vvEPA
          United States
          Environmental Protection
          Agency
            Environmental Research
            Laboratory
            Duluth MN 55804
National Effluent Toxicity
Assessment Center
          Research and Development
                        EPA/600/3-88/034 Sept 1988
Methods for Aquatic
Toxicity Identification
Evaluations

Phase I  Toxicity
Characterization
Procedures

-------
                                           EPA-600/3-88/034
                                              September 1988
         Methods  for Aquatic

Toxicity Identification  Evaluations

 Phase I Toxicity Characterization Procedures
                   Donald I. Mount
           U.S. Environmental Protection Agency
            Environmental Research Laboratory
           Office of Research and Development
                Duluth, Minnesota 55804
               Linda Anderson-Carnahan
                  U.S. EPA Region IV
               Water Management Division
                Atlanta, Georgia 30365
                National EffluentToxicity
                  Assessment Center
                Technical Report 02-88
                              U.S. Environmental Protection Agency
                              Region 5,1 'hnr- r-<  s - ••,
                              77 WestJ&c-."'. •"'       -  -  -.
                              Chicago, JL  60604-.,.. /' 	'  r"J^

-------
                                 Notice
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 the 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 permit. It  will also help  provide  U.S.
Environmental Protection Agency (EPA) and State Pollution Control Agency staff with
the background necessary to overview  and  determine  the adequacy of effluent TIEs
proposed and performed by NPDES permittees.

The  approach is divided into three  phases.  Phase I contains methods to identify the
physical/chemical nature of the constituents causing 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
remove toxicity without specific identification of the toxicants.  Two  EPA TRE  manuals
in  draft stage  (EPA, 1988A; 1988B)  use  parts  of  Phase  I  in  developing  those
approaches.

Phase  II (Mount  and Anderson-Carnahan,  1988)  describes  methods  to  specifically
identify the toxicants if they are non-polar organics, ammonia, chlorine or metals. This
Phase  is  incomplete because methods for other specific  groups, such  as polar
organics, have not been developed. As additional methods are developed, they will be
added.

Phase  III (Mount, 1988) 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.  Phase III has been only infrequently done in TIEs completed to  date. To avoid
Phase  III may be to invite disaster because the suspected  toxicants were not the true
ones.

Phases I and II depend  on acute toxicity and cannot be used  for effluents that do not
have it.  Importantly, however,  that limitation does  not mean  that  effluents having
chronic limits cannot be evaluated  using  these methods.  So long as  there  is acute
toxicity, even though it may be non-lethal toxicity, the methods can  be used.

These  methods are not  mandatory but  are intended to aid  those  who need  to
characterize, identify or confirm  the cause of toxicity in effluents or similar aqueous
samples such as leachates. Where we lack experience,  we have so indicated and tried
to  provide suggested avenues. All tests need not be done  on every sample; the tests
are, in  general, independent. Experience has taught us, however, that skipping tests is
likely to  result in  wasted time, especially during the early  stages of Phase I. An
exception  is when  one  only  wants to know  if  a specific  substance, for example
ammonia, is causing the toxicity or  if there are other  toxicants than  ammonia.
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.
                                       in

-------
                                 Abstract
This manual  describes  procedures for characterizing the physical/chemical  nature  of
toxicants in acutely toxic effluent samples. To detect the presence and potency of the
toxicants as the sample is manipulated the measurement of toxicity using organisms is
essential. The final step is to separate the toxicants from other sample constituents  to
simplify the analytical process. Usually the toxicants must be concentrated for analysis.

Sample manipulations  to alter toxicity include sparging, pH changes, filtration, solid
phase extraction and addition of chelating and reducing agents. The results will often
reveal information about the physical/chemical characteristics of the toxicants.

Subsequent  manuals in preparation describe methods to specifically identify the
toxicants and to confirm that the suspected constituents are the true toxicants.
                                        IV

-------
                                          Contents


                                                                                           Page

Foreword   	   iii
Abstract  	   iv
Contents  	   v
Figures 	   vii
Tables  	   ix
Acknowledgments  	   x

    1.     Introduction   	   1-1
          1.1   Background  	   1-1
          1.2   Conventional Approach to TIEs   	   1-1
          1.3   Toxicity Based Approach  	   1-3

    2.     Health and Safety  	   2-1

    3.     Quality Assurance	   3-1

          3.1   General  	   3-1
          3.2   TIE Quality Control Plans  	   3-1
          3.3   Cost Considerations/Concessions 	   3-1
          3.4   Variability   	   3-2
          3.5   mtra-Laboratory Communication   	   3-2
          3.6   Record Keeping   	   3-2
          3.7   Phase I Considerations  	   3-2
          3.8   Phase II Considerations  	   3-3
          3.9   Phase III Considerations   	   3-3

    4.     Facilities and Equipment  	   4-1

    5.     Dilution Water   	   5-1

    6.     Effluent Sampling and Handling   	   6-1

    7.     Toxicity Tests	   7-1

          7.1   Principles   	   7-1
          7.2   Test Species  	   7-1
          7.3   Toxicity Test Procedures  	   7-2
          7.4   Test Endpoints   	   7-3
          7.5   Feeding  	   7-4

-------
                               Contents (continued)


                                                                                    Page

8.     Phase I Toxicity Characterization Tests	   8-1

      8.1   Initial Effluent Toxicity Test   	   8-4
      8.2   Baseline Effluent Toxicity Test   	   8-4
      8.3   pH Adjustment Test   	   8-6
      8.4   pH Adjustment/Filtration Test  	   8-12
      8.5   pH Adjustment/Aeration Test  	   8-18
      8.6   pH Adjustment/Cis Solid  Phase Extraction Test	   8-22
      8.7   Oxidant Reduction Test  	   8-26
      8.8   EDTA Chelation Test	   8-29
      8.9   Graduated pH Test  	   8-33

9.     Time Frame and Additional Tests  	   9-1

      9.0   Time Frame for Phase I Studies  	   9-1
      9.1   When Phase I Tests are Inadequate   	   9-1

10.   References  	   10-1
                                          VI

-------
                                         Figures


Number                                                                               Page

1-1.    Conventional approach to TIEs  	   1-2
1 -2.    Flow chart for toxicity reduction evaluations   	   1-4
6-1.    Example data sheet for logging in samples  	   6-2
7-1.    Schematic for preparing effluent test concentrations using simple
         dilution techniques   	   7-4
8-1.    Overview of Phase I effluent characterization tests   	   8-2
8-2.    Example of data sheet for initial toxicity test   	   8 - 5,
8-3.    Example of data sheet for definitive baseline toxicity data  	   8-7'
8-4.    pe-pH diagrams for the C02, H2
-------

-------
                                          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  	  8-13
8-3.     Toxicity of Methanol to Several Freshwater Species  	  8-26
8-4.     Toxicity of Sodium Thiosulfate to  Ceriodaphnia, Daphnia, and
           Fathead Minnows	  8-27
8-5.     Toxicity of Disodium EDTA to Ceriodaphnia dubia and Fathead Minnows in
           Water of Various Hardnesses and Salinities   	  8-30
                                             IX

-------
                            Acknowledgments
The authors wish to acknowledge the help from many members of the Environmental
Research  Laboratory (ERL-Duluth)  staff for their  help and advice in  preparing this
manual. Throughout the text, we have referred to "experience" to suggest an approach
and we mean  the  collective  experience of the ERL-D effluent research  group. That
group consisted of Liz  Durhan and Teresa Norberg-King of ERL-Duluth,  Joe Amato,
Larry  Burkhard, Art Fenstad, Jim Jenson, Marta  Lukasewycz, Greg Peterson, Eric
Robert, and Jim Taraldsen of American Scientific International, Inc. (AScI), Duluth.

Special mention should be made of the invaluable help Dorette Gueldner (AScI) has
been, not only in  typing but  in figure preparation,  "on-the-spot" corrections, and
generally seeing that details were in order. Joe contributed an immense amount  of
work  upon  which  we  based many  of the perceptions  as  well  as  specific
recommendations.  Jim, Greg, and Gary Ankley (ERL-D)  completed tests, generating
data  for some of the tables.  Larry and  Joe helped  in final reviews and last minute
changes.  We  gratefully acknowledge  Teresa's  work  in  diligently reviewing the
document, preparing the figures and tables, and making editing changes on the drafts
and final version.

Outside the effluent group,  but still within ERL-D, Evelyn Hunt was very helpful  in
refereeing the  review comments.  Without the always present backing and support  of
Nelson Thomas (ERL-D), the effluent  group would not have been able to complete
such a task.

Rick Brandes,  U.S. EPA Permits Division, Washington, D.C., has been a strong and  at
times a crucial voice in support of all the work  upon which this manual is based. His
funding support has enabled  more than  a doubling  of the  staff of the National Effluent
Toxicity Assessment Center  (NETAC) at ERL-Duluth.

Finally,  we  want  to  specifically  recognize  Bill Clemente  of  Battelle Laboratories,
Columbus, OH, for the most thorough,  by far, and useful technical review comments
on the  first draft  of the manuscript and largely  on  his  personal  time!  Burlington
Research, Inc., Burlington, NC, and EA  Engineering, Science & Technology, Sparks,
MD reviewed the first draft.

The  manual is truly the result of the  efforts  of  many  people.  We welcome your
suggestions for improvement so that  future revision can  make  the methods  more
useful.

-------
                                              Section 1
                                            Introduction
1.1  Background

The  Clean Water Act (CWA) 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
prohibited."  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  amount."  State  narrative and
numerical  water quality  standards  are  used  in
conjunction with  EPA criteria and other toxicity data
bases  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
discharges are met,  EPA has issued a "Policy for the
Development  of  Water   Quality-Based   Permit
Limitations for Toxic  Pollutants" (Federal  Register,
1984). This national  policy recommends an integrated
approach  for controlling toxic pollutants that utilizes
whole  effluent  toxicity  testing  to  complement
chemical-specific  analyses.  The use  of whole
effluent  toxicity testing is necessitated  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 (Peltier
and Weber, 1985; Horning and Weber, 1985). These
cost-effective methods 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 organism
toxicity limits and biomonitoring  conditions in permits,
a substantial number of unacceptably toxic  effluents
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
reduction  evaluations (TREs).  The  object of  the
Aquatic  TRE is to  determine  what measures  are
necessary  to  maintain  the effluent's  toxicity  at
acceptable  levels.  Such  evaluations, however, have
often proven to be very complicated.

The  goal of the  TRE  will  be  set by  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 this, and  because specific TRE actions may also
be required,  communication between the regulators
and TRE investigators 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 whole
effluent aquatic organism toxicity limits or permittees
required, through  special conditions, to  reduce  or
eliminate effluent toxicity. This  document does  not
address human health toxicity concerns such as from
bioconcentration, water supply and recreational  uses.
Neither are the methods  applicable to identifying  the
cause of chronic toxicity except for those  effluents
which also display acute toxicity.

