pH adjust-
ment/C,,, SPE test and the filtered sample (approxi-
mately 30-35 mL) is readjusted to the pH / of the
effluent with HCI and used to conduct a toxicity test on
day 2.
In filtering effluent samples with high solids content,
it may be necessary to use more than one filter for the
235 mL of effluent. If so, the filter preparation step must
8 16
-------
Figure 8-7. Overview of steps needed in preparing the filter and the dilution water blanks for the filtration and/or the C1( solid phase extraction column tests.
Da\
Da
__=
Preparing
the Filter
M
Preparing
Dilution
Water
Blanks
y2
l_
HCI —
-»•>
t
50 mL of pH 3
* .
Filter
i
Discard water
t-*~ HCI
50 mL of pH 3
*
prepared filter
i
30 mL
*
f
|20mL|
•^- 1
Through C18
NaOH SPE Column
i *. j
f
Toxicity
Test
<
Toxicity
Test
300 mL High Purity Water
I
1
-^-* — h
50mLofpH/(unadj.) SOmLofpH 11
t
Filter
i
Discard water
Dilution Water
1
\
j
f
Filter
f
Discard water
NaOH -
*»
SOmLofpH / 50 mL
t
prepared filter
I
30 mL
^JaOH t
T
Through C18
SPE Column
I
»
Toxicity
Test
t
of pH 1 1
i
prepared filter
~~\ i
20 mL 30 mL
1
••
IHCI -** t
?
i
20 i
Through C18
SPE Column
HCI-*- 1
{
Foxicity Toxicity
Test Test
1
Toxi
Te
JaOH
mL
city
st
-------
M
«
2
c
2
X
a.
2
1
•
o'
I
CO
§
D)
a.
=
co
a.
o
0)
O
Q)
Q.
00
d>
LU
— l
t
crJ
»»
JL
i
T
T-
Q.
**—
O
_t
E
in
CVJ
X
Q.
O
I
E
10
CM
CO
X
Q.
O
1
b
IT)
CO
CVJ
— ».
0)
«i
XJ
£
a
0)
a
k.
«=
8
Q.
Q.
$
CO
a.
CD
*^~
-*-
>
CO
Q
>
1
c
If)
CO
E
o
CM
_ j
CO
E
OJ
i
E
CO
"E
o
o
CM
O
T
1
1
O
*
1
1
i
T
T—
co C
13
, LU
Jc a.
H W
OCMZ
fS
2 LU
•C ^«
h- (/)
oo*~C
cJ ^
fS
Ol 1 1
x: Q-
h- CO
*
T
i
o
i ••
t
i
1 1
II 1
CH>>
Gt5
H
'G t?5
OH
P
<\i
IQ>
'S l/i
§ H
"o "55
•
^~
CM
2r
o t!
*c? 0?
O r-
'O tf)
•~
— 2? -
CD
t/i
1
>; "
o ,
*— 00
CO §
§ ,2"
" Jl_
"S
52 •§"
ra
^ tf>
a B
Li- CP
H
-------
be repeated to provide additional filtration blanks or
several filters can be prepared at one time by stacking
them together in the filter housing. Alternatively, it may
be possible to centrifuge samples high in 'suspended
solids and filter the supernatant through a single filter. If
this option is taken, the toxicity of the supernatant must
be tested and compared to the toxicity in the baseline
effluent test.
In the above procedures, either separate effluent
filtration systems should be used or the filtration system
must be cleaned between pH adjusted aliquots to pre-
vent any carry-over of toxicity or particles. This means
all equipment should be thoroughly rinsed with 10%
HNO,, acetone, and high purity water between aliquots
of effluent with the exception of stainless steel equip-
ment where dilute solutions should be used.
The pH of the pH adjusted blanks and effluent
aliquots, designated for day 2 toxicity tests, should be
checked periodically throughout the work day. Adjust-
ments should be made as necessary in order to main-
tain the pH /' of these solutions.
Day 2: Prior to initiating the toxicity tests, the pH of
the pH 3 and 11 blanks and filtered effluent aliquots
should be measured and readjusted to pH /'. Toxicity
tests performed on all three (pH 3, pH/, and pH 11)
filtration blanks involve testing without dilution. Based
on the 24-h initial LC50 of the day 1 effluent, toxicity
tests performed on the effluent aliquots filtered at pH 3,
pH 11, and pH / are set up at 4x-, 2x-, 1 x-, and 0.5x-
LC50 as described in Section 8.2. Measurement of
exposure pH should be made daily on concentrations
around the mortality and the highest tested concentra-
tion, concurrently with survival readings. A sample data
sheet for the filtration tests is shown in Figure 8-9.
Interferences/Controls and Blanks:
Controls prepared for the baseline toxicity test serve
as a check on the quality of organisms, dilution water
and test conditions. Results of the pH adjustment test
(Section 8.3) will indicate whether or not toxic levels of
NaCI have been produced through pH adjustment only.
Results of the effluent filtration tests at each pH
should be compared with the filtration dilution water
toxicity blank performed at the corresponding pH to
determine the validity of the toxicity test outcome. No
significant mortality should occur in any of the filtration
blanks. If unacceptable mortality of organisms occurs in
either the pH 3 or pH 11 adjusted filtration blanks,
further investigation will be necessary to determine
whether lethality resulted from toxicants leached from
the filter at pH 3 and/or pH 11, or whether the increase
in dilution water ionic strength (via acid and base addi-
tion) is responsible for the problem. Additionally, if the
pH 3 and/or 11 filtration, aeration (Section 8.5) and C18
SPE (Section 8.6) dilution water blanks have approxi-
' mately the same final concentration of acid and base,
any ionic strength related toxicity should also be de-
tected in them.
If a filtration toxicity blank shows unacceptable acute
toxicity but the corresponding filtered effluent is equally
or less toxic than the baseline toxicity test, it is possible
that the dilution water toxicity blank removed the final
traces of toxic filter artifacts. In some cases, the efflu-
ent matrix may have also prevented the artifacts from
leaving the filter or masked their presence. Alternatively
the observed filtered effluent toxicity may represent the
net effect of toxicant removal via filtration plus contami-
nation by filter artifacts.
Results/Subsequent Tests:
The LCSOs for the aliquots of pH 3, pH / and pH 11
filtered effluent are compared to the baseline effluent
LC50 to determine whether any of these processes
resulted in a significant change in effluent toxicity.
If toxicity can be removed by filtration, either with or
without pH change, one has a method for separating
the toxicants from other material in the effluent. This
knowledge itself provides an important advance be-
cause further characterization and analyses will be less
confused by non-toxic constituents. Usually further char-
acterization will be the next step. Tests must be de-
signed to determine whether the mechanisms causing
removal are precipitation, sorption, change in qquilib-
rium or volatilization. One necessary step is to tecover
the toxicity from the filter. If this cannot be done and the
loss is not by volatilization, then the whole experiment
may have little utility. Comparisons of pressure and
vacuum filtration may reveal if volatilization is involved.
If characterization of the toxicant is also achieved
through other tests, filtration can be used to remove the
toxicants. Then if the suspected toxicant is the true
one, its concentration should be lower or zero after
toxicity is removed by filtration.
If any or all of these pH/filtration combinations re-
sult in less effluent toxicity (not attributable to the ef-
fects of pH adjustment alone), it may be possible to
confirm the findings of the test. A transfer of the solids
contained on the filter back into the filtrate at pH / can
be attempted by reversing the flow of the filtrate through
the filter or by rinsing the solids off the filter with filtrate.
The toxicity exhibited by this solution should be similar
to that of the original effluent, provided that the final
concentration of solids in the test solution approximates
the solids level in the sample that was filtered. For
precipitates formed as a result of pH changes or for
contamination of suspended solids facilitated by pH
adjustment, time must be allowed for the precipitate to
redissolve in the pH /' filtrate or for a new equilibrium to
be set up between the contaminants on the solids and
in solution. The results of this test are not likely to be
quantitative due to the recovery problems inherent in
the process.
In order to determine whether the effluent matrix
affects the toxicity of filterable particles (e.g., its ionic
strength, dissolved organic carbon content), the filtered
material can also be added to a volume of pH/dilution
water equal to the volume of effluent that was passed
8-19
-------
Figure 8-9. Example data sheet for filtration test.
Test Type: Filtration Species/Age:
Test Initiation (Date & Time): No. Animals/N
Source of Aniri
o. Reps:
nals:
Investigator: Dilution Water/Control:
Sample Log No., Na
Date of Collection:
me: Test Volume:
Other Info:
pH/Conc.
(% Effluent)
3l4x-LC5Q
3/2x-LC50
3/Ix-LCSO
3/0.5x-LC50
31 blank
pHi/4x-LC50
pHi/2x-LC50
pHillx-LCSO
pHi/0.5x-LC50
pH il blank
11I4X-LC50
ll/2x-LC50
llllx-LCSO
ll/0.5x-LC50
11 1 blank
Oh
pH
24 h
A pH DO
Survival Readings:
48 h
A pH DO
1
II
72 h
A pH DO
96 h
A pH DO
1 LC50 1 LC50 1 LC50 1 LG50
Note: See baseline data sheet for control data.
Volumes and Strength of Solutions Added:
HCI NaOH
pH3
pHi
pH11
Comments:
8-20
-------
through the filter. The toxicity of this dilution water,
spiked with effluent solids, can be compared to the
toxicity of the unfiltered (baseline) effluent and the fil-
trate spiked with its own solids.
pH The additional tests
may or may not provide the relevant information.
to
I S
sonicate
pH-adjusted
and possibly cyanide are not readily stripped using
techniques described in the Procedure below, and one
should not assume that they will be removed. Under
both air and nitrogen sparging, a removal process, in
addition to volatilization, may occur. Sparging can re-
move surface active agents from solution by the pro-
cess of sublatipn (lifting, up, carrying away). Surface
fcftve^ a|efftss have li molectilaY^tfuBture that includes
a polar end (either ionic or nonionic) and a relatively
large non-polar, hydrocarbon end. Some examples of
eiagents are resin acids, soaps, deter-
r e c o vtoxicant(s)from
withasolv ,_.„.._
water). For so m9SntseffarafXe?Nbgization polymers and coagulation
pH 11 ad j u sROrMnfe3?6tu)Bec' m chemical manufacturing processes.
filtration test. These results may cause one to suspect The process of sublation occurs because during
either cationic surfactants a s toxicant(s) sparging, surface active agents congregate at the liq-
94) We have had succe syi§l/gaBrinteiffai;%
-------
Test Organisms:
Test organisms, 75 or more, of the same age and
species.
Procedure .-
Day1: Six different solutions are aerated in this
test; pH 3, pH j, and pH 11 effluent, and pH 3, pHz,
and pH 11 dilution water, (cf., Sections 8.3 and 8.4,
respectively for preparation of pH adjusted effluent and
dilution water). A flow chart for the effluent samples of
the pH adjustment/aeration test is shown in Figure 8-
10. Each sample is transferred to a 50 ml cylinder
containing a small per-fluorocarbon stir bar. The diam-
eter and length of the pH probe must be such that it
can be placed into the solution during aeration. The
taller the water column and the smaller the bubbles, the
better the stripping will be. Each solution should be
moderately aerated (approximately 500 mL air/min) for
a standard time, such as 60 min. Formation of precipi-
tates may or not be important and should be noted.
The pH of the acidic and basic effluent and dilution
water aliquots should be checked every 5 min during
the first 30 min of aeration and every 10 min thereafter.
If the pH of any pH 3 or pH 11 solution drifts more than
0.2 pH units, it must be readjusted back to the nominal
value. The volume and concentration of additional acid
and/or base added to the solutions should be recorded
so that the final concentration of Na* and Cl- in each
solution can be calculated following final pH readjust-
ment. Solutions should be stirred slowly during any pH
readjustment. Again, precautions must be taken in or-
der to minimize the amount of acid and base added.
Note that the aeration time does not include the time
intervals during which aeration is temporarily discontin-
ued to readjust pH. A constant pH is not maintained in
the "pHz" effluent because this solution represents the
generalized effects of aeration on the effluent without
regard to pH. Only slight changes in the pH of the
dilution water at its initial pH are expected since such
water is usually at air equilibrium before the start of the
manipulation.
The sparged sample must be removed from the
graduated cylinder for toxicity testing so that any toxi-
cant that may have been sublated is not redissolved in
the sparged sample. This may happen if the sample is
simply poured from the cylinder, and sublation would
never be suspected. Therefore, one way to transfer the
effluent sample is by pipetting it out of the cylinder,
exercising care to prevent any sample from contacting
the sides of the cylinder above the liquid level. For
example, when using a 100 mL graduated cylinder for
the aeration vessel, a 50 ml pipette can be used to
remove the 30 mL sample. At this point it may be
possible to recover a sublated toxicant from the sides
of the cylinder. This must be done at the end of the
aeration step; see Results/Subsequent Tests section
below for details.
Sparging air contaminated with oil (droplets or va-
por) or any other substance is unacceptable. Contami-
nated air is probable from air lines containing oil or in
cases where the source of the air is contaminated (e.g.,
boiler room). Small air pumps, sold for home aquaria
are adequate, but only if the room air is free of chemi-
cals or contaminants. Chemistry laboratories where con-
centrated chemicals and solvents are used often might
not have suitable air quality.
Following aeration, the pH of each solution (includ-
ing the 35 mL portions of pH unadjusted effluent (pHz)
and dilution water) is returned to the pH of the initial
effluent or dilution water using the necessary volumes
of NaOH and HCI. Returning all effluent solutions to the
initial pH of the wastewater will ensure that a valid
comparison can be made with the baseline LC50. The
pH of each sample must be checked periodically
throughout the remaining work day and readjusted as
necessary. If stable pH can be attained prior to toxicity
test initiation, the pH during the test is likely to change
less.
Day 2: Before initiating the toxicity tests, the pH of
all of the aerated effluent and blank solutions should be
checked and adjusted to pHz. Toxicity tests are per-
formed on a single 100% concentration of all three
dilution water blanks (pH 3, pHz'and pH 11). Tbese
dilution water blanks will provide information on toxic
artifacts resulting from aeration.
Based on the 24-h initial LC50 of the day 1 effluent,
toxicity tests are performed on each aerated effluent
solution at concentrations of 4x- (or 1 00%), 2x-, Ix-,
and 0.5x-LC50 (cf., Section 8.2). The pH of each test
concentration should be measured and recorded daily.
An example of the data sheet for the aeration test is
given in Figure 8-I 1.
Interferences/Controls and Blanks:
Dilution water controls prepared for the baseline
toxicity test also act as controls on organisms, dilution
water and test conditions for this test. Results of the pH
adjustment test (Section 8.3) will suggest whether or
not toxic levels of NaCI may have been reached as a
result of the addition of acids and bases to the effluent.
No significant mortality should occur in any of the
three aeration blanks. If there is significant mortality,
the cause must be found and corrected before the test
can be meaningful. To determine which factor(s) caused
blank toxicity, the toxicity of pH adjusted aerated dilu-
tion water can be compared to that in the same pH
adjusted, unaerated dilution water. Approximately the
same quantities and concentrations of acid and base
should be added to both of these samples of dilution
water to make them comparable. Blank toxicity in the
pH adjusted and unadjusted aerated dilution water sug-
gests contaminated air. Other possible causes include
contaminated equipment, such as electrodes or glass-
ware (especially where low or high pH solutions were in
contact), or the addition of too much acid or base.
Another approach to this blank question involves evalu-
ating the pH adjustment/filtration and pH adjustment/
C1( SPE blanks (Sections 8.4 and 8.6). Assuming the
8-22
-------
o
I
m
co
2
0)
O
X
-S>
Q.
CO
(f>
E
LU
M
«
I
8
I
a
•Q
CL
O»
1
a>
a
2
o>
a
O
I
IT)
CO
I
-------
Figure 8-11. Example data sheet for aeration test.
Test Type: Aeration Species/Aqe: .
Test Initiation (Date & Time): No. Animals/No. Reps:
Source of Animals:
Investiaator: Dilution Water/Control:
Sample Log No., Na
Date of Collection:
me: Test Volume:
Other Info:
pH/Conc.
(% effluent)
314-x-LCSO
3/2X-LC50
3llx-LC50
3/0.5x-LC50
3 1 blank
pHi/4x-LC50
pHi/2x-LC50
pHi/lx-LC50
pHi/0.5x-LC50
pH il blank
11/4X-LC50
11I2X-LC50
11/lx-LCSO
ll/0.5x-LC50
111 blank
0 h
PH
Survival Readings:
24 h
A pH DO
48 h
A pH DO
72 h
A pH DO
96 h
A pH DO
Note: See baseline data sheet for control data.
