&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

-------