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
problem. The  following  discussion is  meant  to
exemplify  the need  for  a  logical  approach  which
builds on the effluent data as it is being collected.
                                                 1 - 1

-------
Traditionally, when an effluent has been identified as
toxic  or is suspected of  being toxic  to aquatic
organisms, a sample  of the  wastewater  is analyzed
for the 126 "priority pollutants". The concentration of
each  priority  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.  The goal of this  exercise is to determine
which  pollutants in  the  wastewater sample  are
responsible  for  effluent  toxicity   (Figure  1-1).
Unfortunately, determining the source  of an effluent's
toxicity is rarely this straightforward.
 Figure 1-1. Conventional approach to TIEs.
                                 Evaluate Effluent
                                 Constituents and
                                      Their
                                  Concentrations
                   Search Literature
                      for Aquatic
                   Organism Toxicity
                    Data on Effluent
                     Constituents
The first problem encountered in this course is one of
effluent variability.  Because  toxicity  is  a  generic
response, 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
periodically  acting to  cause  effluent toxicity.
Experience  has shown  that the  latter  may be a
frequent occurrence especially in  privately  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(s).  Monitoring
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  because the toxicants  are  not amenable to
analysis by these procedures.  For this 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(s).

A  second problem with  the conventional approach
involves the focus  on the priority pollutants.  These
have  become  known as the  "toxic  pollutants,"
conveying an  implication that  they  constitute the
universe of toxic chemicals. 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
monitoring 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.  Gas  Chromatography/Mass  Spectrometry
(GC/MS)  will  not  identify cadmium.  Inductively
Coupled Emission Spectroscopy (ICP) may  not detect
it when the  concentration is low. The absence of a
measurable  quantity of  any  substance is  often
interpreted as  meaning  that it  is not present in the
effluent.

The toxicants may  be present  at low concentrations
because  only  small concentrations  of  highly  toxic
chemicals are needed to produce toxicity. If 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
ug/L, 380 grams,  or less than a pound per day of the
compound  is  necessary  to cause  lethality  in the
effluent of a 100  million  gallons per day treatment
plant. Even  with  a removal efficiency of  99%, only
100 pounds per  day loading  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 million gallons  per day  may be
next to impossible  to  identify  by  usual  methods  of
establishing 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 compounds (G. Veith,  ERL-Duluth, personal
communication). This limitation  is related to selection
and efficiency of  solvent extraction  techniques,
analyte volatility  and thermal 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 sensitive and
require  higher  concentrations  of analytes  for
detection. To  detect lower concentrations,  more
                                                 1 -2

-------
specific  methods  are usually more  sensitive.  To
choose specific methods one must have knowledge
of the  toxicants-knowledge  which  does  not  exist
since that is the purpose of the analyses.

Surprisingly, even  with these limitations,  one usually
sees lengthy  lists  of  effluent constituents when
analyses are performed on wastewater. In the case of
GC/MS  chromatograms, large peaks of  non-toxic
effluent  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  the toxic  one.  Using reference
spectra,  many may be tentatively identified as several
different compounds which only serves to increase,
not decrease,  the  number of possibilities.  No aquatic
toxicity  data will  be available  for  most  of these
compounds and toxicity data  must  be  generated.
Compounds often  must be synthesized  in order  to
test  them  because they  are not commercially
available. For those  compounds 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 likely  to fail  because  this
approach does not include matrix effects and toxicant
bioavailability.  For  example, several  metals may be
present in an  effluent sample at concentrations  well
above the toxic  threshold.  These metals  may not be
the source of the effluent's toxicity, however, because
they  are not  biologically  available.  Characteristics
such as  total organic carbon  (TOC),  total suspended
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  utilizes 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
toxicants 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  conventional approach can be  used to
enhance the success 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
constituents  to simplify  analyses   and  enhance
interpretation of analytical data.  Secondly, throughout
the TIE, one must establish whether or not the toxicity
is  consistently caused  by  the same substances.
 Failing  to  establish  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
 analytical  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
 procedures (Phase  I). Phase II (toxicant identification)
 and  Phase III confirmation will usually follow Phase I.
 Two other EPA manuals (EPA, 1988A; 1988B) can be
 consulted for  more information  on bench  scale  and
 pilot  plant  effluent toxicity  treatability  studies  and
 source control options.

 Phase I characterizes physical/chemical properties of
 the  effluent toxicant(s) using effluent  manipulations
 and   accompanying   toxicity  tests.  Each
 characterization test in the Phase I series is designed
 to alter or  render biologically unavailable a group of
 toxicants such as oxidants,  cationic metals, volatiles,
 non-polar  organics  or  metal  chelates. Aquatic
 toxicity tests,  performed  on  the  effluent before  and
 after the individual characterization treatment, indicate
 the effect of the treatment and provide information on
 the nature of the toxicant(s).  By  repeating the  toxicity
 characterization  tests using  samples of a particular
 effluent collected over time, these screening tests will
 provide information  on  whether  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  certain cationic  metals, non-polar  organics,
 oxidants, substances  whose toxicity is pH dependent
 and  others.  Information  on  physical/chemical
 characteristics of the toxicants will indicate filterability,
 degradability,  volatility, and  solubility.  Either  of  two
 choices is available in the second  phase of testing:
 toxicant treatability or toxicant identification studies.

 The  toxicant  identification  option  is  described  in
 Phase II (Mount and Anderson-Carnahan,  1988).
 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, chlorine, and ammonia. Additional procedures
                                                 1 -3

-------
Figure 1-2. Flow chart for toxicity reduction evaluations.
                                                  Phase I
                                         Toxicant Characterization Tests
          Treatability
          Approach
                                            Treatability Approach
                                             or Identify Toxicant
                                                          Identify Toxicant(s)
                                                 Phase II
                                         Toxicant Identification Analyses
                                                 Phase III
                                        Toxicant Confirmation Procedures
                                            Based on Site Specific
                                               Considerations
                   Toxicity Treatability
                      Evaluations
~- —
^
r
Source
Investigation
                                           Control Method Selection
                                             and Implementation
                                            Post Control Monitoring
for other  toxicants  will  be  added  as  they  are
developed. Once the toxicants have been adequately
isolated from other compounds in the effluent  and
tentatively  identified  as the causative  agents,  final
confirmation  (Phase III)  can begin.

Like  Phase  I,  Phase  III  (Mount,  1988)  contains
methods generic  to all toxicants.  No single  test
provides irrefutable proof  that a certain chemical is
causing 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 on untreated wastes because toxicity from
other constituents  that are toxic  in  untreated  waste
but removed 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 so that repeated testing can characterize and
subsequently identify and  confirm the  toxicants in
Phases II and III. Therefore, enough testing should be
done to assure consistent presence of toxicity before
TIEs are initiated. This is done not to validate a given
test but to establish consistent presence of toxicity.

The methods described  herein are applicable  only  to
acute toxicity. Much work remains to be done before
chronic toxicity methods are developed and proven.
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  chronic toxicity.

To  be successful,  TIEs must be  conducted  by
multidisciplinary teams whose team 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  often
decay  during  storage. Often  subsequent tests cannot
be designed  until the results of the previous ones are
known. Obviously then,  waiting a week for analytical
or toxicological results may preclude more work when
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
then effectively being used.
                                                   1  -4

-------
                                             Section 2
                                         Health and Safety
Working with effluents of unknown composition is the
nature of toxicity identification evaluations. Therefore
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
aeration is likely to minimize volatiles and chlorinated
effluents are less likely to contain viable pathogens.

Exposure to the wastewater  during collection and its
use in the laboratory should  be minimized. Inhalation
and  dermal adsorption  can be reduced  by  using
plastic gloves,  laboratory aprons or coats, safety
glasses,  respirators,  and  laboratory  hoods. Further
guidance on health and safety  for toxicity testing 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 toxic to  humans.  Analysts
should familiarize themselves  with  safe  handling
procedures for  these chemicals  (DEHW,  1977;
OSHA,  1976). Use of these compounds may  also
necessitate specific waste disposal practices.
                                                2- 1

-------

-------
                                              Section 3
                                         Quality Assurance
3.1  General
Quality  assurance  is  composed of two aspects,
quality  verification  and quality control.  Quality
verification 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  taking
place such as the number of samples to be taken and
the mode of collection, standard 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 (QAP) may increase the probability of
success.

In  preparing a QAP, enough detail should  be  included
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 QAP should be provided in  a
narrative form  that encourages users to  think  about
quality assurance. To be  effective, the QAP  must be
more than a paper exercise simply restating standard
operating  procedures.  It  must  increase
communication  between  clients,  program planners,
field  and laboratory personnel and data analysts.  The
QAP must make clear the specific responsibilities of
each  individual.  The  larger  the staff,  the  more
important this becomes. While QAPs may  seem to be
an inconvenience, the amount of effort they require is
commensurate with the benefits derived.

3.2  TIE Quality Control Plans

A  successful TIE  is dependent upon  a strong quality
control program. Obtaining  quality TIE data  is more
difficult because  the constituents  are  unknown  in
contrast to quality control procedures for a  standard
analytical method for a specific chemical. In such an
analysis, one knows the characteristics of the analyte
and the implications of the analytical procedure being
utilized. Without knowledge  of  the  physical/chemical
characteristics of  the analyte, however, the impact of
various 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 inappropriate to another.

The problem of quality control is further exaggerated
because quality control procedures for aquatic toxicity
tests  may be radically different from those  required
for individual chemical  analyses.  This  additional
dimension  to quality  control requires  a  unique
framework 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  standard  aquatic toxicity tests,  samples
with  low dissolved oxygen are  usually aerated. This
practice may, however, result in a  loss of toxicity if
the toxicant is volatile or subject to oxidation.

3.3  Cost 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  experiments 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
Peltier and Weber (1985) (involving  a relatively high
degree of quality  control) must be weighed against
the degree of complexity involved, the time  required
and the number 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
                                                3- 1

-------
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 results
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
pollutant. 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 results  must be  of  sufficient quality to  ensure
ecosystem  protection.  The use of dilution water  in
toxicity  tests  which  mimics  receiving  water
characteristics will help to ensure that the effluent will
remain non-toxic  after  being  discharged into the
environment. 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.
3.4  Variability
The opportunities to retest any effluent to confirm the
quality of initial results will  be limited at best.  In
addition  to  the shifting chemical and toxicologic
nature of the discharge over time, individual effluent
samples  stored in  the laboratory change.  Effluent
constituents 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.


3.5  Intra-Laboratory Communication
Quality control  procedures  in  chemistry and  biology
can be quite different.  For  example, phthalates  are a
frequent analytical  contaminant  requiring special
precautions that are not of  toxicological concern. The
toxicological 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 advance.  One  cannot assume
compound  stability;  therefore,  time  delays between
the biological  and  chemical  analysis  of  a  sample
cannot be tolerated.
3.6  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 crucial in  later stages  of the
evaluation. Investigators must record test results  in a
manner such  that  preconceived  notions  about  the
effluent toxicants  are  not unintentionally  reflected  in
the data. TIEs required by state  or federal pollution
control agencies may require that  some or all records
be reviewed.


3.7  Phase I Considerations
Effluent toxicity is "tracked" through Phases I, II  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 ground truth. Therefore,  the toxicity
test  results  must  be  dependable. System  blanks
(blank  samples  carried through  procedures  and
analyses identical 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  intentionally 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,  contaminated  air or  nitrogen  sources, poor
mixing  of test solutions, contaminants  leached from
filters,  pH probes  and  SPE columns,  the  reagents
added  and their contaminants. LC50 data  for several
reagent  chemicals  and  common  aquatic  test
organisms are  provided as  needed in subsequent
sections of this document.