Volumes and Strength of Solutions Added:
HCI NaOH
pH 3
pHz
pH ll
Comments:
8-24
-------
concentration of acid and base in the final blank solu-
tion is approximately the same in all dilution waters for
the three tests, toxicity in the aeration blanks but not in
the filtration or C1g SPE blanks suggests that aeration
rather than pH adjustment has led to contamination.
Compare the toxicity from the effluent baseline test to
the toxicity of all three aerated effluent samples. When
the baseline toxicity is significantly less than that of any
one of the aerated samples, toxicity was added or
created during effluent manipulations. This check is
especially important because pH adjustment of aerated
effluent may have required larger quantities of acid and
base as compared to the pH adjustment test (Section
8.3).
If nitrogen was used for stripping, DO depletion is
likely to have occurred. If a relatively large surface-to-
volume ratio is used (such as the 10 ml volume in a 1
oz plastic cup) during the overnight holding period, DO
should not be a problem.
Results/Subsequent Tests:
The LCSOs for the aliquots of pH 3, pH i, and pH 11
aerated effluent are compared to the baseline effluent
LC50 to determine whether any of these processes
resulted in a significant change in effluent toxicity. If a
substantial reduction in toxicity is seen for any or all of
the three aerated effluent solutions, one must next
determine whether the separation of effects was caused
by sparging, oxidation, or sublation. This is done by
repeating those tests in which toxicity was reduced,
substituting nitrogen for air in the stripping process.
Use of nitrogen eliminates oxidation as a removal pro-
cess. If side-by-side effluent stripping tests with air and
nitrogen provide the same results, toxicant removal is
probably caused by the sparging process. If only the
test(s) conducted with air succeeds in reducing or re-
moving effluent toxicity, oxidation is a probable cause.
An effluent sample may contain toxicants removed
through sparging and oxidation. An example of this is
where aeration at pH 3 and pH j reduces toxicity, but
nitrogen stripping removes the toxicity only in the pH 3
effluent. Using the Procedure described above ammo-
nia should not be air-stripped; however, if different
aeration vessels are used and greater surface area is
used, ammonia can be reduced/stripped. If toxicity is
reduced at pH 11 compared to pH/aerated, pH 3
aerated samples, or the baseline test, measuring the
ammonia concentration after aeration c.an be informa-
tive.
To determine if toxicity is due to sublatable com-
pounds, toxicant recovery can be attempted by adding
dilution water to the emptied cylinder used for sparging
(preferably graduated cylinders with ground glass stop-
pers), stoppering it -and shaking it vigorously, making
sure that the sides of the cylinder are thoroughly rinsed
with the dilution water. This dilution water is then tested
for toxicity. Recovery of the sublated material provides
further evidence that a surface active compound was
present. A more concentrated solution of the sublated
material can be obtained by using a larger sample
volume for sparging, such as 90 ml, and less dilution
water (i.e., 30 mL) to recover the toxicant( This
would result in a nominal concentration of 3x the whole
effluent concentration of the toxicant( To avoid spuri-
ous results in cases where sublated toxicity is recov-
ered, a dilution water blank should be run. A dilution
water sample should be subjected to the sparging step
and to the toxicant recovery step. In that way, if toxicity
is inadvertently being added to samples during the
manipulation (from contaminated glassware or contami-
nated air or nitrogen supply) it should occur in the blank
sample. As some sublated compounds are difficult to
recover, from the glass surface, a solvent rinse with
methanol may result in more efficient recovery. If sol-
vents are used and because of the low concentration
factor involved, most of the solvent will have to be aired
down in order to have an adequate water concentration
to perform the toxicity test. If solvents are needed, one
is wise to scale up the volumes to obtain higher con-
centrations for testing. However, not all kinds of surface
active agents are removed to the same degree, and
some are not removed at all (Ankley et al., 1990). This
is probably due to factors such as matrix effects and
solubility. Recovery of a sublated toxicant can be diffi-
cult. Consequently, reduction of toxicity by sparging
with air and nitrogen can be an indication of a Wxicant
which is a surface active agent, but a lack of toxicity
removal does not rule out the presence of these com-
pounds.
In the pH adjustment/aeration test, removal of toxi-
cants by precipitation resulting from pH change alone
should also be detected by the pH adjustment/filtration
test. Oxidation of compounds can cause precipitation. If
oxidation is the cause, the pH adjustment/filtration test
will not change toxicity. If nitrogen sparging has re-
moved the toxicant, the "volatile toxicant transfer" ex-
periment described below may provide separation of
the volatile toxicant from other constituents. Our experi-
ence with this technique is limited to a few effluents. To
perform the "volatile toxicant transfer" experiment, a
closed loop stripping/trapping apparatus is used (Fig-
ure 8-12). This apparatus consists of a pump which can
circulate air or nitrogen gas, two airtight fluid reservoirs,
perfluorocarbon tubing, and diffusers. The arrangement
should be such that air or nitrogen can be passed
through the effluent in one reservoir and then through
the dilution water in the second reservoir before cycling
back to the first reservoir. The reservoir of the dilution
water serves as a trap that will collect the volatilized
toxicant( Of utmost importance to this experiment is
an air tight system. The time to equilibrium of the
volatile toxicant will be dependent upon the efficiency
of the sparging process and the rate of volatilization of
the toxicant which may be affected by pH. For ex-
ample, using a glass or plastic pipette to aerate the
samples may not effectively sparge the entire volume
of sample. To optimize the toxicant recovery, use of
gas washing bottles (for example, 125 nil and 500 ml
bottles from Kontes Glass Co., Vineland, NJ) fitted with
glass frit diffusers is suggested because they sparge
the sample volumes more effectively.
8-25
-------
Effluent
-X
Dilution
Water
Figure 8-12. Closed loop schematic for volatile chemicals.
Numerous operating conditions can be selected,
each providing different information. This system should
not be operated as a conventional purge and trap
system. The reason is that since one does not know
the identity of the toxicant( the conditions for trapping
are not known. Initially, the objective should be to get
measurable toxicity moved into the dilution water me-
dium in the trap. This will establish that there are at
least some volatile toxicants present. At this stage the
goal is not to move all the toxicant(s) to the dilution
water in the trap. If the same concentration of the
toxicant in the effluent can be transferred to the dilution
water as exists in the unaltered effluent, the data are
easiest to interpret. For this purpose the volume of
sparged effluent should be large and the dilution water
volume in the trap small. The nitrogen gas is recirculated
so that if the trap is inefficient in removing the toxicant(s)
from the nitrogen, the toxicant(s) will not be lost from
the sample. Because conditions to optimize transfer
cannot be selected until the chemical identity is known,
longer sparging times should be chosen,
The first experiments should involve no pH changes
if any measurable change in toxicity occurred in the
earlier tests without the pH 3 orpHll adjustments.
The reason for this selection is that drastic changes in
pH can cause many unknown effluent changes, and
artifacts are more likely to occur. Of course if pH changes
are required to change toxicity, then pH will have to be
altered. When pH is altered, then equilibrium objec-
tives, mentioned above, are not possible and the entire
process takes on characteristics of more conventional
8-26
-------
purge and trap experiments. The usual resin traps
described in EPA methods are not suitable because
the trap cannot be tested for toxicity. The trapping
medium must be, or be able to be, made into a toxicity
testable water.
In those instances when sparging affects toxicity
only when accompanied by a pH change (pH 3 or
pHI i ), the method to be used to operate it as a
conventional purge and trap is as follows. The trap's
dilution water volume should be small relative to the
sample volume and its pH should be opposite that of
the sample pH (e.g., if the sample pH is 3, then the trap
pH should be 11). One can no longer conclude any-
thing about the original effluent equilibrium, and the
procedure is one of separation. Toxicity in the trap may
or may not be caused by the same substance as that
which causes the original effluent toxicity. Obviously, all
the precautions mentioned above regarding NaCI addi-
tion and other adjustments must be tracked with blanks
just as in any other experiment. We have not found
many effluents where the transfer technique is useful,
but for those effluents where it works, it is a powerful
tool. We have found the volatile toxicant transfer ex-
periment to be useful with some samples (i.e., sediment
pore water), where two pH dependent toxicants (e.g.,
ammonia and hydrogen sulfide) are suspect. There is
sometimes an appreciable loss of toxicity after the pH 3
aeration step in samples with ammonia toxicity, yet it is
unknown whether the toxicity loss is due to volatiliza-
tion of hydrogen sulfide (or some other pH dependent
toxicant) at low pH, or is an artifactual decrease of
ammonia toxicity due to a downward pH drift in the test
(cf., Section 6.3). In this case, a trap such as the one
described previously for transferring a volatile toxicant(s)
at altered pH is useful. Water in the trap that volumetri-
cally concentrates the toxicant at two or more times its
whole sample concentration may be successfully tested
for toxicity. We suggest, in the case of suspected hy-
drogen sulfide toxicity, testing the trap water at pH 6,
as the toxicity of hydrogen sulfide is enhanced at that
pH. One caution in this setup is that the volatility of
some pH dependent toxicants such as hydrogen sulfide
makes it imperative that the experiment be initiated
immediately after adjusting the pH to minimize their
loss.
8.6 pH Adjustmen t/C,, Solid Phase
ously, relative degrees of water solubility exist, Many
highly toxic pollutants found in effluents at very low
concentrations are not considered to be water soluble
despite the fact that they are present at toxic concen-
trations.
Compounds extracted by the
PH
C,,chloroform. The
pH a n d a pH,i i g h
u m n
B C]g c a
C1§ S P E co
pH'smn degradation,
pHed. To ensure column integrity, the
pHbe used on the SPE columns will be either
pH pH11) in this manipulation.
Manufacturer's data should be consulted for tolerable
column pH ranges and for exact column conditioning
ml. the 235 pH) filtered
8.4), 3 5nL
pH) i s h e I d pHo r the
mLit. Now, the additional 200
C1gough the (pH 3 pH i,
pHd
mLre concentrations (10
use
Obvi-
pHphdjusted or i
on the toxicity of the effluent (the 24-h initial LC50).
For the blanks, 30 mL of pH adjusted (pH 3 or pH
11) and/or filtered (pH i) dilution water is needed. The
last 10 ml of the post-column water should be used for
blank toxicity tests.
Apparatus:
Six 250 mL graduated cylinders, eight 25 mL gradu-
ated cylinders, glass stoppered bottles for acid and
base solutions, pH meter and probe, stir plate,
perfluorocarbon stir bars, pump with sample reservoir,
perfluorocarbon tubing, ring stands, clamps, three 3 mL
C SPE columns (200 mg sorbent), automatic pipette
(10 mL), disposable pipette tips (10 mL), eye dropper
or wide bore pipette, 30 mL glass beakers of 1 oz
plastic cups, light box and/or microscope (optional).
Reagents:
HPLC grade methanol, high purity water, 1 .0, 0.1,
and 0.01 N NaOH, 1.0, 0.1, and 0.01 N HCI (ACS
grade in high purity water), buffers for pH meter calibra-
tion, acetone and methanol for cleaning the pump and
reservoir, and vials to collect methanol eluate.
Test Organisms:
Test organisms, 135 or more, of the same age and
species.
post-coh
a-27
-------
Procedure:
Day 1: This procedure is performed with effluent
samples adjusted to the various pH's; however, the
manipulations have three distinct steps (Figure 8-13)
which are generally the same for each pH. Prior to
attaching a new column to the apparatus, the reservoir
and pump must be cleaned with acetone, methanol,
and high purity water.
Step 1 involves conditioning the solid phase extrac-
tion columns for each pH. Column conditioning proce-
dures may vary with the manufacturer of the column.
The procedures described below are modifications of
the conditioning steps used with Baker* C,, SPE col-
umns (J.T. Baker Chemical Company, Phillipsburg, NJ).
Using a flow-rate of 5 mL/min,15mL of HPLC
grade methanol is pumped through the column and
discarded. Next 15 ml of high purity water, adjusted to
pH 3 with HCI, is placed in the sample reservoir. Care
must be taken in timing the addition of solutions after
the methanol has passed through the column. While
the mixing of methanol with subsequent solutions must
be minimized, the column must also be prevented from
going dry following the methanol wash and dilution
water or sample application. The amount of time needed
between introduction of solutions to prevent any col-
umn drying will be unique to each investigator's appa-
ratus. This timing should be determined before per-
forming this procedure with actual effluent samples. If
the column dries at any time after introduction of the
methanol during conditioning, the column must be re-
conditioned (with methanol).
As the last volume of pH 3 high purity water is
entering the column (Step 1),thepH 3 adjusted, filtered
dilution water is placed into the reservoir (Step 2).
Again, the column must not be allowed to dry before
the pH 3 dilution water enters the column. The pH 3
high purity water passing from the column should be
measured to determine the point at which the dilution
water begins to leave the column. This pH 3 high purity
water used to condition the column is discarded, Next,
30 ml of the filtered pH 3 dilution water is collected,
and the last 10 ml aliquot collected is used for the
toxicity blank to detect toxicity leached from the col-
umn. This aliquot will have to be pH re-adjusted to the
initial pH of the dilution water using NaOH, and it is
reserved for day 2 toxicity testing. Care should be
taken to minimize changes in sample volume and ionic
strength during pH readjustment.
As the last several mL of filtered pH 3 dilution water
are entering the column, the 200 ml volume of filtered
pH 3 effluent is placed in the sample reservoir (Step 3).
Again, the column sorbent must not be allowed to dry
between the dilution water blank and the effluent. Col-
lect a 30 ml aliquot of post-column effluent after 25 ml
of the sample passes through the system. A second
post-column 30 ml aliquot is collected after a total of
150 ml of the sample passes through the column.
Collection of the first post-column sample after 25 mL
of sample has passed the column ensures that any
dilution water left in the system will not be present in
the post-column sample. The second subsample of
post-column solution provides information on column
overloading and toxicant these
30 ml pH i u s i n g t \
NaOH. The total NaOHJ m e <
necessary forpH adjustment sho uThfeseb e re
aliquots are reserved for day 2 toxicity testing. Columns
re-used but should be saved
elution (see the Results/Subsequent Tests section).
Receiving water should not be used as the dilution
water because trace organic and metal contaminants
or organics (such as humic acid) may be present. If for
any reason, such a water is needed, the same column
should not be used for concentrating the toxic sample.
ml dilution water ml lank, '.
pH checked for toxicity. The
For pH /, the above procedure is repeated using a
clean reservoir and pump and a new conditioned^ ml
C . SPE column for the filtered phi effluent (Figure 8-
13). The pH of the post-column dilution water and post-
column effluent should be measured.
In the final C1J( SPE manipulation, pH 9 (readjusted
from pH 11) dilution water and effluent are processed
as described above. While use of pH 11 effluent offers
the likely advantage of shifting a larger number of basic
organic compounds farther towards the predominately
unionized form, and therefore removal, the JI SPE
column cannot withstand a pH above 10. For this rea-
son, the pH 11 filtered dilution water and sample aliquots
prepared in Sections 8.3 and 8.4 are readjusted to pH
9 with HCI before they are put through the column. The
15 mL of high purity water used to rinse the column
following methanol conditioning must also be read-
justed to pH 9 with NaOH. The 10 mL aliquot of post-
column pH 9 dilution water and both 30 mL aiiquots of
post-column pH 9 effluent are further adjusted to their
pH i's respectively, prior to toxicity testing. The total
volume of acid added for pH readjustment is recorded.
The pH of all aliquots of the chromatographed dilution
water and effluent should be checked and readjusted
as appropriate throughout the remainder of the work
day.
Day 2: The pH of all of the post-column dilution
water and effluent aliquots should be checked and
readjusted ifpH has drifted overnight. Toxicity tests are
performed on a single 100% concentration of all three
of the dilution water blanks. These blanks will provide
information on the presence of toxicity leached from the
C,8 column at different pH's.
The six 30 mL post-column effluent aliquots are
tested for toxicity using an exposure series based on
the 24-h LC50 of the original effluent. Chromatographed
effluent aliquots are tested at concentrations of 4x-,
8-28
-------
Figure 8-13. Step-wise diagram for preparing the C,, solid phase extraction column samples.
step 1
step 2
step 3
Prepare three Ci 8 SPE
• 15 ml methan
• 15mLhlghpui
DONOTLETSORBE
1 Columns
ol o
Ity water v\
ki-r /"»/"» RDV
NT oU UKY
I
HCI -*• t
30 mL at pH 3
Rltered Water
f
Prepared Column
i
Collect 10 mL Sample
NaOH -+•
\ i
Toxicity Test
HCI -». t
200 mL at pH 3
Rltered Water
t
Prepared Column2
from Step 2
|
Collect 30 mL Sample
a tier 25 mL and 150 ml
NaOH -»•
! I
ToxicityTest3
Dilution Water
1 1 !
t
30 mL at pH i
Rltered Water
f
Prepared Column
*
Collect 10mL Sample
I
f
Toxicity Test
Effluent Sample
' I '
i
200 mL at pH i
Rltered Water
t
Prepared Column?
from Step 2
|
Collect 30 mL Sample
after 25 mL and 150mL
1 '
ToxicityTest3
scard Methanol &
fater After Rinses
1
F r\<»«
30 mL at pH 9
Filtered Water
1
Prepared
Column
1
Collect 10 mL Sample
\
-*- HCI J
1
1
Toxicity Test Day 2
'
f -*-NaOH
200 mL at pH 9
Rltered Water
" 1
Prepared
i
Oi\
Column2
gp^
f
Collect 30 mL Sample
after 25 mL and 1 50 mL
i
^.HCI
i ^^..^ ,^__
M
1
Toxicity Test3 Day 2
|
11f column will be eluted with 1 mL methanol (cf., Results/Subsequent Tests), collect methanol column blank
before dilution water is passed over column. Column should go to dryness and will have to be
re-conditioned (Step 1) before proceeding to Step 2.