Frequently toxic artifacts are unknowingly introduced.
For example, pH meters  with refillable electrodes 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
electrodes 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
                                                 3-2

-------
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  disposable test chambers
are recommended to avoid  problems 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  caused  either  by  other
contaminants or  by residues of  cleaning  agents.
Since the organisms are sensitive to all chemicals at
some concentrations, 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
approximately the same age  ensure  quality data.
Standard  reference toxicant  tests  should   be
performed with  the aquatic  test species on a  regular
basis and control  charts  should be developed  (Peltier
and Weber, 1985). During Phase I it will  not be known
how much the toxicity of the reference toxicant varies
over time  compared to  the toxicant(s).  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.


3.8  Phase // 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
identifications  are  made,  interferences  can  be
determined. Likewise instrumental response, degree
of toxicant  separation, and detector sensitivity can be
determined as identifications proceed.


3.9 Phase III 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, confidence
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 chemical analysis and toxicity testing are minimized
in this phase in an effort to decrease the chance  for
production  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,  spiking experiments  are
important in determining recovery for the toxicant(s).
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, 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(s).
The  equipment required for  Phase I  characterization
tests can be found throughout Section 8. The facility
and equipment needs in Phase II of the TIE will  be
site-specific   and  will  depend  both  on  the
physical/chemical characteristics 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 doing  toxicity tests and performing  the
usual  water   chemistry  analyses.   Phase  III
requirements are largely  limited to equipment found in
a typical toxicity  testing lab and equipment necessary
for the analyses of the toxicant(s).

Because of the equipment needs and  time required to
conduct  the  evaluations,  complete  on-site effluent
TIEs using  a mobile laboratory  are  generally  not
feasible. 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   TIE tests. Ready  availability  of test
organisms is important because  often  needed tests
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
purchase  them.  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
conditions 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
conditions. In some cases, the dilution  medium may
cause complete mortality in 48 hours, 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  comparison  of the toxicity of  two
chemicals, A and  B. If one determines LC50s  for A
and B  and concludes  that A is twice as  toxic  as B,
lethal  conditions are being compared 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 question is, "which
is more toxic?".

Where these methods  are built on tolerance, chronic
toxicity endpoints  cannot  be used and  that is why
these methods are intended only 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 methods would
not  allow  sufficient  survival  time  or  health  for
reproduction. Attention  would also have to  be given to
acclimation,  feeding and general living conditions.
Because of these factors, choice of dilution water in
Phase I is of much less concern both because these
are acute tests  and because of the  many additives
used  which change the mixture much more than the
dilution water changes it. In general for Phase I,  any
water which is of consistent quality and will support
growth and  reproduction  of the  test species is
suitable.  In  Peltier and Weber (1985), a  variety of
dilution water  choices is provided 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 excluded
to the extent possible and  where absolute toxicity is
more   important  than relative toxicity,  practices,
including choice  of dilution  water,  need  to  follow
conventional  toxicological  methodology.  Tolerance
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
changes. 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 Phase I and II except that
the desired tested conditions will  often  dictate its
characteristics. For example, in Section 8.6, the same
column 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  should be  based  on 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 are representative of the discharge
over time. Guidelines for determining the number and
frequency of samples required to  represent  effluent
quality are contained in  the "Handbook 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
concentrations  of  individual toxicants,  different
toxicants,  changing  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, TOG and TSS. The toxicant(s) may be present
at such a low level that it does not significantly affect
the quantity 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 recorded.  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 and conductivity
should  be measured.  Total  residual  chlorine,  total
ammonia, alkalinity, dissolved  oxygen  (DO),  and
organic  carbon  measurements  may   also  be
appropriate. Toxicity  should be measured periodically
during storage to document any changes.

Investigators should not be surprised  to find that well
operated municipal and industrial  treatment systems
discharge unacceptably toxic wastewaters. Effluent
guideline-based limits  reflecting  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 wastewater analysis.

For chlorinated effluents, whether sampling should be
done before chlorination depends  on  the question to
be  answered.  Sometimes  the question  may  be
whether there are  toxicants other  than  chlorine
present.  Dechlorination prior to toxicity character-
ization  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 depend
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
                                                 6- 1

-------
Figure 6-1.   Example data sheet for logging in samples.
  Sample Log #:	
  Date of Arrival:
  Facility: _
  NPDES #:
  Location:
  Contact:
  Phone Number:
  Sampler.^
Sample Type:
    Grab
        Collected,
        Date
                                                                                                     AM/PM
                                                                                                     AM/PM
    Composite
        Collected From_
        Date	
        To	
                                                                                                     AM/PM
                                                                                                     AM/PM
                                                                             Date
Sample Conditions Upon Arrival:
    Temperature	
    PH	
                                                                          Total Alkalinity_
                                                                          Total Hardness
    Conductivity/Salinity _
    Chlorine
                                                                          Total Ammonia
                                                                          Total Organic Carbon_
  Condition of treatment system at time of sampling:
  Status of process operations/production (if applicable):
  Comments:
                                                          6-2

-------
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 maximum effluent toxicity, it is
more difficult to catch  peaks in  toxicity and  Phase I
sampling may require more time.

Peltier  and  Weber (1985)  have  discussed the
advantages and disadvantages of grab and composite
sampling  and  have  also  detailed  methods  for
sampling intermittent discharges.

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  characterization tests, one  of  which  entails
filtering the effluent. 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, 3.5 liters 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  used  in  each
characterization 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
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 effluent.  Likewise, there is not  a set
number of samples which  should be analyzed  in
Phase 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.
Table 6-1.   Volumes Needed for Phase I Tests
   Characterization Step
Volume for
Each Step
    Total
Volumes1 (ml_)
Chemical analyses              -2      <1000

pH 3 Adjustment               30       - 300
    filtration                 235
    solid phase extraction       200
    aeration                  35

pH 11 Adjustment               30       —300
    filtration                 235
    solid phase extraction 3      200
    aeration                  35
Unadjusted pH effluent (pH|)
initial test
baseline
filtration
solid phase extraction
aeration
EDTA chelation
oxidant reduction
Gradual pH changes
pH6
pH7
pH 8

40
80
235
200
35
100
100

40-500
40-500
40-500
-590







-120-1500



1 Total volume is — 3.5 L; this is maximum needed, does not
 include subsequent testing.
2 Amount  for  this step is  dependent on  effluent
 characteristics.
3 The pH is readjusted to pH 9 before it is put through the
        column.
                                                  6-3

-------

-------
                                              Section 7
                                           Toxicity Tests
7.1  Principles
Acute  lethality tests  with  aquatic  organisms are
utilized  throughout  the  toxicity characterization
procedures 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 tool 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
because  it is  not   specific.  This  non-specific
response necessitates a  generic  chemical/physical
characterization of toxicants  during  Phase  I testing
before specific identification is begun in Phase II.

A further repercussion of this universal response  is
the probability  of  artifactual  toxicity.  Because the
analyst 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  can
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 and controls 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 evaluation.
7.2  Test Species

Just as  different  analytical methods  have  different
detection  levels for the  same chemical,  different
species  have  different sensitivities  to the  same
toxicants.  The  major difference  is  that the  toxicity
measurement is non-specific to chemicals and so in
an unknown mixture (an effluent) one must determine
whether a different  toxicity value  for  the effluent is
caused by 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 and adaptability to test conditions, there are other
important  considerations.  An  effluent  toxic to  two
species, having equal or different LC50s may be toxic
because  of different toxicants.  Differences of 1.000X
in sensitivity are common  and differences of 10,OOOX
occur among species exposed to  a  single chemical.
Anyone involved in identifying the cause of toxicity of
an  effluent will  be  concerned  when  someone  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  Ceriodaphnia or  Daphnia,
species well suited to TIE methods.  TIE test species
selection is obvious in these instances. Where toxicity
concern is based  on species (trout  or mysids),  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  toxicity  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  suggestions  are
certainly  not all   inclusive. The  final confirmation
(Phase III) methods are designed to show whether the
wrong  toxicant was  identified.  However,  many
resources  may be consumed  before   reaching  that
                                                 7 - 1

-------
stage and earlier assurances should be  obtained if
reasonable, to save time and cost.

One approach  is to compare the  LC50  values  of
whole  unaltered effluent  samples for  the species
originally raising the  toxicity  concern and the selected
species 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(s). 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.

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  as well. Because different
species have dissimilar  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).

Symptoms are useful, especially  if one is comparing
like organisms.  Comparing fish symptoms to Daphnia
symptoms could be very misleading but comparing
Daphnia magna symptoms  to those of Ceriodaphnia
dubia symptoms should be relatively safe.  If one finds
like  symptoms, the  evidence  is  not  convincing
because many toxicants cause specific symptoms but
if symptoms are distinctly different,  the evidence 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 go through the symptoms so fast, or  skip some
symptoms,  that  they would  appear  completely
different from  those of the less sensitive species.
Experience will  reveal additional techniques 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 enough  so
that marine species  can be used in the TIE. However,
the tolerances  of marine organisms to the additives
and  effluent manipulations have not been determined.
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
different freshwater species.  When   Phase  III  is
reached, marine species should be used, but in that
phase, manipulations and  additives are  minimal and
little  ancillary 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
refraction) are  non-specific  for  NaCMhe  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  species  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, removing stress  (due to  low dissolved
oxygen (DO), other  contaminants, and  lack of space)
is important because such stresses may change the
sensitivity of the  organism  to  the contaminant  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. Therefore,
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.

The  reason  for  this  discussion under test methods is
that  effort must be  made  to make  the tests used in
Phase I  as cheap  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
measuring water chemistries had been done on these
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 careful  assurance that the stresses are similar
among comparisons. For  example, it does not matter
if the test organisms are acclimated to a pH change.
                                                7-2

-------
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, erroneous 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
several  purposes of Phase III is to catch  errors or
artifacts that may creep into Phases I and II.

One need not use standard  acute methods in Phase  I
for  these  reasons. The following mechanics  of
performing an  acute test with cladocerans and newly
hatched  fathead  minnows has   been found  by
experience to be very cost effective and is 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 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
cup  in the series contains  only dilution water and
serves as  the control. Mixing the  solutions prior to
aliquot transfer 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  Na2S203 and EDTA (Phase I), or effluent
fraction solvent concentrates  (Phase II) are  added to
effluent or dilution water.

The need for duplicates will depend on the  accuracy
and  precision  required  of  the  test results.  Tests
requiring  a  measure of accuracy in  the   form  of
confidence 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),  designed  to
define effluent toxicity upon arrival  in  the laboratory
and periodically during testing, respectively,  Phase  I
toxicity tests usually  do not require  preparation  in
duplicate.
The test organisms  of  uniform  age  should  be
randomly placed  in  each test  cup to better  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 transferred  with  the organism is reduced to a
drop (50 pL), 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.  Such situations may
be  encountered in  Phase II studies, when limited
volumes  of  effluent  fraction  concentrates may
necessitate the testing of  cladocerans in volumes of
even 1 mL or less.