2 Use same column used with the dilution water unless receiving water is used (see text for details).
3 Same test as depicted in Figure 8-8.
8-29
-------
2x-, lx-,0.5x-LC50 (cf., Section 8.2). The pH of each
solution tested should be measured daily and recorded
along with organism survival. A sample data sheet for
the C,a SPE test is shown in Figure 8-14.
Interferences/Controls and Blanks:
Controls on test organism performance, dilution
water quality, and test conditions are provided by the
control from the baseline toxicity test. The pH adjust-
ment and filtration tests (Sections 8.3 and 8.4) provide
information on the effects of pH adjustment and filtra-
tion on eff iuent toxicity apart from any additional changes
caused by C18 SPE. Effluent and blank test results from
these two tests must be evaluated prior to interpreting
the results of the C18 SPE test, both in terms of identify-
ing any toxic artifacts added during filtration and pH
adjustment and in allocating toxicity reduction to the
three components potentially impacting effluent toxi-
cants in the C)8 SPE test.
Of those methods discussed so far, the C,a tech-
nique requires the greatest manipulation. More prob-
lems are likely to be encountered with toxic blanks
because in addition to those factors associated with pH
adjustment and filtering, the C method also involves
use of resin and methanol. Blanks for toxicity must be
chec: ed in the same manner as before for acid and
base addition, filter artifact toxicity, pH drift, as well as
toxicity from the C,, column. In addition to these, some
effluents behave in a peculiar way after passing through
the SPE column (cf., discussion below).
Results/Subsequent Tests:
The above unique properties of some effluents and
the potential for column blank toxicity problems make
interpretation of the test results more subjective.
If toxicity is not reduced in post-column effluent, not
too much credence should be placed on the results.
One needs to go back and sort through the possible
causes.
If none of the Phase I treatments reduced toxicity
(including the C SPE column) or if the toxicity was
reduced by the C18 column, it is useful to elute the
column with 100% methanol. Of course, a column blank
must also be evaluated; this is a methanol elution
following the column conditioning with methanol and
high purity water. The column must go to dryness
before collecting the blank of methanol. After the metha-
nol blank is collected, the column must be recondi-
tioned as described in Step 1 in Figure 8-13 and the
column must not go to dryness before starting the
dilution water over the column. If a 1 ml volume of
methanol is used (for 200 ml of effluent on a 3 ml
SPE column) and the sorption and elution efficiency
are 100%, any substances retained by the columns will
b.e concentrated 200x. To test the eluate, 150 nL of the
methanol fraction is diluted to 10 mL with dilution water.
The resultant methanol concentration is 1.5%, which is
below the 48-h or 96-h LC50 for all species given in
Table 8-3. This provides a concentration of effluent
constituents 3x whole effluent concentration. This small
amount of concentration over whole effluent allows
detection even if some loss occurred either in sorption
or elution.
If the post-column effluent is not toxic or less toxic,
and the methanol eluate is toxic, the next step is to
proceed with the Phase II C,. SPE procedure to identify
the toxicant(s) removed by the column. At this point it
should not be assumed that toxicity removed by the C,,
SPE column is due to non-polar organic compounds.
While metals are not non-polar organic chemicals, they
can be removed from some effluents using the C,a SPE
column. However, metals generally are not eluted with
methanol and therefore the fractions are not toxic. Met-
als may be eluted with dilution water adjusted to pH 3
or pH 11. Surfactants can be sorbed by the C18 SPE
column just as other non-polar organics, and some
elute with methanol.
If neither the post-column effluent nor the methanol
fraction is toxic, the toxicant is probably retained by the
C,8 SPE column but is not covered by the methanol
elution. When interpreting these results it is important
to consider that when a sample is passed over the C,a
column there can be other mechanisms besides re-
verse phase SPE by which a toxicant(s) can tje re-
moved. For example, the C18 column packing may
remove toxic compounds by filtering them out of solu-
tion, e.g., the toxicant may be associated with solids in
the effluent, and the 40 (im C1S packing material may
remove the solids. The toxic compounds could also be
physically adsorbed or ionically bound onto the surface
of the column packing and the methanol elution cannot
recover the toxicant from the column. Perhaps the
toxicant(s) has been removed by the reversed phase
SPE mechanism but the methanol does not recover the
toxicant(s) because either methanol is too polar a sol-
vent or the toxicants have too low a solubility in metha-
nol. In this case a different solvent system may be
needed to remove the toxicants, e.g., methylene chlo-
ride, hexane or pH adjusted water. Both hexane and
methylene chloride are much more toxic than metha-
nol, and if hexane or any other "toxic" solvent is used,
solvent exchange or some other method must be used
to remove the solvent in order to effectively track toxic-
ity throughout the procedure. Another possibility that
should be considered is that the toxicant has decom-
posed. Whatever the mechanism of toxicant removal,
the interpretation of the loss of toxicity should be evalu-
ated carefully.
After passage through the Cia column, some efflu-
ents exhibit artifactual toxicity which is not observed in
the post-column dilution water blanks. Artifactual toxic-
ity can arise from two sources: a) pH drift and/or b)
biological growth in the post-column effluent. (Problems
with pH drift are discussed in Section 8.3; consult that
section for further information.) Artifactual toxicity from
biological growth in the post-column effluent can be a
major problem for some effluents, particularly municipal
effluents. This growth has not been observed in all
effluents, but for the effluents where it did occur, it was
present in nearly every post-column sample. The post-
8-30
-------
^J
0
Q.
0
1
T3
(0
£
a)
c
•Q
(0
h»
X
a
8
a
H- h
1
1
o
CO
1
1
.0
Q
C
c
d
e
^
n>
P
03
n
c
a
c
•>
-------
Ta ble 8-3. Toxicity of methanol to several freshwater specie*
Species
Ufeslage
LC50 (%. v/v (95% CD)
24 h
48 h
72 h
96 h
Cenodaphnia dubia
Daphma magna
Daphnia pu/ex
Hyaleiia azteca
Salt-no gairdneri
Pimephaies promeias
Lepomis macrochirus
<6 h'
124 h1
<48h'
124 h2
<14h4
juvenile'
juvenile'
S24h'
28-32 d6
juvenile8
>3.0
2.7
(2.6 - 2.9)
2.4
(2.2-2.6)
NR
2.66
(2.3-2.8)
2.5s
(1.9-2.8)
2.5
(2.5-2.7)
4.0
3.8
(3.7-3.9)
2.4
(2.2-2.7)
>3.0
2.7
(2.6 - 2.9)
2.0
(1 .9-2.2)
(2.5-3.7)
NR
NR
2.5
(2.5-2.7)
4.0
38
(3.7-3.9)
2.4
(2.2-2.7)
NR NR
NR 2.5
(2.5-2.7)
3.7 3.7
(3.2-4.2) , (3.2-4.2)
NR 3.7
(3.6-3.9)
NR 1.9
(1 .8-2.3)
1 Data generated at ERL-Duluth. C. dubia were <24 h old at test initiation and fed. (Tested in soft reconstituted water (DMW); static
unmeasured.)
2 Randall and Knopp, 1980. (Tested in spring water: static and unmeasured.)
3 48-h EC50.
1 Bowman et al., 1981. (Tested in well-water: static and unmeasured.)
5 18-hLC50.
' Poirier et al., 1986. (Tested in Lake Superior water; flow-through and concentrations measured.)
Note. (-) = Confidence interval cannot be calculated as no partial mortality occurred; NR = Not reported.
column samples exhibited a turbid, often filamentous
growth and sometimes, lower than normal DO levels in
the toxicity tests. In one effluent, this growth was caused
by methylotrophic bacteria. Methanol occurs in post-
column effluent samples because the methanol used in
conditioning (activating) the C,8 SPE column is slowly
leached out of the column and into the effluent as it
passes through the column.
Methods for eliminating or controlling this type of
artifactual toxicity problem are currently limited. For
some effluents, the most promising method appears to
be additional filtering of the post-column effluent through
a 0.2 (im filter to remove bacteria prior to testing. Whole
effluent filtered through a 0.2 urn filter serves as a
control for the toxicity test with the post column/filtration
manipulation. Filtration is easy to perform and allows
useful post-column toxicity data to be obtained, pro-
vided it does not alter or reduce toxicity in the post-
column effluent. If toxicity is removed in the 0.2 u.m
filtered post-column effluent, but not in the 0.2 u.m
filtered whole effluent, repeat the experiment filtering
the whole effluent with 0.2 u.m filter and testing the
post-C,, 0.2 urn filtered effluent. The post-column (0.2
urn filtered effluent) may need to be filtered (0.2 urn)
again. If toxicity is removed by filtration, see Section
8.4 Results/Subsequent Tests. If this growth in the
post-column cannot be eliminated but toxicity occurs in
the methanol eluate, then proceed with Phase II identi-
fication. If growth is not eliminated and no toxicity
occurs in methanol eluate, then use of different sol-
vents to condition the Clg column may reduce growth
(e.g., acetonitrile). Control of the turbid growth may
also be possible by performing daily renewals with
post-column effluent. Initially, more post-column sample
would have to be collected (60 ml rather than 30 ml),
and a portion should be refrigerated. If control of the
artifactual toxicity caused by the turbid growth cannot
be achieved, other sorbents (e.g., XADs, activated car-
bon) may have to be used. Another possible method of
controlling the growth may be by the use of antibiotics
but we have not investigated this approach.
Observation and judgement must be used to detect
problems occurring from artifactual toxicity and only
through experience can one recognize when they oc-
cur. Failure to recognize them will result in the conclu-
sion that the C18 SPE column did not remove toxicity
when it in fact may have done so.
If toxicity occurred in the methanol eluate from a
POTW effluent, and the C. dubia were more sensitive
than fathead minnows, it might be cost-effective to try
adding a metabolic blocker, piperonyl butoxide(PBO),
to the effluent and eluate. We have frequently found
non-polar organics in POTW effluents and have identi-
fied organophosphate pesticides (OP's) as the toxicant(s)
8-32
-------
to C. dubia (Amato et al., 1992; Norberg-King et al.,
1991). Most metabolic blockers used in aquatic toxicol-
ogy have been used with fish; however, OP's are gen-
erally less acutely toxic to most fish than to cladocer-
ans. PBO is a synthetic methylenedioxyphenyl that can
block the toxicity of various chemicals that need to be
metabolized in the cytochrome P450 cycle to be toxic.
In tests with cladocerans, sublethal additions of PBO to
the whole effluent and/or the methanol eluate test have
been useful for implicating some metabolically acti-
vated OP's as the toxicant(s) (Ankley et al., 1991).
Experiments showed that for C. dubia, D. magna, and
D. pulex, PBO blocked the acute toxicity of parathion,
methyl parathion, diazinon, and malathion, but not di-
chlorvos, chlorfenvinphos, and mevinphos. For those
OP's where PBO reduced the toxicity, the reduction
was greater in the first 24 h; the toxicity of the OP's in
an effluent may be expressed after 24 h.
To perform the PBO addition test for cladocerans,
a PBO water stock is prepared and microliter quantities
are added at various sublethal concentrations (final
concentrations of PBO are 500, 250, and 125 jig/L).
(Note: the 48-h LCSO's for C. dubia, D. magna, and D.
pulex are 1,000, 2,830, and 1,620 u.g/L, respectively).
The PBO additions can be set up in a similar manner to
the EDTA and oxidant reduction tests (Section 8.7 and
8.8) using a 3 x 3 matrix of PBO and effluent concen-
trations. Toxicity reduction with the addition of PBO
would suggest the presence of toxic levels of metaboli-
cally-activated compounds such as OP's. However if
toxicity was not changed, it does not mean those types
of compounds (i.e., OP's) will not be present. Further
tests with PBO will be described in the second edition
of Phase II.
8.7 Oxidant Reduction Test
Principles/General Discussion:
This test is designed to determine to what extent
constituents reduced by the addition of sodium thiosul-
fate (Na2S203) are responsible for effluent toxicity. Chlo-
rine, a commonly used biocide and oxidant, is fre-
quently found at acutely toxic concentrations in munici-
pal effluents. Chlorine is unstable in aqueous solutions
and decomposition is more rapid in solutions when
chlorine is present at low concentrations. Phase I initial
aeration tests will provide information on chlorine toxic-
ity as will the oxidant reduction test. However, this
oxidant reduction test does not simply affect chlorine
toxicity. Also neutralized in this test are other chemicals
used in disinfection (such as ozone, and chlorine diox-
ide), chemicals formed during chlorination (such as
mono and dichloramines), bromine, iodine, manganous
ions, and some electrophile organic chemicals. Fre-
quently, the reduced form of the toxicant has a much
lower toxicity.
Although the thiosulfate addition test was initially
designed to determine if oxidants (such as chlorine) are
responsible for effluent toxicity, thiosulfate can also be
a chelating agent for some cationic metals. Conse-
quently, reductions in effluent toxicity observed with
this test may be due to the formation of metal com-
plexes with the thiosuffate anion (Giles and Danell,
1983). Cationic metals that appear to have this poten-
tial for complexation (based upon their equilibrium sta-
bility constants) include cadmium (2*), copper (Cu2*),
silver (Agu), and mercury (Hg2*) (Smith and Martell,
1981). However, the rate of formation of the complex is
specific for various metals and some cationic metals
may not be rendered non-toxic in the 48-h or 96-h
period used for the toxicity test due to a slow complex-
ation rate.
Recent work using C. dubia has shown that sodium
thiosulfate (and EDTA) can remove the toxicity of sev-
eral cationic metals (Hockett and Mount, in preparation)
from dilution water and effluents. The toxicity of copper,
cadmium, mercury.-silver and selenium (as selenate) at
4x the 24-h LC50 of each in moderately hard reconsti-
tuted water was removed by the levels of thiosulfate
typically added in this test. Mercury toxicity was re-
moved with the addition of thiosuffate for 24 h but not
48 h, indicating it may not have been completely
complexed by the thiosulfate. In addition, tests with
zinc, manganese, lead, and nickel and thiosulfate, indi-
cated that the metal toxicity was not removed by thio-
sulfate. However, with these metals and the addition of
EDTA, the toxicity to C. dubia was cpmplexSb (cf.,
Section 8.8, EDTA Test). Knowing which metals are
bound by both thiosulfate and EDTA, and which metals
are complexed with only one or the other additives can
be very helpful in narrowing down the possible toxicant.
Data on the toxicity of sodium thiosulfate to
Ceriodaphnia dubia, Daphnia magna and fathead min-
nows are given in Table 8-4. Data generated at ERL-D
show that for Ceriodaphnia, both feeding and lower
hardness waters results in greater thiosulfate toxicity,
and this trend appears to be the same for fathead
minnows (Table 8-4). In effluents, some of the added
thiosulfate will combine with certain oxidants present,
thereby lowering the concentration of the reactive and
toxic thiosulfate. Therefore, the LC50 values indicate
that less toxicity due to thiosulfate (Table 8-4) might be
expected in effluents than in dilution water (i.e., recon-
stituted water) where no oxidants are present to react
with the thiosulfate. More importantly, when an effluent
concentration of 4x the LC50 is tested, toxic oxidant
levels should not be excessively high. As a result there
should not be a need to add very large amounts of
thiosulfate to neutralize toxic oxidants in the test solu-
tion.
Additions of sodium thiosulfate for this test can be
approached in either of two ways; a gradient of thiosul-
fate concentration can be added to several test cham-
bers containing the same effluent concentration or as a
dilution test where a 3 by 3 matrix of effluent concentra-
tions and thiosulfate concentrations are used.