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 Phase III. However, sometimes, in order to  maintain
the desired conditions in the test  (such as maintaining
a specific pH) frequent specific measurements  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 will change sometimes quickly  if equilibrium is
not already established. POTW effluents are not in air
equilibrium 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 hold at 8.2-8.5.
If pH is  important to test  interpretation,  pH  must  be
monitored throughout the test.

7.4  Test Endpoints

Little effort should be expended in calculating LC50
values  for Phase  I toxicity tests. There is no use in
applying  sophisticated and complex programs to test
results. A number of  methods of  estimating the LC50
from the acute toxicity data  are  described in  Peltier
and Weber (1985).  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.  For specific chemicals the TU  is equal  to the
concentration of the compound present in the effluent
divided by the  LC50  of the compound.  The TU  of
whole effluent is 100% divided   by  the LC50 of the
effluent.  For example, if  the   48 hour LC50  of
compound A  is 3 mg/L, a solution  of  1 mg/L of this
compound  contains  0.33  TU.  If  the LC50  of  an
effluent is 25%, the effluent  contains 4 TU (100/25).
By normalizing  the  concentration term to a unit  of
                                                7-3

-------
 Figure 7-1.    Schematic for preparing effluent test concentrations using simple dilution techniques.
                      Effluent
                      • Add 10 mL to each replicate
                       for the high concentration.
                      • Add 10ml_to next
                       concentration
                 Dilution
                  Water
     • Add 10 mLto each
       replicate except in
       the high concentration
                  High
                  Cone
                                                                                 Waste
toxicity (such as the LC50), the TU allows the toxicity
of chemicals  and/or  effluents  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 concentrations each equal one
LC50.  Phase  III contains  more  discussion about
adding toxic units; however one  must be cautious in
summing  toxic  units.  Unless toxicants  are strictly
additive,  simple summation of  toxic units  will  be
incorrect.
7.5  Feeding

Most species used in acute testing are not fed during
the test. In our  use of  Ceriodaphnia,  especially  in
Phase II, it was necessary to  perform tests in water
with  very low  dissolved  solids.  In such water,  very
small  concentrations of  contaminants  (e.g.,  metals
from glassware) can be toxic. Therefore, we routinely
feed all animals in test exposures at  the beginning  of
the tests.  This includes  the  100% effluent.  This
practice  has become standard  in all three  phases.
The decision will be  species specific and dependent
on the  characteristics  of the effluent.  Consistency
throughout the  three  phases is most important. All  of
the data for Ceriodaphnia given in tables in Section  8,
are based on tests in which animals were fed.
                                                  7-4

-------
                                              Section 8
                              Phase I Toxicity Characterization Tests
The first phase of a TIE involves a characterization of
the toxic effluent.  The  characterization information
gathered 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 toxicity  removal  or alteration is performed on
the whole effluent. Acute toxicity tests utilizing aquatic
organisms are  used to determine whether the toxic
chemicals  have  certain  physical  or chemical
characteristics.  Two  objectives  are  accomplished
during 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.  For ammonia and  chlorine, Phase I results may
be  rather convincing that  they  are the  cause  of
toxicity but otherwise Phase I only provides evidence
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 development
of  bench-scale wastewater treatment processes
(EPA,  1988A;  1988B) or in  choosing  separation and
analytical procedures for toxicant  identification as
described in Phase II.

The tests described in  this  section are designed for
acutely toxic effluents.  Methods  for  chronic toxicity
have  not yet been developed.  The methods  in this
section are based on the use of small  test organisms.
If  larger species  are  used  modifications  to these
methods will have to be  made.

Analysis  of samples should begin as soon as practical
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 the laboratory as well as  in transit,
the bulk effluent should  be  held  near 4°C and kept
headspace free.  Those  characterization  solutions
(Sections 8.3-8.6) held  prior to  serial dilution  and
toxicity  testing  should  also be  kept  at 4°C  and
covered  to minimize loss by evaporation. If the test
organism to be used is  sensitive to supersaturation,
then supersaturation must be removed.  Ceriodaphnia
are not very "sensitive" to such situations but newly
hatched fathead minnows are. Once in the laboratory,
testing on individual samples may continue indefinitely
provided that whole  effluent  toxicity stabilizes.  The
degree of toxicity can remain similar, but the cause of
toxicity can change with age. Especially in the early
stages,  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
Section 8 is designed to change toxicity of groups of
constituents  (Figure 8-1). Toxicity  before and after
the characterization treatment will indicate for  which
groups the toxicity was changed. All but one of these
tests  is  performed at the  same time in  order to
minimize confounding effects resulting  from sample
degradation over time. While it is not critical that each
characterization treatment be performed  at exactly the
same time, the toxicity  tests should be  initiated at
approximately the same time. One species of  test
organism should be used throughout the initial stages
of Phase I.  Other  species may be useful  in the later
stages of Phase I.

For each day following receipt of the effluent sample
the various steps  are given in  Table 8-1. Day 1 is
when the  sample arrives in  the  laboratory. On  day 1
initial  physical/chemical  measurements  are taken  for
the effluent  sample and an  initial toxicity test is
conducted  on an  aliquot of the sample.  This LC50 is
used  to  set the  desired exposure  conditions for
subsequent  toxicity tests  and  is termed  "initial"
toxicity to  distinguish it from  the  "baseline" toxicity
described below.   Other  aliquots of the sample are
adjusted  to pH 3  and  11,   filtered, aerated  and/or
chromatographed  using a C-\Q solid phase extraction
(SPE) column.  Following these manipulations, each
effluent aliquot is  readjusted to the  initial pH  (pH|) of
the effluent. These aliquots  and  the remainder  of the
effluent are then held at 4°C overnight. 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 hour  LC50 of the
                                                 8- 1

-------
    Figure 8-1.   Overview of Phase I effluent characterization tests. (Note: pH( stands for initial pH.)
       Baseline Toxicity
        Test (Day 2)
                                            Toxic Effluent Sample
                           Initial Toxicity Test
                               (Dayl)
Acid

PH|
                             Base
                                                                        C-, 8 Solid Phase
                                                                        Extraction Tests
                                                                            (Day 2)
Acid

pH,
Acid

pH|
pH6

pH 7

pH8
day 1 initial toxicity test. This procedure also allows
pH  adjusted effluent aliquots more time to  stabilize,
and  additional  pH  adjustments  can  be  made as
necessary.

On  the  second day the aliquots of effluent  prepared
on  day 1  are  diluted to 4X-,  2X-, 1X-, and  0.5X-
LC50 (24  hour)  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.) Other toxicity tests
involving  the  addition  of chelating and  reducing
agents  and less severe  pH adjustments  are  also
conducted. A  second toxicity  test  is  begun  using
unaltered  effluent, now  24 hours old. The results of
this and subsequent whole effluent tests are referred
to as the "Baseline" effluent LC50.

Toxicity test results  are  read on subsequent testing
days and depending 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  tasks scheduled for  each day
is approximately one full day.  If at 24 hours,  less than
50% lethality of test organisms exposed to 100% day
1 effluent has occurred, the sample can be discarded
and a new sample collected with relatively little loss of
resources  or  time.  Alternatively,  the  test  can be
continued to  48 or 96 hours which  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 100%,  50%,
25%, 12.5%, 6.25% effluent.

Several Phase I  characterization tests are relatively
broad in scope, intended to include more than one
class of toxicants. Therefore, if a significant change in
effluent  toxicity  is   seen   following  these
characterization  procedures, additional  tests are
needed to further delineate 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
effluent 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  replace  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 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
                                                  8-2

-------
 Table 8-1.   Outline of Phase I Effluent Manipulations
 	Description	Section
 DAY 1 SAMPLE ARRIVAL
 Chemical analyses                              6.0
      • temperature
      • pH
      • chlorine
      • hardness
      • alkalinity
      • conductivity
      • ammonia
      • TOC
      • DO
 "Initial" toxicity test                              8.1
 Sample manipulation:
      • pH adjustment (pH 3, pH(, pH 11)             8.3
      • pH adjustment/filtration                      8.4
      • pH adjustment/aeration                      8.5
      • pH adjustment/Gig solid phase extraction        8.6
 DAY 2 TOXICITY TESTING:
 Warm effluent samples from Day 1 and set-up
    toxicity tests
      • baseline toxicity                           8.2
      • pH adjustment samples                     8 3
      • aeration samples                          8.4
      • filtration samples                          8.5
      • Gig solid phase extraction samples            8.6
      • oxidant reduction samples                   8.7
      • EDTA chelation samples                    8.8
      • graduated pH samples                      8.9
 Read mortality on "initial" toxicity test                8.1
 DAYS 3 AND 4 MONITORING  TESTS:
 Read 48 h mortality initial toxicity test                8.1
 Read 24 h and 48 h mortality on all tests from Day 2  8.2-8.9


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\Q SPE  column  volume,  increased reagent
concentrations).

Another outcome of the Phase I characterization test
series may be  that several tests succeed in partially
removing  effluent  toxicity. In  this situation,  one may
be dealing with several toxicants, each  with different
physical/chemical characteristics, or  a single toxicant
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
removal  is enhanced as  compared  to the reduction
provided  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
characterization  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 overlap  somewhat in
their  abilities   to  remove groups  of  toxicants.  For
example,  increasing pH  may cause  a  metal  to
precipitate and EDTA may also remove  its toxicity. In
any case, results of this nature are useful in  selecting
Phase II options.

When several  treatments  are applied  to the  same
sample, tests  must be designed to ensure that toxicity
does not result from the  additives used (acid,  base,
EDTA)  rather than from  the effluent's toxicants. The
assumption must not  be  made  that toxicants  are
either additive or synergistic.  Our  experience  shows
that independent action  (one  or  more of  multiple
toxicants act independently of  the  rest, as though the
others  were   not present)  is not  uncommon  in
effluents. Experience also shows that one should not
use selected   tests to  confirm a  suspicion that  a
certain  toxicant  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 the helpful and more often channel thinking and
delay final solution. 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 ammonia or  whether  there  are
toxicants present other than salt. These questions are
quite  different from the former case where one is
playing "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
characterization 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.  A  toxicity blank,
however,  does  poorly in identifying artifact toxicity if
toxicity  is  affected by  the  wastewater  matrix. For
example,  the  toxicity  of a Phase  I reagent, EDTA,
may be completely different in dilution water and in
effluent. If so, a toxicity  blank  is inappropriate for the
chelation test. A "toxicity control",  on the other hand,
involves  a comparison of the  test solution  and the
                                                   8-3

-------
Baseline  effluent.  In this case, the comparison must
demonstrate that  the  effluent  test solution has  not
become more  toxic than the unaltered effluent.  If it
has, the test  procedure  has produced  artifactual
toxicity.  For  some treatments, valid blanks or toxicity
controls cannot be made.

No  procedure should  be assumed  to be  free of
artifactual toxicity. Many of the Phase I toxicity tests
involve  relatively  severe or  unorthodox effluent
manipulations.  Blanks  and  controls  must   be
consistently  and conscientiously  used to detect  the
introduction of toxic artifacts  or other changes to the
effluent that increase sample  toxicity.

8.1  Initial Effluent Toxicity Test

Principles/General Discussion:
The major  purpose of the  Initial effluent  test is to
provide an estimate of the  24 hour LC50 for purposes
of setting exposure concentrations in Phase I tests.

Volume Required:
Initial toxicity test is performed in duplicate exposure,
and 40 ml of effluent is needed.
Apparatus:
Disposable one ounce test chambers,  automatic
pipette (10 mL), disposable pipette tips (10 ml_),  eye
dropper  or  wide  bore pipette,  light  box  and/or
microscope (optional).

Test Organisms:
Test organisms, 60 or more, of the same age  and
species.

Procedure (Day 1):
A  concentration series using 10 mL in duplicate of
100%, 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

Principles/General Discussion:
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
treatment in the laboratory, must be determined.  The
portion of the effluent sample, tested for toxicity the
day after it arrives in the  laboratory (day  2),  will be
referred  to as the "Baseline effluent".  The Baseline
effluent LC50  will be  compared  to results of toxicity
tests  initiated  on day  2 on aliquots  of  the  effluent
carried through  characterization tests.  Such a
comparison will demonstrate whether  the removal or
alteration  of  various  groups of  toxicants changes
effluent toxicity. Thus, by comparing these results, an
indication  of  the physical/chemical  nature  of  the
toxicants can  be obtained.  If the Baseline  effluent
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 revised  to
reduce delay in testing. If Phase  I testing is extended
to additional days, Baseline tests must be done each
time on succeeding days, and used for comparison to
these later tests.


Volume Required:
The Baseline  toxicity  test is performed  in duplicate
using  10  mL per  replicate.  The  total  volume
necessary will depend on the 24 hour  LC50 of the
day 1 initial effluent test, but 80  mL will be adequate
for most species.


Apparatus:
Disposable  one  ounce test chambers  or glass
beakers,  automatic  pipette  (10  mL),  disposable
pipette tips  (10  mL), eye  dropper  or wide bore
pipette, light box and/or microscope (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. One  test series  will
provide exposures at effluent dilutions  of 100%, 50%,
25%,  12.5%   and  6.25% (or as  appropriate if  the
effluent is  more  toxic).  This  series will  enable a
comparison of the  results of the Baseline  (day  2)
effluent test to the Initial effluent LC50 (cf.,  Section
8.1).  In testing^ the  day 2  effluent, any obvious
physical  changes, (e.g., formation of  precipitates,
odors) which  occurred during storage, should be
noted.

The exposure levels in the  other test series will be
based on the 24  hour LC50 of the Initial  (day  1)
toxicity and will include day 2 effluent concentrations
at  4X-,  2X-,   1X-,and  0.5X-LC50. Most  of  the
toxicity tests  with  the characterization solutions  will
also  be  performed  using  these same  exposure
concentrations. If the  24 hour  LC50 of the  initial
effluent is greater than 25%,  the  series  should  begin
at 100%,  include four exposure  concentrations and
the lowest concentration  should  equal  one-half  the
                                                 8-4

-------
Figure 8-2.   Example of data sheet for initial effluent toxicity test.
Test Type: INITIAL
Test Initiation (Date
Investigator:
EFFLUENT
& Time):

Sample Loq #, Name:
Date of Collection:


                                                                   Species/Age: _
                                                                   No. Animals/No. Reps:_




                                                                   Source of Animals:	




                                                                   Dilution Water/Control:,




                                                                   Test Volume:	




                                                                   Other Info:	

Cone. (% effluent)
100
50
25
12.5
6.25
Control








24 h
A B










LC50 B
c> I

Survival Readings'
48 h
A B










ILC50 H
c> 1


72 h
A B










ILC50 1
ci §


96 h
A B










HLCSO
Hci

             Comments:
                                                     8-5

-------
LC50. In this case, the method for making dilutions
described earlier  must be changed. If the 24  hour
LC50 of the day 1  initial effluent is equal to 25%, the
second exposure series will be unnecessary because
the first series described fulfills the requirements for
comparison  to the initial effluent test and  character-
ization  solution  toxicity test results. Throughout this
section, the alternate series will be assumed  when
4X-LC50 exceeds  100%.

A sample data sheet is shown  in Figure 8-3. In order
to  compare the Baseline effluent  toxicity  and  the
toxicity  of  the  effluent  aliquots subjected  to
characterization tests, all of the day 2 toxicity tests
should  be initiated at approximately the same time.

The  Baseline toxicity  test should  be repeated each
time additional  characterization tests are  performed
on the sample  after the  initial Phase  I battery. The
Baseline test will  serve as the basis for determining
the effects produced by the additional characterization
tests, and will also provide information on degradation
of  sample   toxicity. If  sample toxicity  is  greatly
changed compared to  the toxicity  of the sample on
the day of its arrival (e.g., greater  than 50% change)
it is  advisable to  discard the  remaining sample and
collect  a fresh one.

/nterferences/Contro/s and Blanks:
The  control treatment in this  test  is  used  for
comparison to several subsequent tests and provides
an important reference for diluent water acceptability.
Mortality in  these controls will negate other work.

Results/Subsequent Tests:
Baseline LC50's should be  generated  for as long  as
the effluent sample is being used and a Baseline test
should  be started  every time the 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
adjustment  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  well  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
impacted by solution pH  are  acids and bases. To
understand 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, Ka, for the proton transfer
reaction.

            HA + H.O = H 0+ + A~
                    £      
-------
Figure 8-3.   Example of data sheet for definitive baseline effluent toxicity data.
Test Type: BASELINE EFFLUENT
Test Initiation (Date & Time):
Investigator:
Sample Log #,
Date of Collect

Name:
ion:

                                                                        Species/Age:_
                                                                        No. Animals/No. Reps:.




                                                                        Source of Animals:	




                                                                        Dilution Water/Control:_




                                                                        Test Volume:	




                                                                        Other Info:     	

Cone. (% effluent)
100
50
25
12.5
6.25
4X-LC50/
2X-LC50/
1X-LC50/
0.5X-LC50/
Control


• L(
c


24 h
A B












350 Hll
1 •

Survival Readings:
48 h
A B












LC50 Si
ci •


72 h
A B












LC50 H
ci •


96 h
A B












ILC50
C,

                  Comments:
                                                          8-7

-------
B: unprotonated base
Kb: thermodynamic equilibrium constant for the base.
e.g., C6H5NH2  + H20 =
                                    +  OH'
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+]
Note:
e.g.,   C6H5NH3-1-  + H2O = H30+ +  C6H5NH2
       conjugate acid
             K =
              a
                                =2.34xlo-6
                   [C6H5NH3+]
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
pKa 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 pKa  of  ammonia) will be  found  as
50% NH4+  and 50%  NHs. At one pH unit above the
pKa (i.e., 10.25) roughly 90% of the ammonia will  be
in the  un-ionized  form (NHs) and the remainder will
be in the NH44- form.  At pH 8.25, one unit below the
pKa of ammonia, approximately  90% of the ammonia
will be in the NH4+ form, and approximately 10% will
be in the NHa form.
The above can be summarized by the following:
                     Predominant Species
                   Organic
                                Inorganic
          pH>pK,
              acid  RCOO", RCO"   A'
              base  RNH2         B
          pH < pK,
              acid  RCOOH, RCOH  HA
              base  RNH3*        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 un-ionized  form as  compared  to
the  ionized form. For  example,  NH3  is  generally
recognized as the toxic form of ammonia while NH4 +
is  of far less  concern.  A second  implication  of this
effect relates  to toxicant solubility.  Un-ionized 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 (Section  8.5) or  extraction
with non-polar  solvents or  solid  phase column
techniques  (Section 8.6).  Likewise,  changes  in
compound  solubility  with pH  change many 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
activity, (in a  simple sense,  whether the system  is
aerobic or anaerobic), one can see how various forms
of  manganese are  created and eliminated as pH
shifts.
                                                     Figure 8-4.   pt-pH diagrams for the CO2, H2O, and Mn-CO2
                                                                 systems (25°C). Solid phases considered:
                                                                 Mn(OH)2(s) (pyrochroite),  MnCO3(s)  (rhodo-
                                                                 chronsite), Mn3O<(s) (hausmannite), r-MnOOH
                                                                 (manganite), r-MnOz (nsutite). (Reprinted with
                                                                 permission from Stumm & Morgan, 1981.)
                                                                                                   c
                                                                                                   CD
                                                                                              -  1.5
                                                     Each  of the different  forms  of  the 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
                                                 8 - 8

-------
chemical constituents  present in the  water.  The
hydrolysis rate of organics is greatly affected by pH,
and pH changes may alter organic toxicity as well.

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  membrane  permeability at  the respiratory
surface 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 sometimes results in reduction or loss
of toxicity. If the kinetics of the  pH driven reaction (on
return  to the original  effluent  pH)  are  slow or
irreversible, pH adjustment alone  may be effective in
evidencing  toxicants  affected  by  pH  change. Some
organics  may also degrade due to pH change.

Another purpose of the "pH Adjustment Test" is to
provide  a blank  for  subsequent Phase I pH
adjustment tests  performed in combination with other
operations. This  test will  demonstrate whether toxic
concentrations of ions have been  reached as a result
of the addition of acid and base.

Toxicity blanks for sample ionic  strength increases,
which are based on  adding  the  same  volumes and
strengths of the  acid and  base *o  dilution  water, do
not give comparable results  when  added to the
effluent.  Effluents  already   contain  substantial
concentrations of major  anions and cations, which are
not found in dilution water. Further, the  volumes and
strengths of  the  acid  and  base  necessary, for
example, to lower  an effluent  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.