For the gradient approach, concentrations of so-
dium thiosulfate equal to and lower than the thiosulfate
LC50 for the test species being used are added to
several containers with effluent at the 4x-LC50 (or 100%)
concentration (cf., Figure 8-1 5). If the test species is
not listed in Table 8-4, the thiosulfate LC50 will have to
-33
-------
be determined. Time to mortality may also be useful in
addition to observing mortality at a fixed time (i.e., 24,
48 or 96 h). Time to mortality measurements are impor-
tant when no dilutions of the effluent are used.
The dilution approach has the advantage in that
LCSOs can be calculated to see how much the toxicity
was reduced. For this test a matrix of three effluent
concentrations and three levels of thiosulfate concen-
trations are used. The choice of the thiosulfate concen-
trations to add to the effluent is based on the thiosulfate
LC50 for the test species being used in an appropriate
dilution water (Table 8-4). Three sets of effluent solu-
tions (i.e., 4x-LC50,2x-LC50,1x-LC50or 1 00%, 50%,
25%) are prepared. To the first set, thiosulfate is added
to each test solution at one-half the thiosulfate LC50; to
the second set, thiosulfate is added at one-fourth (0,25x)
the LC50; and to the third set, thiosulfate is added at
0.125xtheLC50. In this approach the concentration of
thiosulfate remains constant over each effluent dilution
series. The test results are compared to the baseline
test result to determine the amount of toxicity removal.
For cases where oxidants account for only part of
the toxicity, sodium thiosulfate may only reduce, not
eliminate the toxicity. The thiosulfate addition test is
useful even when chlorine appears to be absent in the
effluent. As discussed above, oxidants other than chlo-
rine occur in effluents and this test should not be
omitted just because the effluent is not chlorinated.
Likewise, removal of toxicity by thiosulfate does not
prove that chlorine was the cause of effluent sample
toxicity. Refer to the Results/Subsequent Tests section
below for additional options.
Table 8-4. Toxicity of sodium thiosulfate to Ceriodaphnia dubia, Daphnla magna, and fathead minnows
Water
Species Type
Ceriodaphnia dubia' SRW
SRW
SRW
SRW
MHRW
HRW
VHRW
VHRW
VHRW
B annul a magna2 SRW
Pimephales prome/as' SRW3
SRW
MHRW3
MHRW
HRW3
HRW
VHRW3
VHRW
Life
Stage
S24h
12t h
<24h
S24h
S24h
<24h
S24 h
S2i h
S24 h
NR
<24h
<24h
12i h
<24h
<24h
£24 h
S24h
S2t h
24 h
2.5
( — )
1.3
(1.0-1.7)
1.5
(1 .2-2.0)
2.0
(152.7)
1.7
(1 .2-2.2)
6.6
(5.8-7.5)
5.0
(3.8-6.5)
6.6
(5.8-7.6)
5.0
(3.9-6.4)
2.2
(NR)
8.4
(7.6-9.3)
7.4
IB.3-S.B1
9.5
( — )
9.8
11.6
(9.7-I 3.9)
12.9
(11.8-14.1)
13.3
(11.8-15.7)
13.4
(11.8-15.3)
LC50 (a/I
48 h
0.85
I a. 72-1 .0)
a. ss
IB. 72-1 .1)
a. as
i a. 83-i .1)
084
(0.71-0.99)
a. as
IB.B2-I .6)
1.6
( — )
3.3
(2,5-4.3)
1.8
(1 .4-2.3)
1.2
(0.91-1.6)
1.3
(NR)
8.4
17. 6- 9. 31
7.1
(5.9-8.4)
8.9
(8.1-9.7)
9.8
8.5
(7.2-10.0)
12.9
(11.8-14.1)
7.8
(6.5-9.2)
12.1
(10.8-1 3.4)
) (95% Cft
72 h
7.9
(7.4-8.5)
7.1
(5.9-8.4)
8.2
(73-9.2)
9.8
7.9
(6.9-9.1)
12.5
(11.3-13.8)
6.7
(5.5-8.2)
12.1
(10.8-13.4)
96 h
*
7.3
(6.4-8.3)
5.9
(4.3-7.9)
4.7
(3.9-5.6)
9.8
7.9
(6.9-9.0)
11.7
(10.4-13.0)
6.1
(4.8-7.9)
12.1
(10.8-13.4)
' Data generated at ERL-Duluth; both species were <24 h old at test initiation and C. dubia were fed.
1 Dowden and Bennett, 1965.
3 Data generated at ERL-Duluth and values represent those from 7-d growth and survival tests and daily renewals.
Note: (-) = Confidence interval cannot be calculated as no partial mortality occurred; NR = Not reported; VSRW = very soft reconstituted
water; SRW = soft reconstituted water; MHRW = moderately hard reconstituted water; HRW = hard reconstituted water; VHRW = very
hard reconstituted water.
8-34
-------
Volume Required:
A maximum volume of 100 ml effluent is required
for the oxidant reduction test. The exact volume re-
quired will depend on the 24-h initial LCSO.For the
gradient addition, six effluent aliquots at 4x-LC50 or
(100%) are required, each having different thiosulfate
concentrations. For the dilution test, three sets of three
effluent solutions are prepared (Le., 4x-, 2x-, Ix- 24-h
LC50) and three different concentrations of sodium
thiosulfate (e.g., 0.25, 0.5, and 1.0 g/L) are each added
to a set of dilutions.
Apparatus:
Glass stirring rods, glass volumetric flask for so-
dium thiosulfate, 1 mL glass pipettes, automatic pi-
pette, disposable pipette tips, 10 -1000ul pipettes,
eye dropper or wide bore pipette, 30 ml beakers or 1
oz plastic cups, light box and/or microscope (optional),
pH meter and probe.
Reagents:
Regardless of whether the 1 x 6 gradient addition
test or the dilution test is to be done, the sodium
thiosulfate stock concentration should be 10x the so-
dium thiosulfate LC50 concentration for the test spe-
cies being used.
Test Organisms:
Test organisms, 50 or more, of the same age and
species.
Procedure:
Day 2: To perform the gradient thiosuifate addition
test, transfer six 10 ml aliquots of effluent diluted to 4x-
LC50 (or 100%) into six test chambers. Add 1 .0, 0.8,
0.6, 0.4, and 0.2 mL of the appropriate concentration of
the thiosulfate stock to five aliquots and mix. Do not
add any to the sixth. The container receiving 1 ml of
thiosulfate should now contain the approximate con-
centration of sodium thiosulfate equal to the LC50 of
the test species. Figure 8-15 contains an example form
for recording the data. A suggested schedule for ob-
serving time to mortality is shown on the data form.
To perform the thiosulfate dilution addition test,
prepare three sets of effluent dilutions (i.e., 4x-LC50,
2x-LC50, 1x-LC50) and add the appropriate amount of
thiosulfate (i.e., 0.5x,0.25x, and 0.125x thjosulfate LC50)
for the test species to each set of dilutions. Figure 8-16
is an example data sheet for the thiosulfate addition
test using this effluent dilution approach. The baseline
test conducted at the same time will provide informa-
tion on effluent toxicity without thiosulfate added.
interferences/Controls and Blanks:
Controls prepared for the baseline toxicity test act
as a check on the general health of test organisms,
dilution water quality and test conditions.
When the time to mortality in the various thiosulfate
exposure concentrations in the gradient addition test is
compared to the treatment without thiosulfate, one can
determine whether the addition of thiosutfate increased
the time to mortality at some thiosulfate concentration.
If, in all of the effluent exposures, the time to mortality
decreases, then thiosulfate is affecting toxicity. If all
test solutions cause mortality in the thiosuifate effluent
dilution test, but this trend does not occur in the baseline
test, the thiosulfate may be causing the toxicity. In
either case, the test should be repeated with weaker
sodium thiosutfate additions. If the toxicity is unchanged,
perhaps not enough sodium thiosutfate was added, and
the test can be repeated using a higher range of the
thiosulfate additions.
If a significant loss in effluent toxicity is apparent
over the first 24-h period after sample arrival in the
laboratory (i.e., initial LC50 c baseline LC50), it may be
necessary to conduct future oxidant reduction tests for
Phase I immediately upon arrival of the sample in the
laboratory.
Results/Subsequent Tests:
If oxidants are causing toxicity, time to mortality
should increase somewhere in the range of tested
thiosulfate additions or the toxicity should be reduced
from the baseline LC50. No change in toxicity suggests
either no oxidant toxicity or not enough thiosulfate was
added. The experiment should be repeated, increasing
the concentration of thiosulfate added.
When the LCSOs from the sodium thiosulfate addi-
tion dilution test indicate toxicity was reduced when
compared to the baseline LC50, thiosulfate has either
reduced or complexed the toxicant( If the highest
addition of thiosulfate increases the toxicity of the
sample, the thiosulfate itself may be at a toxic concen-
tration. However, if the LC50 for the next lower addition
of thiosulfate of the effluent dilutions reduced and/or
removed toxicity, then more tests for oxidants or metals
should be explored.
If oxidant toxicity is evident, a measurement of free
chlorine should be made and the concentration com-
pared to the chlorine toxicity value for the test species
used. For identification it may be necessary to measure
mono and dichloramine since they have different
toxicities than free chlorine (see Phase lii for confirm-
ing mixtures as toxicants). A comparison of the aera-
tion and C,e SPE test results to the oxidant reduction
test results may provide even more information on the
physical/chemical nature of the oxidants.
For those effluents where chlorine is measurable,
dechlorination may be achieved by the use of sulfur
dioxide (SO,) gas. This technique used was developed
by T. Wailer (personal communication, University of
North Texas, Denton, TX). (Note: Caution in handling
the SO, should be exercised because it is an extreme
irritant.) As with thiosuifate, SO. may also reduce com-
pounds other than chlorine, mis information can be
useful when one needs to know if substances other
than chlorine are causing the toxicity.
To dechlorinate using SO,, the following procedure
is used. Place 10 mL of high quality distilled water into
8-35
-------
Ftgurt 8-1S. Exampl* data »h»«t for th« oxldrnt reduction teat wh»n using a gradient of sodium thlosulfat*
concentrations.
Test Type: Oxidant Reduction
Test Initiation (Date & Time):
Investigator:
Sample Log No., Name:.
Date of Collection:
Species/Age:
No. Animals/No. Reps:.
Source of Animals:
Dilution Water/Control:.
Test Volume:
Other Info:
4x-LC50:
TRC:
Of 100%
ml Stock
Added
to %
Effluent
1.0
0.8
0.6
0.4
0.2
0.0
Oh
PH
Survival Readings:
2h
ApHDO
4h
ApHDO
8h
ApH DO
. *.
24 h
ApHDO
48 h
ApHDO
72 h
ApHDO
96 h
ApHDO
•
Note: See baseline data sheet for control data.
Stock Concentration - Q/L
Comments:
6-36
-------
Figure 8-16. Example data sheet tor the oxldant reduction teat whan affluent dllutlona ara imed.
Test Type: Oxidant Reduction
Test Initiation (Date & Time):
Investigator:
Sample Log No., Name:
Date of Collection:
Species/Age:
No. Animals/No. Reps:_
Source of Animals:
Dilution Water/Control:
Test Volume:
Other Info:
4x-LC50:
TRC:
or 100%
%
Effluent
4x-LC50
2x-LC50
Ix-LCSO
4x-LC50
2X-LC50
Ix-LCSO
4x-LC50
2X-LC50
lx-LC50
Cone.
Thiosulfate
O.Sx LC50
0.5* LC50
O.Sx LC50
0.25x LC50
0.25x LC50
0.25x LC50
0.125xLC50
0,125xLC50
0.125x LC50
mL
Stock
Used
Oh
pH
•
. Survival Readings:
24 h
A pH DO
h
48 h
A pH DO
72 h
A pH DO
96 h
A pH DO
V
Note: See baseline data sheet for control data. Baseline test also serves as toxicity blank for this
additive test.
Stock Concentration
Comments:
g/L
8-37
-------
a graduated cylinder. Bubble the SO, gas directly into
the water for about 5 min to prepare SO2-saturated
water. (Caution: This saturation procedure must be
done in a hood!) For a first attempt at the amount of
SO, to add without the TRC measured we use 2 p.L of
the SO,-saturated water per 100 ml of effluent with
TRC values of O-5 mg/L). This amount of SO, is not
acutely toxic and is effective at removing most com-
monly encountered TRC concentrations. For measured
chlorine concentrations, proportional amounts of SO,-
saturated water have been used as follows: for 0.02
mg/L TRC add 3.6 ^L SO,-saturated water to 1 L of
sample; for 0.16 mg/L TRC, add 12 \iLJL S0,-saturated
water; for 1.3 mg/L TRC add 39 jj.L/L SO, saturated
water: and for 2.1 mg/L TRC add 64 jiUL SO,-satu-
rated water (T. Waller, personal communication), An-
other technique to remove the chlorine is being ex-
plored at present. The use of sodium bisulfite solutions
added in the same way as sodium thiosulfate solutions
are added is being explored.
In cases where both the oxidant reduction test and
EDTA chelation test reduce the toxicity in the effluent
sample, there is a strong possibility that the tpxicant(s)
may be a cationic metal(s). For example, thiosutfate
and EDTA both reduce the toxicity of copper, cadmium,
and mercury, At this point, the Phase II methods for
identification for cationic metal(s) toxicants should be
investigated.
8.8 EDTA Chelation Test
Principles/General Discussion:
To determine the extent to which effluent toxicity is
caused by certain cationic metals, increasing amounts
of a chelating agent (EDTA; ethylenediaminetetraacetate
ligand) are added to aliquots of the effluent sample.
The form of the metal (e.g., the aquo ion, insoluble
complex) has a major effect on its toxicity to aquatic
organisms (Magnuson et al., 1979) and specific metal
forms are more important in aquatic toxicity than the
total quantity of the metal.
EDTA is a strong chelating agent, and its addition
to water solutions produces relatively non-toxic com-
plexes with many metals. The success of EDTA in
removing metal toxicity is a function of solution pH, the
type and speciation of the metal, other ligands in the
solution, and the binding affinity of EDTA for the metal
versus the affinity of the metal for the tissues of the
organism (Stumm and Morgan, 1981). Because of its
complexing strength, EDTA-metal complexes will often
displace other soluble forms such as chlorides and
oxides of many metals. Among the cations typically
chelated by EDTA are aluminum, barium, cadmium,
cobalt, copper, iron, lead, manganese (2*), nickel, stron-
tium, and zinc (Stumm and Morgan, 1981). EDTA will
not complex anionic forms of metals such as selenides,
chromates and hydrochromates, and forms relatively
weak chelates with arsenic and mercury. For those
metals with which it forms relatively strong complexes,
the toxicity of the metal to aquatic organisms is fre-
quently reduced: EDTA has been shown to chelatethe
toxicity to C. dubia due to copper, cadmium, zinc,
manganese, lead; and nickel (Hockett and Mount, In
Preparation) in both dilution water and effluents. How-
ever, it was also found that EDTA did not complex the
toxicity of silver, selenium (either as sodium selenite or
sodium selenate), aluminum (AI(OH)4"), chromium (ei-
ther as chromium chloride or potassium dichromate), or
arsenic (either sodium m-arsenite or sodium arsenate)
when tested using moderately hard water and C. dubia.
Since EDTA chelates calcium and magnesium (al-
beit weakly) the choice of the level of EDTA to add was
originally (EPA, 1988A) based on the premise that
calcium and magnesium had to be chelated before
toxic metals would be. However, recent work has shown
that the toxicity due to cationic metals was reduced
regardless of water hardness. Therefore the mass of
chelating agent required should be approximated be-
cause excess EDTA becomes toxic when present above
a certain concentration. The range of EDTA concentra-
tions that will adequately bind the metals but is not
toxic appears to be smaller than that for sodium thiosul-
fate and oxidants.
Table 8-5 contains LC50s of EDTA for Ceriodaphnia
and fathead minnows at various hardness and salinity
values. Note that the concentration of EDTA tolerated
by organisms increases directly with both water hard-
ness and salinity. By measuring the hardness and sa-
linity of the effluent, the range of EDTA concentrations
that should not be toxic in an effluent sample can be
estimated. "Salinity" not due strictly to NaCI will have
different effects on toxicity. This calculation, for predic-
tion of the EDTA concentration, is more involved than
is at first apparent. The data in Table 8-5 indicate that
over the physiological range of hardness and salinity,
hardness affects the toxicity of EDTA more than NaCI.
The usual methods for measurement of salinity (con-
ductivity meter, salinometer or refractometer) do not
specifically measure sodium chloride. The choice of
EDTA concentrations should always be based first on
hardness and secondly on salinity when the salinity is
known. The particular combination of hardness and
salinity present in an effluent sample may have to be
tested to get an accurate EDTA LC50. If the salinity is
composed of ions other than sodium and chloride, the
hardness of the dilution water should be made equal to
the effluent hardness and the additional "salinity" added
in the form of other major cations and anions such as
potassium, sulfate and carbonate.
An EDTA LC50 value derived in a standard dilution
test water (such as reconstituted water) is likely to be
much lower than the LC50 of EDTA added to an efflu-
ent. For example, the values contained in Table 8-5
represent worst case conditions presented by EDTA in
dilution water. Likewise, the toxic concentration of EDTA
in one effluent will probably not be the same as the
concentration causing toxicity in a different effluent or
even a different sample of the same effluent. For this
reason the concentrations of EDTA added to the efflu-
ent should bracket the expected LC50 based on clean
water with a similar hardness and salinity value as per
a-38
-------
Figure
-------
a graduated cylinder. Bubble the SO, gas directly into
the water for about 5 min to prepare SO,-saturated
water. (Caution: This saturation procedure must be
done in a hood!) For a first attempt at the amount of
SO, to add without the TRC measured we use 2 u.L of
the SO,-saturated water per 100 ml of effluent with
TRC values of O-5 mg/L). This amount of SO, is not
acutely toxic and is effective at removing most com-
monly encountered TRC concentrations. For measured
chlorine concentrations, proportional amounts of SO,-
saturated water have been used as follows: for 0.02
mg/L TRC add 3.6 ul SO,-saturated water to 1 L of
sample; for 0.16 mg/L TRC, add 12 nL/L S0,-saturated
water; for 1.3 mg/L TRC add 39 juL/L SO, saturated
water; and for 2.1 mg/L TRC add 64 jiUL SO,-satu-
rated water (T. Waller, personal communication). An-
other technique to remove the chlorine is being ex-
plored at present. The use of sodium bisulfite solutions
added in the same way as sodium thiosulfate solutions
are added is being explored.
In cases where both the oxidant reduction test and
EDTA chelation test reduce the toxicity in the effluent
sample, there is a strong possibility that the toxicant(s)
may be a cationic metal(s). For example, thiosulfate
and EDTA both reduce the toxicity of copper, cadmium,
and mercury. At this point, the Phase II methods for
identification for cationic metal(s) toxicants should be
investigated.
8.8 EDTA Chelation Test
Principles/General Discussion:
To determine the extent to which effluent toxicity is
caused by certain cationic metals, increasing amounts
of a chelating agent (EDTA; ethylenediaminetetraacetate
ligand) are added to aliquots of the effluent sample.
The form of the metal (e.g., the aquo ion, insoluble
complex) has a major effect on its toxicity to aquatic
organisms (Magnuson et al., 1979) and specific metal
forms are more important in aquatic toxicity than the
total quantity of the metal.
EDTA is a strong chelating agent, and its addition
to water solutions produces relatively non-toxic com-
plexes with many metals. The success of EDTA in
removing metal toxicity is a function of solution pH, the
type and speciation of the metal, other ligands in the
solution, and the binding affinity of EDTA for the metal
versus the affinity of the metal for the tissues of the
organism (Stumm and Morgan, 1981). Because of its
complexing strength, EDTA-metal complexes will often
displace other soluble forms such as chlorides and
oxides of many metals. Among the cations typically
chelated by EDTA are aluminum, barium, cadmium,
cobalt, copper, iron, lead, manganese (2*), nickel, stron-
tium, and zinc (Stumm and Morgan, 1981). EDTA will
not complex anionic forms of metals such as selenides,
chromates and hydrochromates, and forms relatively
weak chelates with arsenic and mercury. For those
metals with which it forms relatively strong complexes,
the toxicity of the metal to aquatic organisms is fre-
quently reduced, EDTA has been shown to chelate the
toxicity to C. dubia due to copper, cadmium, zinc,
manganese, lead; and nickel (Hockett and Mount, In
Preparation) in both dilution water and effluents. How-
ever, it was also found that EDTA did not complex the
toxicity of silver, selenium (either as sodium selenite or
sodium selenate), aluminum (AI(OH)4), chromium (ei-
ther as chromium chloride or potassium dichromate), or
arsenic (either sodium m-arsenite or sodium arsenate)
when tested using moderately hard water and C. dubia.
Since EDTA chelates calcium and magnesium (al-
beit weakly) the choice of the level of EDTA to add was
originally (EPA, 1988A) based on the premise that
calcium and magnesium had to be chelated before
toxic metals would be. However, recent work has shown
that the toxicity due to cationic metals was reduced
regardless of water hardness. Therefore the mass of
chelating agent required should be approximated be-
cause excess EDTA becomes toxic when present above
a certain concentration. The range of EDTA concentra-
tions that will adequately bind the metals but is not
toxic appears to be smaller than that for sodium thiosul-
fate and oxidants.
Table 8-5 contains LC50s of EDTA for Ceriodaphnia
and fathead minnows at various hardness and salinity
values. Note that the concentration of EDTA tolerated
by organisms increases directly with both water hard-
ness and salinity. By measuring the hardness and sa-
linity of the effluent, the range of EDTA concentrations
that should not be toxic in an effluent sample can be
estimated. "Salinity" not due strictly to NaCI will have
different effects on toxicity. This calculation, for predic-
tion of the EDTA concentration, is more involved than
is at first apparent. The data in Table 8-5 indicate that
over the physiological range of hardness and salinity,
hardness affects the toxicity of EDTA more than NaCI.
The usual methods for measurement of salinity (con-
ductivity meter, salinometer or refractometer) do not
specifically measure sodium chloride. The choice of
EDTA concentrations should always be based first on
hardness and secondly on salinity when the salinity is
known. The particular combination of hardness and
salinity present in an effluent sample may have to be
tested to get an accurate EDTA LC50. If the salinity is
composed of ions other than sodium and chloride, the
hardness of the dilution water should be made equal to
the effluent hardness and the additional "salinity" added
in the form of other major cations and anions such as
potassium, sulfate and carbonate.
An EDTA LC50 value derived in a standard dilution
test water (such as reconstituted water) is likely to be
much lower than the LC50 of EDTA added to an efflu-
ent. For example, the values contained in Table 8-5
represent worst case conditions presented by EDTA in
dilution water. Likewise, the toxic concentration of EDTA
in one effluent will probably not be the same as the
concentration causing toxicity in a different effluent or
even a different sample of the same effluent. For this
reason the concentrations of EDTA added to the efflu-
ent should bracket the expected LC50 based on clean
water with a similar hardness and salinity value as per
8-38
-------
Table 8-5. EDTA t cCeriodaphnia duW«minnow»;ad
hardnesM* a n sallnltle*
Species
Ceriodaphnia
dubia
promelas
Hardness
Water (mg/L a s Salinity'
Typ« CaCOj) (ppt)
VSRW 10-13 0
SRW 40-48 0
80-100 0
HRW 160-180 0
VHRW 0
0.5
I
2
40-48 3
0
0
0
0
280-320 0
0.5
40-48 1
2
3
24 h
0.04
(0.03-004)
0.12
(0.10-0.13)
0.23
(0.21-0.27)
0.50
(0.42-0.60)
0.71
(0.58-0.87)
0.05
0.12
(0.10-0.13)
0.33
(0.27-0.41 )
0.44
(-)
0.04
(0.03-0.04)
0.14
(0.12-0.18)
0.29
(0.23-0.35)
0.54
(0.43-0.66)
0.81
(0.68-0.97)
LC50{a/LU95%CI)
0.03
(— )
0.11
(-)
0.22
(— )
0.44
( — )
0.41
(0.36-0.47)
0.05
0.11
( — )
0.23
(0.21-0.27)
0.32
(0.23-0.45)
0.03 0.03
(0.03-0.04) (0.03-0.04)
0.14 0.11
(0.12-0.18)(0.08-0.14)
0.27 0.27
(0.22-0.33) (0.22-0.33)
0.50 0.47
(0.40-0.62) (0,36-0.60)
0.81 0.81
(0.68-0.97) (0.68-0.97)
0.03
(0.02-0.04)
0.08
(0.07-009)
0.25
(0.20-0.31)
0.44 ""
(0.34-056)
0.81
(0.52-0.83)
0.11
0.17
(0.13-0.21)
0.23
(0.17-0.32)
0.37
(028-0.48)
Brine from evaporated seawater used as source of salinity. All data generated at EPA ERL-Duluth. All C. dubia were 524 h old and the
fathead minnows were all <36 h old at test initiation. Ceriodaphnia were fed; see section on toxicity tests for details.
Note: (-) = Confidence interval cannot be calculated as no partial mortality occurred: NR = Not reported; VSRW = very soft reconstituted
water; SRW = soft reconstituted water; MHRW = moderately hard reconstituted water; HRW = hard reconstituted water; VHRW = very
hard reconstituted water.
the above discussion. The complexation of metals with
EDTA may not be immediate after the addition of EDTA.
Therefore, it is recommended that the EDTA test solu-
tions be set up first and these solutions allowed to sit
for the duration of pH adjustments and other manipula-
tions before the introduction of test organisms. This is
at least 2 h.
As with the oxidant reduction test, the EDTA can
be added in two ways; a gradient of EDTA can be
added to replicate of one effluent concentration or three
concentrations of EDTA can be added to three sets of
effluent dilutions. The effluent itself is used as a control
rather than a blank based on dilution water as in the pH
adjustment test. The gradient addition test is done by
adding increasing concentrations of EDTA to several
aliquots of the effluent (4x-LC50 or 100%). The goal of
this test is to add enough EDTA to reduce metal toxic-
ity. At some EDTA addition the metals will be chelated
and the EDTA will not be present at toxic concentra-
tions. At lower EDTA additions the metal toxicity is not
removed; in the midrange of the EDTA additions the
metals will be rendered non-toxic by the EDTA, and at
the high end of the range of EDTA additions, the
unreacted EDTA is itself toxic. By using an effluent
concentration of 4x-LC50 (or 100% if the LC50 is greater
than 25%) the potential for exceeding the binding capa-
bility of the added EDTA is lessened, especially for
very toxic effluents (LC50
To conduct the EDTA test using effluent dilutions,
three addition levels of EDTA (using one stock) are
selected (based on the LC50 of EDTA for the species
of choice). Each of these three EDTA levels are then
added to effluent dilution tests in a 3 x 3 matrix. The
EDTA is added to the 4x-, 2x-, and 1x-LC50 test cups
after the effluent solutions are prepared so that the
three EDTA concentrations are constant across each
a-39
-------
set of effluent dilutions. For example, 0.2 mL of an EDTA concentration (concentration of EDTA used in
EOTA solution is added to t tttratibn) cby Q.5/0.20DiO2.5x a Dlt)2$i igi EDTA stock.
4x-LC50,2x-LC50,1x-LC50of 1 00%, 50%, (Note: Molecular weight (MW) of NaaEDTA is 372.3 g.)
and 25%. To the next set of test cups of same effluent
dilution sequence, 0.05 mL of the EDTA stock solution
is added and likewise 0.0125 mL is added to the third
set of test cups.
To determine the amount of EDTA to add, one can
use the hardness titration, the measurement of calcium
and magnesium concentrations, or the concentration of
EDTA at the EDTA LC50 for the species of interest.
These are described in the Procedure below.
Volume Required:
A volume of 100 mL effluent usually is required for
the EDTA chelation test. The exact volume needed will
depend on the 24-h initial LC50 and the particular
option chosen to determine the EDTA addition.
Apparatus:
Glass stirring rods, glass volumetric for EDTA stock
solution, automatic pipette, disposable pipette tips, 10,
100, and 1000 \±L pipettes, eye dropper or wide bore
pipette, 30 mL beakers oMoz plastic cups, light box
and/or microscope (optional).
Reagents:
EDTA (disodium salt, Na, EDTA) stock solution
(see discussion under Procedure), reagents for deter-
mination of effluent hardness and salinity (APHA, 1980;
Methods 314 and 210).
Test Organisms:
Test organisms, 50 or more, of the same age and
species.
Procedure:
Day 2: There are three ways to determine the
concentration of EDTA stock to prepare.
The first and the most accurate approach (when it
can be used) is to measure the hardness of the 4x-24-h
LC50 effluent concentration (or 100% when the LC50 is
>25%) using the standard method for measuring hard-
ness (APHA, 1980). The concentration of EDTA that
produced the endpoint in the hardness titration of the
effluent sample is the concentration of EDTA needed at
the 0.2 mL addition for either the EDTA gradient test or
the effluent dilution test. An example illustrates this
calculation. In a 36% effluent sample (4x-LC50), 5 mL
of 0.01 M EDTA was needed to titrate the hardness
(100 mL sample size). For the gradient test, 7 EDTA
concentrations will be added to several test chambers,
all containing one concentration of effluent (4x-LC50).
The concentration of EDTA required for the hardness
titration is the highest additive concentration. For a 10
mL test volume of 36% effluent, when 0.5 mL of 0.01 M
EDTA stock was added the resultant EDTA concentra-
tion is that which is desired at the 0.2 mL addition. To
provide this EDTA concentration at the 0.2 mL addition
(minimizing the volume addition), increase the 0.01 M
The second approach is used when the hardness
measurement endpoint cannot be discerned because
of interferences. If the hardness cannot be titrated,
measure the calcium (Ca2*) and magnesium (Mg2*) of
the sample using atomic absorption procedures, and
calculate the amount of EDTA needed to chelate the
calcium and magnesium. EDTA binds with both CaJ*
and Mg2* on a 1:1 molar basis. The combined number
of moles of Ca2*(MW=40.1 g) and Mg2*(MW=24.3 g) in
10 mL of effluent at 4x-LC50 equals the number of
moles of EDTA needed for the 0.2 mL addition for a
10mL sample. This calculated concentration should be
added at the 0.2 mL addition for either the gradient or
dilution test. The calcium and magnesium should be
measured at 100% effluent if the LC50 is greater than
25%.
The third approach, and the one we use most
frequently, is to use the EDTA LC50 concentration to
select addition levels. This approach allows for either a
gradient of EDTA additions to 100% (4x-LC50) solu-
tions or EDTA additions to effluent dilution tests. Choice
of the EDTA LC50 must be based on effluent hardness
(and salinity). It may be necessary to determine the
EDTA LC50 for the particular combination of effluent
hardness and salinity and test organism used.
After the concentration for the stock solution of
EDTA has been determined, the EDTA gradient test
can be set up. The EDTA LC50 is generally set at the
0.2 mL addition. To perform the gradient EDTA addi-
tion test, 7 aliquots of the effluent are prepared at a
concentration equal to 4x-LC50, or 100% effluent where
the initial 24-h LC50 is greater than 25%. Next, 0.4 mL
of the appropriate EDTA stock is added to the first 10
mL aliquot of the effluent, 0.2 mL is added to the
second 10 mL sample of effluent, 0.1 ml to the third,
and so on until the sixth 10 mL effluent sample has
received 0.0125 mL The seventh is an effluent blank
used to compare treatment effects on time to mortality
(see Figure S-17). A microliter syringe will be needed
for the smaller additions. If the effluent has a low
toxicity (LC50 = 50-100%) a series of dilution blanks
may be necessary to check for the dilution effect of the
EDTA stock addition. No more than 10% dilution of the
effluent aliquots should be allowed unless a dilution
blank series is included.
To perform the EDTA additions using the effluent
dilution test, three sets of three effluent concentrations
are prepared (1 00%, 50%, 25%, or 4x-, 2x-, 1 x-LCSO),
while the baseline test serves as the toxicity blank.
After the concentration of stock solution of EDTA has
been established, add the EDTA using a 3 x 3 matrix.
This means that the 0.2 mL addition is added to the
100%, 50%, 25%, 0.05 mL is added to another set, and
0.0125 mL is added to the third effluent set (see Figure
8-18). The EDTA is added after all the dilutions are
prepared. To allow the EDTA time to complex the
8-40
-------
Figure 8-17. Example data sheet for EDTA chelation test when using t gradient of EOT A concentrations.
Test Type: EDTA Chelation
Test Initiation (Date & Time):
Investigator:
Sample Log No., Name:.
Date of Collection:
Species/Age:
No. Animals/No. Reps:.
Source of Animals:
Dilution Water/Control:_
Test Volume: ~
Other info:
4x-LC50:
or 100%
mL
Stock
Added
0.4
0.2
0.1
0.05
0.025
0.0125
0.0
Oh
PH
Survival Readings:
2h
ApH DO
4h
ApH DO
8 h
ApH DO
. \.
24 h
ApH DO
48 h
ApH DO
72 h
ApH DO
96 h
ApH DO
*
Note: See baseline data sheet for control data.
Stock Concentration « g/L EDTA
Comments:
8 41
-------
Figure 8-18. Example data sheet for the EDTA chalatlon test when effluent dilutions are used.