Volume Required:
600 mL are needed to  make pH adjustments. A 300
ml_ aliquot of the day 1  effluent sample  is lowered to
pH  3 with four  exposure concentrations,  and the
second  300 mL  sample is raised  to  pH  11.
Approximately 30 mL will  be  needed for the test but
the actual amount depends on the 24 hour LC50 of
the initial effluent test. The remaining 270 mL of each
of  these  solutions  is  reserved  for   the  pH
Adjustment/Filtration, pH Adjustment/Aeration and pH
Adjustment/Cia SPE Phase I tests.
Apparatus:
Burettes for acid and  base titrations, pH  meter and
probe, 2-500  mL  beakers, 2-500 mL  graduated
cylinders, 12-30 mL beakers, stir plate, and stir bars
(perfluorocarbon),  automatic  pipette,  disposable
pipette tips,  eye dropper  or wide bore  pipette, light
box and/or microscope (optional),  pH  meter  and
probe.
Reagents:

1.0,  0.1  and 0.01  N  NaOH,  1.2, 0.12 and  0.012 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 Adjustment test is
shown in Figure 8-5.  Using a burette,  and  stirring
constantly, 1.0  N NaOH is added dropwise to a 300
mL aliquot day  1 effluent until the solution pH nears
11.0.  (Note:  overshooting results  in the addition  of
more  salts  and volume   may  cause  toxicant
decomposition  and should  be avoided.)  In order  to
minimize any over-adjustment  of 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 11.0.  If pH  11  is
exceeded, 0.12 N HCI must  be used to lower the pH
to  11.0.  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.
Volumes and strengths of base (and any acid added)
should be recorded.  A  30 mL  volume is held for the
same length of  time it takes to complete other Phase
I  manipulations  (performed  on day 1)  with  pH  11
effluent.  Once other  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|)  of
the day  1 effluent.  (The  other   aliquots  of  pH  11
effluent are also returned  to pH| at this time.) This is
accomplished by the slow, dropwise addition  of 0.12
N HCI  first and later  0.012  N  HCL  as the pH of the
stirred solution nears pH|. If pH| is exceeded,  the  pH
must be  appropriately increased  with 0.01 N  NaOH.
Again,  the volumes  and  strengths  of acid and  any
base added should be recorded.
The pH of the solution should be checked periodically
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 entire procedure is repeated with a second  300
mL aliquot of effluent, using 1.2 N and 0.12 N HCI to
lower the pH to 3. As with the pH 11 effluent, 270 mL
of  the  pH  3  effluent  is   used  for  the  pH
                                                8-9

-------
 Figure 8-5.   Flow chart for pH adjustment tests.
                                                  Effluent Sample 600 mL at pH|
                            300 ml, adjusted to pH 3
270 mL at pH 3

30 mL at pH 3
     Day 1
                                              4
Use for other
  Phase I
   tests:
                                            HOLD
                                            Use for other
                                              Phase I
                                               tests:
                                                                        i
•Filtration
•Aeration
•C18SPE
                                         'Adjust to pH|
•Filtration
•Aeration
•C18 SPE
                                                                 •Adjust to pH|
     Day 2
                                         Set up Toxicity
                                            Tests
Adjustment/Aeration, pH Adjustment/Filtration, and pH
Adjustment/Cia SPE tests. The remaining 30 mL of
the pH 3 effluent is held until the first day's 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 periodically checking and readjusting
the sample 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 (on 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 recorded. The acute
toxicity  of each  pH-adjusted solution  is  tested at
4X-,  2X-,  1X-,  0.5X-LC50 (the  24  hour  Initial
LC50) as described in Section  7. Test solution pH
should be measured and recorded every 24 hours.  A
sample data sheet is shown  in Figure 8-6.
                                                                Set up Toxicity
                                                                    Tests
                                Interferences/Controls and Blanks:

                                Controls prepared for the Baseline toxicity test act as
                                a  check on the organisms, dilution water, and test
                                chambers for this test as well.

                                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. The 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 adjustment tests, increased toxicity following
                                pH adjustment,  not as a result of NaCI concentration,
                                indicates a pH effect on  toxicity (see below).

                                Results/Subsequent Tests:

                                If  either the pH 3 or pH 11  adjusted  effluent  tests
                                have significantly greater toxicity than the Baseline
                                effluent test, two possible sources of toxicity exist:  1)
                                the  ions  (Na + , Cl") added  by  the  acid  and  base
                                have resulted  in a  solution with  an  ionic  strength
                                intolerable  to  the  test  organism;  or 2)  chemical
                                                 8-1 0

-------
Figure 8-6.   Example of data sheet for pH adjustment test.
  Test Type:  pH ADJUSTMENT
  Test Initiation (Date & Time):
  Investigator	
  Sample Log #, 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/
3/
3/
3/
11/
117
11/
IV
•


24 h
A pH








9 LC50 1

Survival Readings:
48 h
A pH








H LC5°


72 h
A pH








• LC50


96 h
A pH








• LC50
11
               300 mL pH 3
               300 mL pH 11
               30 mL pH 11
               30 mL pH 3
                                      HC1
                                                                                 NaOH
               Comments:
                                                  8-1 1

-------
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 dilution
water-based blank. To help resolve this situation a
listing  of  NaCI  LC50 values  for common  test
organisms  is 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  dissolved substances.  The toxicity  of  the
added NaCI is best determined by adding that amount
directly   to  the  effluent and see   if  the addition
increased effluent toxicity.

If  pH 3 and/or  pH  11  adjustment  results  in a
significant decrease in  effluent toxicity,  it could result
from volume changes  by acid and  base addition or
chemical reactions  driven  by pH   change may  not
have been  re-established  or are irreversible.  These
two possibilities can be checked by adding a  volume
of  dilution water equivalent to the total volume of acid
and base added to the 30 mL effluent volume.  If a
similar  loss in  toxicity of  the  diluted  wastewater
occurs,  the pH Adjustment test  should be repeated
using more concentrated acid and base.

A  reduction or loss of toxicity may also be the result
of  the  degradation of toxicant  at  the  altered  pH
values.  Both organics and  inorganics  can  be so
changed with  a probable  loss in   toxicity. In some
cases,   the  toxicity could   also  be  increased  if  the
degradation product  is more toxic  than the  original
compound.

For most of the Phase I combination pH adjustment
tests, the pH Adjustment test will act as an equivalent
or "worst case" control for changes in test solution
ionic strength and volume.  In effect,  most  of  the
operations  applied to  the  pH  adjusted effluent in
Sections 8.4-8.6  will  not affect  pH or will serve to
drive it  closer to  pH|. This may  not be the case for
the pH  Adjustment/Aeration test, however. Because
pH 3 and  11  must  be maintained throughout  the
aeration process and  because 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.
 8.4 pH Adjustment/Filtration  Test

 Principles/General Discussion:
 The filtration experiment  provides  information on
 effluent  toxicants associated with filterable material.
 Toxic pollutants associated with particles may be less
biologically  available. Aquatic  organisms  can  I
exposed to these pollutants through ingestion of tl
particles, however.  This  route of exposure  may  I
significant for cladocerans  and  other  filter  feede
ingesting bacterial cells and other solids with sorb<
toxicants. The degree to  which any compound exis
sorbed or in solution depends on a number of facto
including particle surface charge (or lack  thereo
surface area, compound polarity  and charge,  solubili
and the effluent matrix. By filtering particles from tr
effluent,  both a source and  a sink of  toxic chemica
may be removed.

In addition to determining the effect of filtration on th
toxicity  of  the whole effluent,  the  effects of  p
adjustment in combination  with  filtration  are  als
assessed. As discussed  in  Section  8.3,  changes i
solution  pH can result in the formation of insolubl
complexes  of metals (Figure  8-4). Similarly, organi
acids  and  bases  existing   in  ionic  form  can  b
transformed  into  the   non-ionic  form  by  pi
adjustment. Shifts in effluent pH can also  act to driv<
dissolved toxicants onto particles in the effluent  (e.g.
shifting  the dissolved/adsorbed equilibrium away fron
the free form). Changes  in  toxicant polarity  resultini
from   solution  pH  change  can   make sorru
particle/toxicant interactions  stronger.  In other cases
the increase  in effluent ionic strength resulting frorr
the  shift  in  pH  may  force  non-polar  organic
compounds onto uncharged  surfaces  to a greatei
extent.

By  filtering pH adjusted  aliquots  of effluents, those
compounds typically in solution at unadjusted pH bu1
insoluble or  associated  with particles to a  greater
extent  at  more  extreme  pHs,  are  removed.  By
removing the  toxicant  contaminated particles  or
precipitated compounds prior to  readjustment of the
sample to pH|, these toxicants are no longer  available
for  dissolution in the effluent. The pH change may
also  destroy or  dissolve  the  particles  thereby
removing  the  sorption surfaces   or  drive  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
degassing the solution during filtration. This  problem
is  potentially  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 reduced,  subsequent  tests must  be
performed to define the nature of the toxicity loss.

The  solid  phase  extraction characterization test
(Section 8.6) requires 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 column  or  the filtration  removed the  toxicity
Filtering may also be useful in connection with  other
Phase I tests.
                                                 8-12

-------
                Table 8-2.   Acute Toxicity of Sodium Chloride to Selected Aquatic Organisms

                                                                  LC50 (g/L) (95% Cl)
Species Water Type
Ceriodaphnia dubia DMWa.b





Daphnia magna \_\fjti.c

RW

Pimephales promelas DMWb

Sottb.d

Lepomis macrochirus Soft RWe

Life-stage
<24h

<24 h
<24 h
<24 h

NR

NR

<24h

11 wk

1-9g

24 h
4.2
(")
3.3
3.0
2.3
(2.0-2.6)
3.3
(NR)
6.4
(NR)
7.9
(7.0-9.0)
7.9
(NR)


48 h
2.3
(2.0-2.6)
2.7
2.1
2.3
(2.0-2.6)
3.1
(NR)
5.9
(NR)
7.9
(7.0-9.0)
-

-

72 h 96 h










6.9 4.6
(5.5-8.7) (7.4-7.9)
7.7
(7.4-7.9)
12.9
(NR)
                a Data for C. dubia and fathead minnows was generated at ERL-Duluth.
                 Both species were < 24 h old at test initiation and C. dubia were fed. Dilution water used was
                 diluted mineral water.
                b Static, unmeasured test.
                0 Dowden and Bennett, 1965.
                d Adelman et al., 1976.
                e Patrick et al., 1968.
                LW = lakewater; RW = reconstituted water; NR = not reported
                (--)  Confidence interval cannot be calculated as no  partial mortality occurred.
Volume Required:

A  235 mL aliquot of pH|,  day 1, 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  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. Each test requires
four exposure concentrations  (10  mL each).  The
exact effluent volume required for the toxicity  test will
be a function of the effluent toxicity (Section  7). The
remaining filtered effluent volumes (200 +  mL) of pH
3, pH 11, and the pH| solution  are each reserved for
the Cis SPE tests (Section 8.6). Excess  volume has
been included to cover losses  occurring  during the
filtration operation.
Apparatus:

Six-250 mL  graduate cylinders,  6-250 mL beakers,
6-50 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  1.0  pm,
without organic  binder),  stainless  steel forceps,
burettes  for acid and base titrations, pH  meter and
probe, stir  plate,  perfluorocarbon stir bars, automatic
pipette (10 mL), disposable pipette tips (10 mL), eye
dropper  or  wide  bore  pipette,  light  box  and/or
microscope (optional).