Test Type: EDTA Chelation
Test Initiation (Date & Time):
Investigator:
Sample Log No., Name:
Date of Collection:
Species/Age:
No. Animals/No. Reps:_
Source of Animals:
Dilution Water/Control:
Test Volume:
Other Info:
4X-LC50:
TRC:
or 100%
%
Effluent
4x-LC50
2x-LC50
Ix-LCSO
4x-LC50
2x-LC50
lx-LC50
4x-LC50
2x-LC50
Ix-LCSO
Cone.
EDTA
mL
Stock
Used
0.2
0.2
0.2
0.05
0.05
0.05
0.0125
0.0125
0.0125
Oh
pH
Survival Readings:
24 h
A pH DO
48 h
A pH DO
72 h
A pH DO
96 h
A pH DO
»•
Note: See baseline data sheet for control data. Baseline test also serves as toxicity blank for
this additive test.
Stock Concentration
Comments:
g/L EDTA
8-42
-------
metals, these samples should be prepared first. The
test solutions should not have test organisms added
until all other manipulations are performed.
The complexation of metals by EDTA proceeds at
a rate which may vary according to the sample matrix.
In studies of some aqueous samples containing metal
toxicity, better success in demonstrating chelation of
metals by EDTA may occur if samples spiked with
EDTA remain refrigerated overnight. The next day they
are warmed to the test temperature and the pH is
adjusted to pHi before placing test organisms in the
chambers. This allows time for the EDTA to chelate
any metals which may be in the sample. The solutions
should be mixed thoroughly after spiking with EDTA
and before adding the test organisms.
For both the gradient EDTA addition and EDTA
dilution tests, the pH of the effluent after addition of
EDTA should be checked. Since EDTA is an acid,
additions of this reagent will lower the pH of the efflu-
ent. The amount of change in solution pH will depend
upon the buffering capacity of the effluent and the
amount of reagent added. If the pH of the effluent has
changed, readjustment of the test solution pHtopHz
should be performed. When stable pHs are obtained
and all other manipulations have been completed, test
organisms are added to the test chambers.
Interferences/Controls and Blanks:
Controls prepared for the baseline toxicity test pro-
vide quality control for test organisms, dilution water
and test conditions. Either the zero ml EDTA addition
in the gradient test or the baseline effluent test serves
as a blank for use in determining the presence of EDTA
toxicity.
For the EDTA gradient test, time to mortality may
be recorded at each EDTA addition and then compared
to the untreated effluent. If time to mortality is shorter in
all treatments than in the untreated effluent, repeat the
test using lower EDTA concentrations. If the baseline
test has less toxicity than the EDTA additions in the
dilution test, then the EDTA may be causing toxicity. If
time to mortality or toxicity is not reduced in any treat-
ment, it may be wise to repeat the additions using a
higher range of EDTA concentrations. Erratic patterns
in mortality cannot be used, and when this occurs it
suggests that this test is not appropriate for the particu-
lar effluent being studied.
The addition of EDTA to the sample often lowers
the pH of the test solution to as low as pH 4.0. If the
presence of a pH-dependent toxicant such as ammonia
is suspected, then the results of this test must be
interpreted cautiously before attributing losses of toxic-
ity to chelation of metals. For instance, a sample which
contains ammonia toxicity and an undetermined amount
of metal toxicity (perhaps none) may show a loss of
toxicity at some EDTA concentrations. A closer look
may reveal that pH drifted in these samples to 7.5 or
lower, rendering ammonia non-toxic in the sample, and
that metal chelation may have had no role in reducing
toxicity. In the same way, the presence of compounds
whose toxicity is exacerbated at low pH (e.g., hydrogen
sulfide) may confound interpretations of this test. In
such a sample, metal toxicity may indeed be reduced
by chelation at some EDTA additions: however, the
increased toxicity of hydrogen sulfide at the lower pH
could mask the metal toxicity loss. In cases such as
these, we have returned pH to the initial value, and
successfully employed a simple method of pH control
(e.g., closing the test vessel) to avoid misleading pH-
dependent toxicant interferences. This strategy may
need to be attempted to ferret out interferences and
obtain useful information from the EDTA addition test.
In certain effluents, EDTA may reduce the toxicity
ofcationic surfactants. This reduction may appear as a
delay in time to mortality. If the EDTA test result is not
likely caused by cationic metals, other Phase I proce-
dures, such as sublation during the aeration step, may
also indicate surfactant toxicity (cf., Section 8.3 aera-
tion test).
Results/Subsequent Tests:
For the EDTA gradient test, if the appropriate EDTA
concentration range is utilized, the time to mortality will
not change from that seen in the exposure 4x-24-h
LC50 of unaltered effluent at low additions of EDTA. In
the 0.2 ml addition, toxicity should be reduced and at
higher additions of EDTA, toxicity will be as high or
higher than the whole effluent itself due to unbound
EDTA toxicity and effluent toxicants other than chelat-
able metals if present. Time to mortality must be used
to detect partial toxicity removal. Toxicity may be re-
moved at all exposures if the lowest addition of EDTA
removes metal toxicity and the highest addition does
not cause E.DTA toxicity. If toxicity is not reduced in any
treatment, either the effluent has no chelatable metal
toxicity or not enough EDTA was added. Increased
toxicity over the toxicity of untreated effluent suggests
EDTA toxicity and a lower EDTA range should be
tested.
For the EDTA dilution test, if the effluent is less
toxic (i.e., LC50 is greater than baseline LC50) in any
of the three EDTA addition dilution tests, then the
indication is that EDTA removed or reduced the toxicity
and therefore metal toxicity is present. If in all three
tests the effluent is more toxic (i.e., treatment LCSOs
are lower than baseline LC50), then the possibility
exists that EDTA itself is causing toxicity and the test
should be repeated using lower EDTA addition concen-
trations. If no LC50 of any of the three additions indi-
cates less toxicity than in the baseline test, the possibil-
ity of the presence of cationic metals causing toxicity in
the effluent is low, but additions of EDTA at higher
levels may need to be explored.
If toxicity is reduced in a systematic manner, pro-
ceed to Phase II methods for specific identification of
the metal(s).
8-43
-------
8.9 Graduated pH Test
Principles/General Discussion:
This test is designed to determine whether effluent
toxicity can be attributed to compounds whose toxicity
is pH dependent. The pH dependent compounds of
concern are those with a pKt that allow sufficient differ-
ences in dissociation to occur in a physiologically toler-
able pH range (pH 6-9). Also, the two forms of the
compound (ionized versus un-ionized) must have de-
tectable toxicity differences to the TIE organism. The
ionizable compounds commonly found in municipal and
industrial discharges include ammonia, hydrogen sul-
fide, cyanide, and some organic compounds (e.g., pen-
tachlorophenol). In addition, pH differences can affect
metal toxicity through changes in solubility and specia-
tion. The effect of pH on ammonia toxicity might be
more readily observed than the effect of pH on the
levels of toxicity of metals, hydrogen sulfide, cyanide,
and ionizable ofganics.
Ammonia is frequently present in effluents at con-
centrations of 5 mg/L to 40 mg/L (and higher). (The
ammonia is measured upon arrival of the sample (Sec-
tion 6) and this information will be helpful for the gradu-
ated pH test.) Levels of 5 mg/L to 40 mg/L are likely to
cause toxicity when several other effluent conditions
occur. Effluent parameters to consider are pH, tem-
perature, DO, CO, content, and TDS. Of these param-
eters, pH has the largest effect on ammonia toxicity,
and for many effluents (especially with POTW efflu-
ents) the pH of a sample rises upon contact with air.
Typically, the pH at air equilibrium ranges from 8.0 to
8.5. Literature data on ammonia toxicity (EPA, 19858)
can be used only as a general guide because of the
large effect of very slight pH changes. The pH values
for most ammonia toxicity tests are usually not mea-
sured or reported fully enough to be useful.
One might expect ammonia to be removed during
the pH 11 adjustment/aeration test. Based on our expe-
rience, however, ammonia is not substantially removed
by the method described in Section 8.5. Other tech-
niques which can be used to remove ammonia related
toxicity may also displace metals or other toxicants with
completely different physical and chemical characteris-
tics. For example, ion exchange resins (e.g., zeolite)
removes ammonia, cationic metals, and possibly or-
ganic compounds through adsorption. For these rea-
sons, the graduated pH test is most effective in differ-
entiating toxicity related to ammonia from other causes
of toxicity, if it is the dominant toxicant.
Ammonia acts as a basic compound in water. The
un-ionized, more toxic form (NH,) predominates above
pH 9.3 and the ionized, essentially non-toxic form (NH,+)
is most abundant below pH 9.3 at 25°C. Through the
pH range of 6-8.5, the percent of ammonia in the toxic
form increases 250x over this range. Importantly, as pH
increases, the percentage of the toxic form becomes
greater but the toxicity of the toxic form is less, and
conversely, as pH decreases, the percentage of ammo-
nia (NH,) decreases, but the toxicity of the NH, in-
creases (EPA, 19858). However, the increase in the
concentration of ammonia occurring in the toxic form
with increasing pH is greater than the decrease in its
toxicity. The net result is an increased toxicity of a
given total ammonia concentration with increased pH.
Temperature also affects the dissociation of ammonia,
but since the temperature is held constant in these
toxicity tests for Phase I, it can be ignored.
Effluent toxicity related to metals may also be de-
tected by the graduated pH test, although these effects
are less well documented in effluents than those asso-
ciated with ammonia toxicity. Acidification of a sample
may increase the bioavailable portion of a metal, and in
some cases (i.e., cadmium, copper, and zinc) this is
countered by a decreasing toxicity of the metal as the
test pH decreases. It. is known, however, that aluminum
toxicity increases as pH diverges from neutral. In ex-
periments in the pH range of 5 to 7 (Campbell and
Stokes, 1985), the toxicities of cadmium, copper and
zinc were shown to increase with increasing pH while
the toxicity of lead decreased with increasing pH. We
have found lead and copper to be more toxic to C.
dubia at pH 6.5 than at pH 8.0 or 8.5, (in very hard
reconstituted water) and nickel, zinc, and cadmium
were more toxic at pH 8.5 than at 6.5. Since these
compounds are also chelatable by EDTA, the results of
both tests (the graduated pH test and the EDTA addi-
tion test) can give information about whether it is an
ionizable compound or a pH sensitive cationic metal.
Other metals have exhibited some degree of pH de-
pendence, but these are not as well defined. Results of
the graduated pH test should be considered in conjunc-
tion with the EDTA addition test (Section 8.8). Whether
the metal toxicity can be discerned will depend in large
part on the concentration of other pH dependent toxi-
cants in the sample. In order to detect metal toxicity,
one must be cautious when selecting a dilution water
when the test solutions are at low effluent dilutions
because artifactually enhanced toxicity due to metals
may be created if the hardness of the dilution water is
much different than that of the effluent. This effect may
be magnified for metals when coupled with the pH
change. A dilution water similar in hardness to the
effluent must be used for this test to reveal metal-
caused toxicity. If more than one pH dependent toxi-
cant is present, the pH effects may either cancel or
enhance one another.
Hydrogen sulfide (H2S) occurs in wastewaters, and
its toxicity can be detected by the graduated pH test.
Dissolved sulfide exists in two forms, H2S and HS'. The
predominant form depends on both pH and tempera-
ture, but since temperature is held constant in these
Phase I toxicity tests it essentially can be ignored. The
un-ionized form (H2S) is more toxic to aquatic organ-
isms, and at pH 6 it comprises over 90% of the dis-
solved sulfide, while at pH 7, 50% is unionized. At a
pH of 8.5, less than 5% of the dissolved sulfide is
present in the unionized form. Since H2S is the more
toxic form, one would expect to observe an increase in
toxicity relative to a decrease in solution pH. When
considering results of this type it is wise to check
toxicity alteration by the pH adjustment/aeration tests
8-44
-------
(Section 8.5). H2S is readily oxidized and also removed
through volatilization; therefore if H2S is the predomi-
nating toxicant, a significant reduction of toxicity should
be observed in the pH adjustment/aeration tests.
The effects of pH on toxicity can be used to detect
the presence of these pH dependent toxicants. By
conducting three effluent tests, each at a different pH,
the effluent toxicity can be enhanced, reduced or elimi-
nated. For a typical example (at 25°C) where ammonia
is the primary toxicant, when the pH is 6.5, 0.180% of
the total ammonia in solution is present in the toxic
form (NH,). At pH 7.5, 1.77% of the total ammonia is
present as NH, and at pH 8.5, 15.2% is present as
NH,. Similar changes in the percent ammonia as NH
forpH's 6.5, 7.5 and 8.5 occur at other temperatures:
for example, the percentages of unionized ammonia at
20°CforpH's 6.5, 7.5, and 8.5 are 0.130%, 1.24% and
11.2%, respectively (EPA, 1979). This difference in the
percentages of unionized ammonia is enough to make
the same amount of total ammonia about three times
more toxic at pH 8.5 as at pH 6.5. Whether or not
toxicity will be eliminated at pH 6.5 and the extent to
which toxicity will increase at pH 8.5 will depend on the
total ammonia concentration. If the graduated pH tests
are done at dilutions symmetrical about the LC50, one
should see toxicity differences between pH 6.5 and 8.5
(cf., Phase II discussion on equitoxic test). The effluent
LC50 (expressed as percent effluent) should be lower
at pH 8.5 than pH 6.5 if ammonia is the dominant
toxicant.
The most desirable pH values to choose will de-
pend upon the characteristics of the particular effluent
being tested. For example, if the air equilibrium pH of
the effluent at 4x the 24-h LC50 is 8.0 it may be more
appropriate to use pH's 6.0, 7.0, and 8.0. The gradua-
tion scheme that includes the air equilibrium (the pH
the effluent naturally drifts to) will allow a comparison of
treatments to unaltered effluent (i.e., baseline test). The
pH's of many POTW effluents rise to 8.5 or higher, so a
gradient of pH's such as 6.5, 7.5 and 8.5 is more
appropriate. In any case, it will be necessary to conduct
the test at more than one effluent concentration (4x-,
2x-, Ix-, 24-h LC50) or with a different graduated pH
scheme to determine what role, if any, the pH depen-
dent compounds play in toxicity.
Perhaps the greatest challenge faced in this gradu-
ated pH test is that of maintaining a constant pH in the
test solution. This is a necessity if the ratio of ionized to
the un-ionized form is to remain constant and the test
results are to be valid. In conducting toxicity tests on
effluents, it is not unusual to see the pH of the test
solutions with effluent concentrations of > 12% drift 1 to
2 units over a 48 to 96-h period (see Procedure for
suggestions on pH control).
Volume Required:
The volume needed is dependent on the test de-
sign chosen to conduct this test. The test chamber
size, number of dilutions, and the toxicity of the effluent
will dictate this; however, 200 mL of test volume'should
suffice for all three pH's.
Apparatus:
Test chambers such as 78 mm L x 50 mm W X50
mm H or 1 oz plastic cups; Hamilton 1 L gas syringe
(Model S-1000, Reno, NV); 35 mm x 14 mm H Corning
plastic petri dish bottoms, rubber stoppers, eye drop-
pers or wide bore pipette, 30 ml beakers or 1 oz
plastic cups, light box, and/or microscope (optional).
Reagents:
Cylinder tank of.CO , 1 .0, 0.1, and 0.01 N HCI
1 .0, 0.1, and 0.01 N NaoH (ACS grade in high 'purity
water), buffers for pH meter calibration.
Test Organisms:
Use 5 for each of three dilutions of the whole
effluent (4x-, 2x-, Ix-, 24-h LC50 or 1 00%, 50%, and
25%) and for each test pH (e.g., pH 6, 7, 8) (Figure 8-
19), as well as a control.
Procedure:
Day 2: Either CO.nt.HCI (or the combination of
both) can be used to lower the pH of the sample. The
pH of most natural waters and some effluents is con-
trolled by the bicarbonate buffering system. Surface
waters normally contain <10mg/L of free CO,.
»*
For the CO, pH controlled tests, the pH is adjusted
with CO, by varying CO, content of the gas phase over
the water or effluent sample. It is necessary to maintain
constant pH's in the static acute test throughout the 48
or 96-h tests. By using closed headspace test cham-
bers, the CO, content of the gas phase can be con-
trolled. The amount of CO, needed to adjust the pH of
the solution is dependent upon sample volume, the test
container volume, the desired pH, the temperature, and
the effluent constituents (e.g., dissolved solids). When
dilutions of an effluent have the same hardness and
initial pH as the effluent, the same amount of C0?wj||.
usually be needed for each dilution, but sometimes
more is needed in the higher effluent concentrations.
Use of a dilution water of similar hardness as the
effluent makes the CO, volume adjustments easier.