Reagents:
Solvents and high  purity water  for  cleaning  pump
reservoir and filter, 0.1 N and 0.01 N NaOH,  0.12 N
and  0.012  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):
The  first  step  for the  Filtration  Test  is filter
preparation, shown in Figure 8-7. First the filters are
prepared,  then  the   blanks,  and finally  effluent
samples at three pH's are filtered  (cf.,  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
effluent sample. Adsorption of toxic  dissolved organic
compounds onto the  filter can lead to  spurious
results.
                                                 8-13

-------
 Figure 8-7.   Overview of filter preparation and dilution water blanks for filtration test and Ctg solid phase extraction column
            test samples.
Da
=
Da
Preparing
the Filter
/1
Dilution
Water
Blanks

r
y2
L
300 mL High
I
HCI — ^r

T
Purity Water
r

1 + NaOH
100mLofpH3 100 mL of pH, (unadj.) 100mLofpH1l
4^
Filer
4-
•^ -^
Filter
Filter
4- 4
Discard water Discard water
Discard water

Dilution Water
I

pH3
4
1
pH
0
•i-

prepared filter prepared filter
4

200 mL 20 mL 200 mL
•*• 4 — NaOH +
4



pH 11
Jr
prepared filter
' *•

20 mL 200 mL 20 mL

NaOH Thru C18 Thru C18
| SPE Column SPE Column
L— + 	 I.
1 r i


Tox. Tox. Tox.
Test Test Test





1

HCI — >4 4- HCI
Thru Gig
SPE Column
HCI — »
4- IT
Tox. Tox. Tox.
Test Test Test

Filter Preparation
To  prepare  the  1.0  iim  glass-fiber  filter  for  use,
wash two 50 mL volumes of high purity water through
the filter.  For  pH 3 effluent filtration test,  the  filter
should be washed with  high purity water adjusted to
pH  3 using HCI. Likewise, the filter used with the pH
11 effluent samples should first be washed with high
purity water  adjusted to pH 11 using  a concentrated
NaOH solution.  Washing  the filters  with water
adjusted  to  the  same  pH  as the  effluent should
preclude  sample contamination with  water soluble
toxicants 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| filtration  blank is simply prepared by
passing  220  mL  of  pH  unadjusted  dilution water
through a prepared filter. The filtered dilution water is
collected and  200 mL of this volume is reserved for
the GIB SPE test blank (Section 8.6). The remaining
20 mL is used on day 2 as a blank in a toxicity test.
Again, excess is included to cover any possible loss
during rinses.

The procedures to prepare the pH 3 and  11 filtration
blanks are somewhat different. To prepare the pH 3
filtration blank, 255 mL of dilution water is  adjusted to
pH 3 with 0.12 and 0.012 N HCI, caution being taken
to  minimize the increase  in  dilution water ionic
strength.  Of the pH 3  dilution  water,  35  mL is
reserved  for use as  a blank in the  Aeration test
(Section 8.5).  The remaining 220 mL of pH 3 dilution
water  is passed through a  filter  previously washed
with pH 3 rinse water.  The  pH  3 dilution water is
collected and  200 mL of  this volume is reserved for
the pH 3 CIB SPE test blank.  The remaining 20 mL is
readjusted to the initial pH of the dilution water (pH0,
using 0.01 N NaOH,  again taking  care not to exceed
pH0  (pH0 symbolizes the  equilibrium   pH   of  the
                                                 8-14

-------
 Figure 8-8.   Overview of steps needed with the effluent for the filtration and C-| g SPE column tests.
    Day 1
                                                                                 NaOH

200 mL
+•

35 mL

Thru C}8
SPE Column*

U- NaOH
4

i
«—
- NaOH
200 mL
+•

35 mL

Thru Cig
SPE Column*

r ^

r *

HCI
HCI
200
mL
~H-

35 mL

Thru C18
SPE Column*
~>
r ^


r ^
* HCI
W
     Day 2
                                  "See Figure 8-13 and Section 8 6 for details.
dilution water) during the readjustment process. This
solution is used on day 2 in a single exposure toxicity
test.

The pH 11 blank is prepared in a similar fashion using
255 mL of dilution water adjusted to pH  11 with 0.1  N
and 0.01 N NaOH. Of the pH 11 dilution water, 35 mL
is reserved for use in the Aeration test. The remaining
volume is filtered using the filter previously washed
with pH 11 rinse water and 200 mL of the filtered pH
11 dilution water is collected for use as the basic C-\Q
SPE blank. The remaining  20 mL is readjusted to pH0
with 0.012 N HCI and used  on day 2 in a  single
exposure toxicity test.


Sample Preparation
The same filter used  to  prepare the  pH0  filtration
blank is now used to filter a 235 mL aliquot of the pH|
effluent. Effluent passed through the filter is collected
and 200 mL  is reserved for the GIB SPE test. The
remaining volume is held  overnight for toxicity tests
initiated on day 2.

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 200 mL and 35 mL aliquots. The 200 mL
aliquot is used in the pH 3 GIB SPE test. The 35 mL
aliquot is readjusted to pH| using  0.1 and 0.01  N
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 day 2 toxicity
testing.

Finally, the filtration step is repeated using 235 mL of
the  pH  11  effluent  (Section  8.3)  and the  filter
originally  used to  filter pH  11 dilution  water.  Again,
200  mL of the pH  11 filtered effluent is  used in  the
basic Ci8 SPE test; 30 mL is readjusted to the pH| of
the effluent with 0.12 N and 0.012 N HCI  and used to
conduct a toxicity test on day 2.

In filtering effluent samples with high solids content, it
may  be necessary to change filters in order to obtain
235  mL of filtered  effluent. If so,  the filter preparation
step  must be repeated to provide additional filtration
blanks. 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
on day 2 and  compared to that of  the Baseline
effluent.

In the above  procedures, the effluent filtration system
must be  cleaned  between  pH  adjusted  aliquots to
prevent any  carry-over. This means all equipment
should be thoroughly rinsed with 10% HNOa, acetone
and high purity water between aliquots of effluent.

The  pH  of the pH  adjusted blanks  and effluent
aliquots, designated for day 2 toxicity tests, should be
checked  periodically  throughout  the  work  day.
Adjustments should be made as necessary in order to
                                                 8-15

-------
maintain  pH0 and  pH|,  respectively,  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 the pH0 and
pH|, respectively.  Toxicity tests performed on all three
(pH 3, pH|, and pH  11) filtration blanks involve testing
without dilution. Based on the 24 hour 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-,  1X-,  and  0.5X-LC50, or  as
described in  Section  8.2. Measurement  of exposure
pH should be made  daily  on  the  highest tested
concentration, 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.  Results  of  the
effluent  filtration  tests  at  each  pH   should  be
compared with the  filtration  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 control organisms occurs in either or both
of  the  pH  adjusted  filtration  blanks,  further
investigation  will  be necessary to  determine whether
lethality resulted from toxicants leached from the filter
at pH 3 and/or 11, or whether the  increase in dilution
water ionic strength (via acid and base  addition)  is
responsible  for  the  problem.   This  can   be
accomplished by repeating  the filtration  step on pH
adjusted  dilution  water providing  a  pH  adjustment
blank (i.e., pH  adjusted  unfiltered dilution  water).
Additionally, if the pH 3 and/or 11 filtration, aeration
(Section  8.5) and  C-\Q  SPE (Section 8.6) dilution
water blanks have approximately  the  same  final
concentration of  acid  and  base,  any ionic  strength
related toxicity should be detected in them.

If a filtration blank shows unacceptable acute toxicity
but the corresponding  filtered effluent is  equally or
less toxic than the Baseline effluent, it is possible that
the dilution water blank removed  the final traces of
toxic filter artifacts.  In some cases, the effluent 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
contamination by  filter artifacts.
Results/Subsequent Tests:

The  LC50s  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 any or all of these pH/filtration combinations result
in less effluent toxicity (which cannot be attributed to
the   effects  of  pH  adjustment  alone),  it  may  be
possible to confirm the findings of the test. This can
be  attempted through a  transfer of  the  solids
contained on  the  filter back  into the filtrate at pH|.
This can be done  by reversing the flow of the filtrate
through the  filter or 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  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
filtrate spiked with  its  own solids.

We  have had limited  experience  with  effluents  in
which acute toxicity  could be removed by filtration at
a  normal  pH  (7.0-8.0).  The additional  tests
suggested  herein  may or   may  not  provide the
intended information.

If toxicity can  be removed by filtration,  either with  or
without  pH change,  one has  a  method for removing
the  toxicants from other material in  the effluent. This
knowledge  itself  provides  an  important  advance
because further characterization and analyses  will  be
less  confused by  those constituents separated from
the  toxicants.  Usually further characterization  will  be
the  next step. Tests  must  be designed  to determine
whether the  mechanisms   causing  removal are
precipitation,  sorption,  change  in  equilibrium  or
volatilization. One  usually necessary step is to learn
how to recover 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. Use
of   pressure  and   vacuum   filtration  may  reveal  if
                                                 8-1 6

-------
Figure 8-9.   Example of data sheet for filtration tests.
  Test Type: FILTRATION
  Test Initiation (Date & Time):
  Investigator:	
  Sample Log #, Name:_
  Date of Collection:	
Species/Age:	
No. Animals/No. Reps:.
Source of Animals:	
Dilution Water/Control:_
Test Volume:	
Other Info:
pH/Conc.
(% effluent)
3/
3/
3/
3/
3/ blank
PH|/
PH|/
PH|/
PH,/
pH| / blank
11
11
11
11
1 1/ blank
Survival Readings:
24 h
A pH















• 1
48h
A pH















72h
A pH















96h
A pH















1 LC50 ^BLC5° ffll-C50
• HI " Hi
                                                    8-1  7

-------
volatilization is involved. If  one gets an idea of the
toxicant 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.
8.5  pH Adjustment/Aeration  Test
Principles/General Discussion:

The Aeration test is designed to determine how much
effluent toxicity can be  attributed to  volatile  or
oxidizable  compounds. The test is  performed with
pH-adjusted and unadjusted effluent. By  comparing
the toxicity test results  for aerated  acidic, pH|  and
basic samples,  toxicity  may  be  changed  and  this
knowledge  can be  used  for further  characterization.
Some compounds  can be removed or oxidized  most
easily at one  pH,  whereas others  are most easily
removed or oxidized at  a different pH.  Thus, the
aeration is performed at several pH values.

Whether a constituent is  completely  removed,  or
sufficiently  removed to reduce  toxicity,  depends  on
many chemical/physical conditions.  At  a minimum,
one must be certain that the geometry of the sparging
process is  always  the same and  that the  duration is
constant. Otherwise, the test is of little value.  The pH
of many effluents  will change,  sometimes  rapidly,
during  sparging and so  pH  must be  frequently
checked and  maintained  during  the entire aeration
period.

Air is used for  sparging so that oxidation is included.
Subsequent tests  with  nitrogen  may  be used to
separate sparging  from oxidation.  We have grouped
them first,  to avoid many  tests  initially, and because
many treatment processes remove such constituents
well.  Oxidation  can change  many  constituents  in
many ways and one must determine if oxidation or
sparging is the mechanism before additional tests can
be designed.  Water  soluble constituents such  as
ammonia and cyanide are not readily stripped by this
test  and one  should not assume that they  will be
removed.
 Volume Required:

 Thirty-five mL volumes of pH 3, pH 11  (see Section
 8.3) and  pH| effluent are  needed  for this test.  A
 maximum volume of 30 mL of each of these solutions
 is required for the toxicity tests on  aerated solutions.
 An excess  volume has  been provided to  allow for
 losses without  removing it from the aeration vessel.
 Each   toxicity  test   utilizes   four  exposure
 concentrations  (10 mL  each).  The exact volume
 required for the toxicity test on each pH adjusted  or
 unadjusted aerated  solution will depend on the toxicity
 of the effluent (the 24 hour Initial LC50).
Apparatus:

Aeration  device or compressed  air system with a
molecular sieve, six glass diffusers, six-50  mL wide
graduated  cylinders, burettes for  acid  and  base
titrations,  pH  meter  and  probe,  stir  plate(s),
perfluorocarbon stir bars, automatic pipette  (10 mL),
disposable  pipette  tips (10 mL), eye dropper or wide
bore pipette, light box and/or a microscope (optional).