In our laboratory, a rectangular chamber (measur-
ing 78 mm L x 50 mm W x 50 mm H) with a small
diameter hole (approximately 20 mm) on one end has
worked well for the CO, graduated pH test. The test
solution volume should be about 10% of the headspace
volume to maintain a large surface to volume ratio
should be maintained. For a 20 ml test volume, with
the CO, gas flushed into air space of the test chamber,
pH's have reached equilibrium in about 1 h. In most
instances, the amount of CO, produced by the inverte-
brates has not caused further pH shifts, but with larval
fathead minnows, the pH can drop from the amount of
CO, they respire as well as decomposition of food.
Therefore, in fish tests, the headspace must be reflushed
daily.
The exact amount of CO to inject for pH's 6.0, 7.0
and 8.0 must be determined through experimentation
with each effluent before the graduated pH test begins.
8-45
-------
Figure 8-19. Example of data sheet for the graduated pH test when effluent dilutions are used.
Test Type: Graduated pH
Test Initiation (Date & Time):
Investiaator:
Sample Log No., Name:
Date of Collection:
Soecies/Aqe:
No. Animals/No. Reps:
Source of Animals:
Dilution Water/Control:
Test Volume:
Other Info:
4x-LC50: or 100%
TRC:
%
Effluent
4x-LC50
2x-LC50
Ix-LCSO
4x-LC50
2x-LC50
Ix-LCSO
4X-LC50
2x-LC50
Ix-LCSO
PH
6,0
6.0
6.0
7.0
7.0
7.0
8.0
8.0
8.0
Oh
PH
Survival Readings:
24 h
A pH DO
48 h
A pH DO
72 h
A pH DO
96 h
A pH DO
*
Note: See baseline data sheet for control data.
Comments:
8-46
-------
The amount of CO, added to the chamber assumes
that the liquid volume to gas volume ratio remains the
same. Generally, as the alkalinity increases, the con-
centration of CO, that is needed to maintain the pH
also increases, inject the CO using a gas tight syringe
and quickly close the test chamber tightly. Place the
test chamber in a position that maximizes the surface
to volume ratio. To prepare the test solutions, use a
dilution water of a similar hardness to the effluent and
transfer the effluent solutions to the test container and
randomly add the test organisms. Then add the pre-
determined amount of CO, to obtain the desired pH's
and close the container. For pH values from pH 8.5 to
6, 0-10% CO, has been needed. If more than 10% CO,
is needed, adjust the solutions with acids and bases
(described below) and flush the headspace with CO,.
Again, the necessary concentration of CO, to use must
first be determined experimentally with effluent test
solutions already adjusted to the appropriate pH. This
may require the test to be set up one day later than the
other Phase I tests.
For some effluents adequate pH control can be
obtained by adjusting the pH with acid or base and
tightly covering the test container (no headspace pH
test). A technique that we use has the 1 oz plastic cups
covered with plastic tissue culture dishes (see Appara-
tus for details). This technique works well with effluents
that have adequate DO content, and'where the BOD is
not high. The procedure for using plastic cups with
tissue culture dish covers is as follows. Adjust three
aliquots of the effluent and the dilution water to the
appropriate pHs. Next, prepare the appropriate dilu-
tions for testing (i.e., 4x-, 2x-, 1x-24-h LC50, or 1 0%,
50%, 25%) and check the pH in one-half hour. If the
pH's have drifted, readjust them with the appropriate
acid or base. Transfer about 35 ml_ of each into the 1
oz plastic cups, and randomly add the test organisms.
Carefully place the cover onto the cup; care must be
exercised because some test water will be displaced by
the lid, and organisms can be lost. Ensure that no air is
trapped under the lid during the sealing process. If air
is trapped, remove the cover, count the number of
organisms, and add an additional small amount of the
appropriate pH adjusted test solution. The test organ-
isms can readily be observed through the clear cover
or the sides of the plastic cup. The cover should be
removed only when all the animals have died as the
tight seal cannot be obtained after initially setting up
the test without adding more test water. Once animals
have died or the test is over, remove the cover and
measure the pH and DO. It is important to measure the
DO because toxicants such as ammonia have different
toxicities when DO is low (EPA, 1985B). Keep in mind
that if all of the test animals have been dead for a
while, the pH and/or DO of the test water could have
changed.
Methods that use continuous flow of a COj/air
mixture, such as tissue cell incubators, may be prefer-
able and give better pH control. At this time we have
not attempted to use a continuous flow of CO, and
cannot recommend a system to use.
8
Maintaining pH above the air equilibrium pH (gen-
erally above 8.3) is difficult to achieve. The pH control
in this high range is much more difficult because the
concentration of CO. must be very low and the micro-
bial respiration can increase the CO, levels in the test
chamber. Use of CO,-free air in the headspace may
work or bubbling a mix of CO,-free air and normal air
through the headspace or test solution may be needed.
Because such small C0a concentrations are needed
and because CO, evolution by microorganisms or test
organisms can significantly alter the CO, concentration,
more frequent flushing of the headspace in static tests
will be needed.
Since many plastics are permeable to CO,, glass
containers may need to be used. Measurements of pH
must be made rapidly to minimize the CO exchange
between the sample and the atmosphere, dvoid vigor-
ous stirring of unsealed samples because at lower pH
values, the CO. loss during the measurement can cause
a substantial pH rise.
For the CO, pH controlled tests, the pH should be
measured at 24, 48, 72, and 96 h and at each reading,
one may need to re-flush the headspace with CO,. A
small amount of experimentation will determiQe the
amount of CO, needed for this step. For trTe no
headspace pH tests conducted in cups with covers, air
bubbles may start to appear after 12 h, and this can
cause the pH to change. An excess of each test solu-
tion may need to be prepared to be added to the test
cup at each 24 h interval to prevent the formation of air
pockets which contribute to pH drift.
We also have been exploring the use of hydrogen
ion buffers to maintain the pH of effluent test solutions.
Efforts to use phosphate buffers were unsuccessful
due to the toxicity of the phosphates themselves. Three
hydrogen ion buffers were used by Neilson et al. (1990)
to control pH in toxicity tests in concentrations ranging
from 2.5 to 4.0 mM. These buffers were chosen based
on the work done by Ferguson et al. (1980). The
buffers are: 2-(N-morpholino) ethane-sulfonic acid (Mes)
(pK, = 6.15), 3-( N-morpholino) propane-sulfonic acid
(Mops) (pKt = 7.15), and piperazine-N,N'-bis (2-
hydroxypropane) sulfonic acid (Popso)(pK, = 7.8).
The acute toxicity of these buffers is low to both
C. dubia and fathead minnows (Table 8-6) and suble-
thal levels can be added to hold the pH of test solu-
tions. For example, 6.25 mM (1.2 g/L) of the Mes buffer
has been adequate to maintain the pH of one effluent
to within ±0.1 pH units. However when used in a sedi-
ment pore water, more buffer was needed (i.e., 25 mM
or 4.9 g/L) but these levels are still below the acute
toxicity for the buffer. Likewise for the Mops buffer, 6.25
mM (1.3 g/L) held the pH of the effluent at ±0.1 pH
units, but 50 mM (10.5 g/L) was needed for the pore
water. The Popso buffer held the pH at 8.2 or 8.5 using
6.25 or 12.5 mM (2.3 or 4.5 g/L, respectively) of buffer
for both the effluent or pore water. The addition of
these buffers did not change the toxicity of a non-toxic
effluent or change the toxicity of a toxic effluent and
sediment pore water.
47
-------
Table 8-6. The toxicity of the Mea, Mops, and Popao buffers to Ctriodaphnla dubia and fathead minnow*
LCSOJo/L)
Buffer
Speclei
Water
Type
24 h
48 h
72 h
96 h
Mes
Mes
C. dubia
C. dubia
LSW
VHRW
15,0
17.4
7.4
12.1
Mops
Mops
C dubia
C. dubia
LSW
VHAW
16.1
>20.9
13.0
11.9
PopsO
PopsO
C. dubia
C. dubia
LSW
VHRW
>2.3
12.7
>2.3
8.3
Mes
Mes
P. promelas
P. promelas
LSW
VHRW
13.9
>19.5
13.9
>19.5
13.9
>19.5
13.9
>19.5
Mops
Mops
P. promelas
P. promelas
LSW
VHRW
>20.9
>209
17.2
>20.9
16.1
>20.9
16.1
>209
PopsO
PopsO
P. promelas
P. promelas
LSW
VHRW
32.3
>36.2
32.3
>36.2
27.9
>36.2
27.9
>36.2
Note: The pH was held to at least ± 0.1 pH unit of desired pH for all tests. Mes buffer tests were at pH 6.2, Mops buffer tests were at pH
7.2, and Ropso buffer tests were at pH 8.2. LSW = Lake Superior water; VHRW = very hard reconstituted water.
While these buffers serve to prevent the pH from
drifting, their addition alone does not actually adjust the
pH value to the desired pH. The buffers are weighed
out and added to the aliquots of whole effluent and
dilution water and both are then pH adjusted with base
to the appropriate values. Serial dilutions are made,
and test organisms are added. While our experience
with the buffers is limited, we have found the amount of
any buffer needed to hold any pH is effluent specific.
Experiments will need to be done to determine the
lowest concentration of buffer needed to maintain the
desired pH. The test solutions need not be covered
tightly to maintain pH; however, pH should be mea-
sured at each survival reading at all dilutions.
In all graduated pH tests, the pH should be mea-
sured at least in the chambers that bracket the LC50
concentration as soon all the animals die. If the pH
drifts more than 0.2 pH units, the results may not be
usable and better pH control must be achieved.
Interferences/Controls and Blanks:
The controls in the CO, chamber or closed cup,
and the baseline test act as checks on the general
health of the test organisms, the dilution water and
most test conditions. If the effluent pH in the baseline
test (at the LC5Q) is close to the pH of the pH adjusted
test solutions (at their respective LCSO's), the toxicity
expressed in the two tests should be similar. Signifi-
cantly greater toxicity may suggest interference from
other factors such as the ionic strength related toxicity
if the pH was adjusted with either HCL or NaOH (cf.,
Section 8.3), or CO, toxicity. Dilution water blanks at
the various pH's are not used because such blanks are
not appropriate since the effluent matrix may differ from
that of the dilution water. The cleanliness of the acids
and bases is checked in the blanks of the pH adjust-
ment test. Other compounds with toxicities that in-
crease directly with pH may lead to confounding results
or may give results similar to ammonia. Phase II con-
tains a suggested test (called the equitoxic test) to
identify ammonia as the cause of toxicity. Monitoring
the acid and base additions may be useful to determine
if artifactual toxicity resulted from the addition of the
salts. Monitoring conductivity of the effluent solutions
after the addition of the acids and bases may also be
helpful in determining artifactual toxicity.
Results/Subsequent Tests:
For the graduated pH test, the pHs selected must
be within the physiological tolerance range for the test
species used (which generally is a pH range of 6 to 9).
In this pH range, the amount of acid or base added is
negligible, and therefore the likelihood of toxicity due to
increased salinity levels is low.
8-48
-------
When ammonia is the dominant toxicant, the efflu-
ent LC50 of the pH 6.5 test solution should be higher
than in the pH 7.5 test, which in turn, should be higher
than the pH 8.5 test. However, ammonia is not the only
possible cause of toxicity. Using the pH at the baseline
effluent LC50, the relative toxicity of each pH adjusted
solution can be predicted if ammonia is the sole cause
of toxicity. For example, if in the baseline effluent toxic-
ity test, the average pH was 8.0 in the 100% concentra-
tion in which no organisms survived and the average
pH for the 50% concentration was 7.5 and all organ-
isms survived, the estimated pH at the LC50 (71%)
could be approximated at 7.7. One would expect greater
than 50% mortality in the pH 8 test solution and signifi-
cantly less in the pH 7 solution. Therefore, if this occurs
one should proceed to Phase II to identify the pH
sensitive toxicant.
If ammonia is one of several toxicants in an efflu-
ent, this procedure may pose problems. For this rea-
son, if effluent total ammonia levels are greater than 20
mg/L, it may be appropriate to include a pH 6 effluent
treatment interfaced with other Phase I tests (cf., Sec-
tion 9). Methods for further identifying and confirming
ammonia as the toxicant can be found in Phases II and
8-49
-------
Section 9
Time Frame and Additional Tests
9.1 Time Frame fur Phase I Studies
The amount of time necessary to adequately char-
acterize the physical/chemical nature of, and variability
in, an effluent's toxicant(s) will be discharge specific.
Among the factors affecting the length of Phase I stud-
ies for a given discharge is the appropriateness of
Phase I tests to the toxicants, the existence of long- or
short-term periodicity in individual toxicants and to a
lesser extent, the variability in the magnitude of toxicity.
An effluent which consistently contains toxic levels of a
single compound that can be neutralized by more than
one characterization test, should be moved into Phase
II more quickly than an ephemerally toxic effluent with
highly variable constituents, none of which are im-
pacted by any of the Phase I tests. The decision as to
when to go beyond Phase I should be based in part on
the regulatory implications and resources involved in
subsequent actions. Where a great amount of resources
is involved, it is crucial that Phase I results be ad-
equate.
There are no clearly defined boundaries between
Phase I and Phase II. The section Results/Subsequent
Tests of the characterization tests in Section 8 provide
further tests to conduct and may be thought of as
intermediate studies between Phases I and II. In terms
of guidance for the time frame of the TIE, several
samples should be subjected to the Phase I character-
ization test battery but not all manipulations have to be
done on all subsequent samples. The decision to do
subsequent tests on these samples to confirm or fur-
ther delineate initial results is a judgement call and will
depend on whether or not the results of Phase I are
clear-cut. The time required to perform a complete
Phase I battery on a sample will depend on many
circumstances, not the least of which is how well orga-
nized and experienced the performing lab is at doing
TIEs.
If the Phase I characterization tests needed to
remove or neutralize effluent toxicity vary by the sample,
the number of tested samples must be increased and
the frequency of testing should be sufficient to include
all major variability. We cannot provide a time frame or
the number of samples to evaluate. Again, judgement
will have to be used but the differences seen among
samples can be used to decide when further differ-
ences are not being found. Phase I characterization
testing should continue until there is reasonable cer- -
tainty that new types of toxicants are not appearing. No
guidance can be given as to how many weeks or
months this may take-each problem for every dis-
charger is unique. The LC50 of samples can be very
different but the same screening tests must be suc-
cessful in removing and/or neutralizing effluent toxicity.
The individual Phase I tests which were previously
successful in changing toxicity should be used as a
starting point for Phase II identification. The first step in
Phase II will often be to reduce the number of constitu-
ents accompanying the toxicants. These efforts may
reveal more toxicants than suggested by Phase I test-
ing. In Phase II one may discover that toxicants of quite
a different nature are also present but were not in
evidence in Phase I. More Phase I characterization
may then be needed. *
Phase I results will not usually provide information
on the specific toxicants. Therefore, if effluent toxicity Is
consistently reduced, for example through the use of
C,, SPE, this does not prove the existence of a single
toxicant. In fact, several non-polar organic compounds
may be causing the toxicity in the effluent over time,
but use of the C,, SPE technique in Phase I detects the
presence of these compounds as a group. Recognizing
this lack of specificity is very important for subsequent
Phase II toxicant identification.
9.2 When Phase I Tests are Inadequate
For some effluents, the Phase I tests described
above will provide few or no clues as to the characteris-
tics of the toxicants. For such effluents, other approaches
must be tried. Some additional approaches are given
below in much less detail than tests in Section 8 be-
cause our experience with them is limited. In addition to
these, one should not hesitate to use originality and
innovation to develop other approaches. As long as
toxicity is used to track the changes, any approach may
be helpful.
Use of Multiple Phase I Manipulations
Our experience suggests that independent action
and less than additive action are much more common
than we realize, at least in effluents. When these inter-
actions occur, interpreting Phase I data may be difficult
and in some instances (especially with independent
action) no apparent effect on toxicity will be seen un-
less Phase I tests are clustered or used in a series.
These steps do not begin until all Phase I manipula-
tions have been completed and the results evaluated.
Tests are continued to further separate and concen-
trate the toxicant(s).
9-1
-------
The pH of effluents plays an amazingly powerful
role in how it affects both the form of toxicants and their
toxicity. Including pH adjustments to different values
than is suggested in Phase I may be helpful. For
example, if the C^-
-------
Phase I Manipulations, regarding complications of de-
termining toxicant interactions in effluents.
After Phase I is completed on a sample, the investi-
gator must carefully evaluate the data, draw conclu-
sions, and make decisions about the next steps that
are needed. Sometimes the next step is obvious, at
other times the outcome will be confusing and the next
step will not be obvious. Several general suggestions,
based on our experience to date, may provide some
help.
As a matter of principle, where multiple toxicants
are involved, experience shows that once one toxicant
is identified, identification of subsequent toxicants be-
comes easier because:
1. The toxicity contribution of the identified
toxicant can be established for each sample.
2. The number of Phase I manipulations that
will affect the toxicity of the known toxicant
can be determined.
3. One can determine whether the identified
and the unidentified toxicant(s) are additive.
4. If some manipulations affect the toxicity
due only to the unidentified toxicants, some
of their characteristics can be inferred.
5. One can determine if the relative toxicity
contributions of identified and unidentified
toxicants varies by sample. Such information
can be used to design tests to elucidate
additional physical/chemical characteristics.
Another suggestion, is that when some Phase I
outcomes are understandable and others are not, con-
centrate on the one or the few that seem to be the most
clear-cut and which have a major effect on toxicity. For
example, if an effluent has 10 TU and 2 TU are re-
moved by the addition of EDTA, 1 TU is removed by
the C1t column and 5 TU are removed by the aeration
manipulation, begin identification on the toxicity removed
by aeration. In another example, suppose the filtration
manipulation reduced the toxicity by 1 TU, both pH 3
and pH 11 adjustment tests showed that the toxicity
increased by 2 TU, the graduated pH test at pH 7
decreased toxicity by 2 TU and the post-c,, SPE col-
umn effluent (at pH i) had 2.5 TU less toxicity than the
whole effluent. Of the 2.5 TU removed by the column,
1.7 TU could be eluted with the 100% methanol. The
next step then is to begin the Phase II identification on
the SPE extractable toxicity because:
1. Widely accepted methods are available for
analyses of many non-polar organic
compounds.
2. The method exists for both separating and
concentrating such toxicants (cf., Phase II).
3. This C, extractable toxicity manipulation
behaved as expected.
4. Many effluents have non-polar toxicity, and
based on those probabilities, that non-polar
toxicity is likely to be real.
In the latter example, the unexplainable pH and
filtration effects might be a result of the behavior of the
non-polar toxicant(s) or could be caused by some as-
sociated artifact. If the non-polar toxicity is identified,
then the results of the pH adjustment and filtration
steps may be explainable.
The third suggestion is to concentrate on those
manipulations affecting toxicity in which the toxicant is
removed from other effluent constituents. In the above
example, the SPE column separated the toxicant(s)
from other non-sorbable constituents. Other examples
of where the toxicant is removed from the other con-
stituents are the filtration and the aeration manipula-
tions.
Separating the toxicant(s) from non-toxicant(s), and
concentrating the toxicant are usually the most produc-
tive efforts to pursue before identification (analyses)
begins. Attempts to begin analysis for suspect toxicant(s)
without this-step is frequently a mistake, and can be
costly.
9.4 Interpretation Examples
In this section, various examples of Phase I results
are given with interpretation suggestions. These'Shpuld
be used only as guides to thinking and not as definitive
diagnostic characteristics. Since almost any toxicant
can be present in effluents, clear-cut logic is not totally
dependable in interpreting results. Rather, one must
use the weight of evidence to proceed, and be aware
that artifacts cannot at this point always be identified.
One should avoid making categorical assumptions
to every extent possible. For example, to assume that
the toxicity is due to a non-polar toxicant(s) because
the toxicity in the post-c,, SPE column effluent was
removed often is an error. Metals may also be the
toxicant adsorbed by the SPE column. However, as in
the example in Section 9.3, if the toxicity can be recov-
ered in the methanol fraction (see Section 8.6, Results/
Subsequent Tests forelution and Phase II for more
details), then the theory that a non-polar toxicant(s) is
causing the toxicity is better substantiated. Metals do
not elute with methanol and therefore do not produce
toxicity in the methanol fraction toxicity test (cf.,
Phase II).
Example /. Nun-polar toxicant(s). The Phase I re-
sults implicating non-polar toxicants are:
1. All toxicity in the post-c,, SPE column
effluent was removed.
2. The toxicity removed was recovered in the
methanol elution of the SPE column.
The above discussion (cf., Section 9.3) has pro-
vided most of the interpretative rationale for these
Phase I results which are typical of non-polar organics.
As stated above, toxicants other than non-polar com-
pounds may be retained by the SPE column but they
are less likely to be eluted sharply. Also, as discussed
9-3
-------
Example V. Ammonia. Ammonia concentrations can
be measured easily, and because it is such a common
effluent constituent, determining the total ammonia con-
centration in the whole effluent is a good first step (see
Section 6). If more than 5 mg/L of total ammonia is
present, additional evaluations should be done. Sole
dependence on analyses is not advisable because there
is little or no additivity between ammonia and some
other toxicants (e.g., such as surfactants). Even though
the ammonia concentration is sufficient to cause toxic-
ity, other chemicals may be present to cause toxicity if
the ammonia is removed.
Three indicators of ammonia toxicity are:
1. The concentration of total ammonia is
5 mg/L or greater.
2. Toxicity increases as the pH increases.
3. The effluent is more toxic to fathead
minnows than to Certodaphnia or Daphnia.
Example VI. Oxidants. In effluents, oxidants other
than chlorine may be present. Measurement of a chlo-
rine residual (TRC) is not enough to conclude that the
toxicity is due to an oxidant.
In general, oxidants are indicated by the following:
1. The addition of sodium thiosulfate to the
effluent reduced or removed the toxicity.
2. Aeration without any pH adjustment
removed or reduced toxicity.
3. The sample is less toxic over time when
held at 4°C (type of container is not an
issue here).
4. Ceriodaphnia are more sensitive than
fathead minnows.
Of course, TRC greater than 0.1 mg/L at the efflu-
ent LC50 concentration (and depending on test spe-
cies) would indicate chlorine as the oxidant causing the
toxicity. In addition, the dechlorination with SO, pro-
vides evidence of chlorine toxicity in the same manner
as the sodium thiosulfate addition test.
9-5
-------
Section 10
References
Adelman, I.R..LL Smith Jr., and G.D. Siesennop.
1976. Acute Toxicity of Sodium Chloride,
Pentachlorophenol, Guthipn®, and Hexavalent
Chromium to Fathead Minnows (Pimephales
promelas) and Goldfish (Carassiusauratus). J. Fish. .
Res. Board Can. 33:203-208.
Amato, JR., D.I. Mount, E.J. Durban, M.T. Lukasewycz,
G.T. Ankley and E.D. Robert. 1992. An Example of
the identification of Diazinon as a Primary Toxicant
in an Effluent. Environ. Tox. and Chem: In Press.
Ankley, G.T., J.R. Dierkes, D.A. Jensen and G.S.
Peterson. 1991. Piperonyl Butoxide as a Tool in
Aquatic Toxicological Research With
Organophosphate Insecticides. Ecotoxicol. Environ.
Safety. 21:266-274.
Ankley, G.T., M.T. Lukasewycz, G.S. Peterson and
D.A. Jenson. 1990. Behavior of Surfactants in
Toxicity Identification Evaluations. Chemosphere.
21:3-12.
APHA. 1980. Standard Methods for the Examination of
Water and Wastewater. 15th Edition. American
Public Health Association, American Water Works
Association Water Pollution Control Federation,
Washington, D.C.
Berg, E. 1982. Handbook for Sampling and Sample
Presentation of Water and Wastewater. EPA-600/
4-82-019. Cincinnati, Ohio.
Bowman, M.C., W.L. Oiler, T. Cairns, A.B. Gosnell and
K.H. Oliver. 1981. Stressed Bioassay Systems for
Rapid Screening of Pesticide Residues. Part I:
Evaluation of Bioassay Systems. Arch. Environ.
Contam. Toxicol. 10:9-24.
Campbell, P.G.C. and P.M. Stokes. 1985. Acidification
and Toxicity of Metals to Aquatic Biota. Can. J.
Fish Aquat. Sci. 42:2034-2039.
DHEW. 1977. Carcinogens • Working With Carcinogens.
Public Health Service, Center for Disease Control,
National Institute of Occupational Safety and Health.
Department of Health, Education and Welfare.
Publication No. 77-206.
Dowden, B.F. and H.J. Bennett. 1965. Toxicity of
Selected Chemicals to Certain Animals. J. Water
Pollut. Control Fed. 37(9):1308-l 316.
EPA. 1979. Aqueous Ammonia Equilibrium • Tabulation
of Percent Un-ionized Ammonia. EPA-600/3-79/
091. Environmental Research Laboratory, Duluth,
MN.
EPA. 1985A. Methods for Measuring the Acute Toxicity
of Effluents to Freshwater and Marine Organisms.
Third Edition. EPA-600/4-85/013. Environmental
Monitoring and Support Laboratory, Cincinnati, OH.
EPA. 19858. Ambient Water Quality Criteria for
Ammonia. EPA-440/5-85-001. Environmental
Research Laboratory-Duluth, MN, and the Criteria
and Standards Division, Washington, D.C.
EPA. 1988A. Methods for Aquatic Toxicity Identification
Evaluations: Phase I Toxicity Characterization
Procedures. EPA-600/3-88-034. Environmental
Research Laboratory-Duluth, MN.
EPA. 19888. Short-Term Methods for Estimating the
Chronic Toxicity of Effluents and Receiving Waters
to Marine and Estuarine Organisms. EPA-600/4-
87-028. Environmental Monitoring and Support
Laboratory, Cincinnati, OH.
EPA. 1989A. Toxicity Reduction Evaluation Protocol for
Municipal Wastewater Treatment Plants. EPA/600/
2-88/062. Water Engineering Research Laboratory,
Cincinnati, OH.
EPA. 19898. Generalized Methodology for Conducting
Industrial Toxicity Reduction Evaluations (TREs).
EPA/600/2-88/070. Water Engineering Research
Laboratory, Cincinnati, OH.
EPA. 1989C. Methods for Aquatic Toxicity Identification
Evaluations: Phase II Toxicity Identification
Procedures. EPA/600/3-88/035. Environmental
Research Laboratory, Duluth, MN.
EPA. 19890. Methods for Aquatic Toxicity Identification
Evaluations: Phase III Toxicity Confirmation
Procedures. EPA/600/3-88/036. Environmental
Research Laboratory, Duluth, MN.
EPA. 1989E. Short-Term Methods for Estimating the
Chronic Toxicity of Effluents and Receiving Waters
to Freshwater Organisms. Second Edition, EPA/
600/4-89/001 and Supplement EPA/600/4-89/001 A.
Environmental Monitoring and Support Laboratory,
Cincinnati, OH.
EPA. 1991 A. Toxicity Identification Evaluation:
Characterization of Chronically Toxic Effluents,
Phase I. EPA/600/6-91/005. Environmental
Research Laboratory, Duluth, MN.
EPA. 1991 B. Methods for Measuring the Acute Toxicity
of Effluents to Freshwater and Marine Organisms.
Fourth Edition. EPA-600/4-90/027. Environmental
Monitoring and Support Laboratory, Cincinnati, OH.
EPA. 1991C. Short-term Methods for Estimating the
Chronic Toxicity of Effluents and Receiving Waters
to Freshwater Organisms. Third Edition. EPA/600/
4-91/002. Environmental Monitoring and Support
Laboratory, Cincinnati, OH.
10-1
-------
Federal Register. 1984. U.S. EPA: Development of
Water Quality Based Permit Limitations for Toxic
Pollutants; National Policy. EPA, Volume 49, No.
48, Friday, March 9, 1984.
Ferguson, W.J., K.I. Braunschweiger, W.R.
Braunschweiger, JR. Smith, J.J. McCormick, C.C.
Wasmann. N.P. Jarvis, D.H. Bell, and N.E. Good.
1980. Hydrogen ion buffers for biological research.
Anal. Biochem. 104: 300-310.
Giles, M.A. and R. Danell, 1983. Water Dechlorination
by Activated Carbon, Ultraviolet Radiation and
Sodium Sulphite. Water Res. 17(6): 667-676.
Hockett, J.R. and D.R. Mount. In Preparation. Use of
Metal Chelating Agent to Differentiate Among
Sources of Toxicity. Manuscript.
Magnuson, V.R., O.K. Harriss, M.S. Sun, O.K. Taylor,
G.E. Glass. 1979. Relationships of Activities of
Metal-Ligand Species to Aquatic Toxicity. ACS
Symposium Series, No. 93. Chemical Modeling in
Aqueous Systems, E.A. Jenne, Editor, pp. 635-
656.
Neilson.A.J., A.S. Allard, S. Fischer, M. Malmberg, and
T. Viktor. 1990. Incorporation of a Subacute Test
with Zebra Fish into a Hierarchical System for
Evaluating the effect of Toxicants in the Aquatic
Environment. Ecotox. and Environ. Safety 20:82-
97.
Norberg-King, T.J., E.J. Durhan, G.T. Ankley and E.
Robert. 1991. Application of Toxicity Identification
Evaluation Procedures to the Ambient Waters of
the Colusa Basin Drain. Environ. Tox. and Chem.
10(7):891-900.
OSHA. 1976. Occupational Safety and Health
Administration. OSHA Safety and Health Standards,
General Industry. 29 CFR 1910. OSHA 2206
(Revised).
Patrick, R., J. Cairns Jr., and A. Scheier. 1968. The
Relative Sensitivity of Diatoms, Snails, and Fish to
Twenty Common Constituents of Industrial Wastes.
Prog. Fish-Cult.:1 37-140.
Poirier, S.H., M.L Knuth, C.D. Anderson-Buchou, L.T.
Brooke, A.R. Lima, and P.J. Shubat. 1986.
Comparative Toxicity of Methanol and N,N-
Dimethylformamide to Freshwater Fish and
Invertebrates. Bull. Environ. Contam. Toxicol.
37(4) :615-621.
Randall, T.L. and P.V. Knopp. 1980. Detoxification of
Specific Organic Substances by Wet Oxidation. J.
Water Pollut. Control Fed. 52(8):2117-2130.
Schubauer-Berigan, M.K. and G.T. Ankley. 1991. The
Contribution of Ammonia, Metals and Non-polar
Organic Compounds to the Toxicity of Sediment
Interstitial Water from an Illinois River Tributary.
Environ. Toxicol. and Chem. 10(7):925-939.
Smith, R.M. and A.E. Martell. 1981. Critical Stability
Constants. Volume 4: Inorganic Complexes. Plenum
Press, NY. ^
Stumm, W. and J.J. Morgan. 1981. Aquatic Chemistry-
An Introduction Emphasizing Chemical Equilibria in
Natural Waters. John Wiley & Sons, Inc., New
York, NY.
Walters, C.I. and C.W. Jameson. 1984. Health and
Safety for Toxicity Testing. Butterworth Publ.,
Woburn, MA.
10-2
-------
TECHNICAL REPORT DATA
tftmt n»d Inicrucnoni on '*« revtnt btfort comettrint)
NO,
EPA/600/6-91/003
4. TITLE ANOSUBTITLI
Methods for Aquatic Toxicity Identification Evaluation
Phase I Toxicity Characterization Procedures (Second
Edition)
PB92-100 072
t. REPORT OATI
February 1991
4, PERFORMING ORGANIZATION CCDE
7 AUTHORIS) , 3 1
Norberg-King, T.J."1",, Mount, D."".. Durban, E."1",'Ankley,
G. T. , Burkhard, L. , Amato, J. , Lukasewycz, M. ,
NIZATIQN
NO
J,PERFORMING ORGANIZATION N A M E"ANU"AOOB ESS
United States Environmental Protection Agency,
Environmental Research Laboratory, 6201 Congdon Blvd.
ouiuih. MN 558^4;-'ASCI Corp, Duluth, MN 55804; DVS".
ii. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADO«6SS
Environmental Research Laboratory
Office of Research and Development
J. S. Environmenta1 Protection Agency
)uluth, MN 55804
13. TYPE OPREPOHTANQPEftlOOCOVERED
14. SPONSORING AGENCY CODS
EPA-600/03
s. SUPPLEMENTARY NOTES
EPA Series Research Report
IS. ABSTRACT
ABSTRACT
In 1988, the first edition "Methods for Aquatic Toxicity Identification Evaluations:
Phase I Toxicity Characterization Procedures" was published (EPA, 1988A), This
second edition provides more details and 'more insight into the techniques described in
the 1988 document. The manual describes procedures for characterizing the
physical/chemical nature of toxicants in acutely toxic effluent samples, with
applications to other types of samples such as receiving water samples, sediment
pore water or elutriate samples, and hazardous wastes. '
7.
Kir WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
ENDED TERMS C, COSATl Fietd/Cioup
Author (s) Contmued:. ~
Anderson-Camahan, L. .
Performing Org. Name and Address Continue^:
Environmental Protection Agency, Region
IV-Policy Planning and Evaluation Branch,
345 Courtland, Atlanta, GA 30365.
OlSTR)BUTlON STATEMENT
Release to public
19,SECURITY CLASS
Unclassified
3\ NO. Of *AG1
20. SECURITY CLASS (Thu p»t*l
Unclassified
22. PRICE
fttm JJJO.I
------- |