Reagents;
0.01 N NaOH, 0.012 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):
Six different solutions are aerated in this test; pH 3,
pH|, and pH  11  effluent, and pH  3, pH0, and pH  11
dilution water, (cf., Section 8.3 and 8.4, respectively
for preparation of pH adjusted effluent and dilution
water.) A flow chart for the  effluent samples of  the
Aeration  test is  shown  in Figure  8-10.  Each sample
is transferred to a 50 mL cylinder containing a small
perfluorocarbon stir bar. The diameter 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 (~10 mL air/minute) for a standard time such
as  60 minutes.  Formation  of precipitates should  be
noted.

The pH of the acidic and basic effluent and dilution
water aliquots is checked every  five minutes during
the first 30 minutes of aeration and every  10 minutes
thereafter.  If the pH of  any solution had drifted more
than 0.2 pH units,  it  must be readjusted back to the
nominal. 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
CI" in each solution can be calculated  following final
pH readjustment.  Solutions should be  stirred during
any pH  readjustment.  Again,  precautions  must  be
taken in order to  minimize the amount of  acid and
base added.  Aeration time does not include the time
intervals  during  which  aeration  is temporarily
discontinued  to readjust pH. A  constant pH is  not
maintained in the "pH|" 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 pH0 dilution water are likely since such
water is usually brought to air equilibrium.

Air contaminated with oil  (droplets or  vapor), natural
gas or  any  other substance  is not acceptable.
                                                 8-18

-------
 Figure 8-10.  Diagram for preparing aeration test samples.
                                     HCI
     Day 1
                      Aerate 35 mL 1 h;
                        Maintain pH
                     NaOH
                                        NaOH
        Aerate 35 mL 1 h;
          Maintain pH
NaOH
Aerate 35 mL 1 h;
   Maintain pH
HCI
HCI
     Day 2
                             'All steps are conducted on dilution water to prepare blanks for testing.
Contaminated 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,  if the room air is
free  of chemicals. Chemistry  laboratories  where
concentrated chemicals are used often  do not have
suitable air quality. Following aeration, the pH of each
solution (including the 35  mL portions  of  pH
unadjusted effluent and dilution water) is returned to
the pH of the initial effluent or dilution water using  the
necessary  volumes  of  0.01  N NaOH and 0.012 N
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 periodically checked  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): Prior to initiating the toxicity tests, the pH of
all of the aerated effluent and blank solutions should
be checked and  adjusted to pH| or pH0. Toxicity tests
are performed on a single  100% concentration of all
three dilution water  blanks  (pH 3, pH0  and pH 11).
These blanks will provide information on toxic artifacts
resulting from aeration.

Based  on  the  24  hour Initial LC50  of  the day 1
effluent, toxicity  tests are performed  on each aerated
effluent solution  at concentrations of 4X- (or 100%),
2X-,  1X-,   and  0.5X-LC50  (cf.,  Section  8.2). The
            pH of each of test concentration should be measured
            and recorded daily. An example of the data sheet for
            the Aeration test is given in Figure 8-11.


            Interferences/Controls and Blanks:

            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. Blank toxicity, especially in all
            three, suggests contaminated  air. Other possible
            causes include contaminated  equipment,  such  as
            electrodes or glassware (especially where low  or high
            pH solutions  were in contact),  or the addition  of too
            much acid or  base.  To determine which  of these
            factors resulted in  blank toxicity,  the  toxicity  of  pH
            adjusted aerated dilution 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  these
            samples of dilution water to make them comparable.
            Another approach to this question  involves observing
            the pH Adjustment/Filtration and pH Adjustment/C^a
            SPE  blanks (Sections 8.4 and 8.6).  Assuming the
            concentration of acid  and base in the  final  blank
                                                 8-19

-------
Figure 8-11.  Example of data sheet for aeration tests.
 Test Type: AERATION
 Test Initiation (Date & Time):
 Investigator:	
 Sample Log #, Name:_
 Date of Collection:	
           pH3
           pH,
           PH11
Species/Age:	
No. Animals/No. Reps:
Source of Animals:	
Dilution Water/Control:,
Test Volume:	
Other Info:	
pH/Conc.
(% effluent)
3/
3/
3/
3/
3/ blank
PH|/
PH|/
PH|/
pH|/
pH| / blank
1V
11/
n/
11/
1 V blank
Survival Readings:
24 h
A pH















48h
A pH















72h
A pH















96h
A pH















                                   HC;
                                                                      NaOH
                                                   8-20

-------
solution  is  approximately  the same  in  all dilution
waters  for  the  three  tests,  toxicity  in  the aeration
blanks  but  not  in  the  filtration or C^Q SPE blanks
suggests that aeration rather than pH adjustment has
led to contamination. Compare the toxicty from the
Baseline test to the  toxicity  of all  three aerated
effluents.  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).

Dissolved oxygen depletion is likely to be caused by
nitrogen stripping.  If a  relatively  large  surface-to-
volume ratio is used (such as the 10  mL volume in a
1  ounce 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|, 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  or  oxidation. 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
process. 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 succeed(s)  in
reducing or removing effluent toxicity, oxidation is a
probable cause.  An  effluent  sample  may contain
toxicants removed  through  sparging and oxidation. An
example would be  where aeration at  pH 3  and pH|
reduces toxicity, but  nitrogen stripping removes the
toxicity only in the pH 3 effluent.

An additional  removal  process,  which is not
volatilization, may   also occur under both  air and
nitrogen sparging.  Materials such as surfactants will
be carried by bubbles and  deposited above the liquid
level on the sides of  the cylinder. If air and nitrogen
both remove toxicity, this zone should be checked for
such deposits.

Removal of toxicants by  precipitation resulting  from
pH change alone in this test 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 sparging has removed the toxicant,
the "volatile toxicant  transfer"  experiment described
below may provide  separation of  the  volatile toxicant
from other  constituents.  Our  experience with this
 Figure 8-12.  Closed loop schematic for volatile chemicals.
                    PUMP
                                      !5__
              •100 mL
              Dilution water
              at 25 °C
                           •1000 mL
                           Effluent
                           at 25 "C
technique is limited to a few effluents. To perform the
"volatile toxicant transfer"  experiment, a closed loop
stripping  apparatus  is  used  (Figure  8-12).  This
apparatus  consists of  a  pump  which can  pump
nitrogen  gas,  two airtight  fluid  reservoirs (cut
graduated cylinders work well),  perfluorocarbon tubing
and diffusers. The arrangement should be that air or
nitrogen can  be circulated  through one reservoir and
then through the second before returning to the first
reservoir.

Numerous operating conditions  can be selected, each
telling something different.  This system should not be
operated as  a conventional purge and  trap system.
The reason is that since one does not know, as yet,
the identity of the toxicant,  conditions for trapping are
not  known.  The objective  initially  should  be to get
measurable toxicity  moved into 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  to the trap. If the  same
concentration of the toxicant can be transferred to the
trap 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 trap volume
small. The nitrogen  sparging gas is recirculated  so
that if the trap is inefficient in  removing the toxicant
from the nitrogen, the toxicant will  not  be lost from
the sample. Because conditions cannot be selected
to optimize transfer,  longer sparging times should  be
chosen.
                                                 8-21

-------
The first experiments should involve no pH changes if
any measurable  change in toxicity  occurred  in  the
earlier tests without pH change. The reason for this
selection is that drastic changes in pH can cause so
many  unknown effluent changes,  and artifacts  are
likely to occur. Of course if only pH changes caused
toxicity changes,  then  pH  will  have to be  altered.
When  pH  is altered,  then  equilibrium objectives,
mentioned  above,  are  not possible  and the entire
process takes on characteristics of more conventional
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.

If sparging  only affects  toxicity with pH  change, then
the system should be operated  in  a  conventional
manner.  That  is,  the  trap  volume should be small
relative to the sample volume  and  the trap pH should
be opposite the sample pH  (e.g., if the sample pH is
3, then the  trap pH should be  10).  One can no longer
conclude  anything  about the  original  effluent
equilibrium  and the procedure is  one of separation.
Toxicity in  the trap may or may not be  the same
substance as the one causing  original effluent toxicity.
Obviously,   all the precautions  mentioned  above
regarding NaCI addition 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.

8.6 pH Adjustment/CIS Solid Phase
     Extraction  Test

Principles/General Discussion:
The solid phase  extraction (SPE)  test is designed to
determine  the extent  of effluent  toxicity caused by
those organic compounds and metal chelates that are
relatively non-polar. The effluent is passed  through a
small column packed with an octadecyl (Cis) sorbent.
Compounds in the effluent interact through solubility
and polarity with the sorbent  and  are extracted from
the  effluent  onto  the  sorbent.  This   type  of
chromatography in which the mobile  phase  (the
effluent)  is  polar and the solid phase (C18 sorbent) is
non-polar,  is referred to as  reversed  phase, SPE.
Any organic compound  present  in water  can be
considered "soluble" by virtue of  its  presence in the
water.  Obviously  relative degrees  of water solubility
exist.  Many highly toxic pollutants found in effluents at
very low concentrations  are not  considered water
soluble despite the  fact they  are  present  at toxic
concentrations.

Compounds extracted  by the C18  sorbent  from a
neutral aqueous solution are usually soluble in hexane
or chloroform. The C18 sorbent can  also be used to
extract organic acids and bases as organic acids and
bases  can  be  made  less polar by  shifting the
equilibrium to  the  un-ionized species.  By  adjusting
the effluent to a low pH and a high pH, some of these
compounds will exist predominately in the un-ionized
form and  will sorb  to the C18  column. Because  of
C18  column degradation, the use of pHs above  10
and  below  2 are  not used.  To ensure column
integrity, pH of the  effluent will  be lowered  only to 3
and  raised  only   to  9  (not  11)  in  this  test.
Manufacturer's data should  be consulted for tolerable
column pH ranges  and for exact column conditioning
proceedures  which must  be  done  to get  proper
performance.


Volume Required:
A maximum volume of  35 ml_ of solution at each pH
is  required  for the pH adjustment/filtration toxicity
testing, and the additional 200 ml_ volume is pumped
through the C\Q column.  The  remaining  200  mL
volumes of filtered pH 3, pH 11, and pH| effluent from
Section 8.4 are used in this test. Each toxicity test is
conducted on four  exposure concentrations (10  mL
each). The exact volume required for the toxicity test
on each pH adjusted or pH| post-column effluent will
depend on the toxicity of the  effluent  (the 24 hour
Initial LC50).


Apparatus:

Six-250   ml  graduated  cylinders,  8-25  mL
graduated  cylinders,  burettes for  acid and base
titrations,  pH  meter  and   probe,   stir  plate,
perfluorocarbon stir  bars, pump with sample  reservoir,
perfluorocarbon tubing, ring stands,  clamps, 3-3 mL
Cis  SPE  columns (200  mg  sorbent), automatic
pipette (10 mL),  disposable pipette tips (10  mL), eye
dropper or wide  bore  pipette, light  box  and/or
microscope (optional).


Reagents:
HPLC  grade methanol, high purity  water, 0.01  N
NaOH, 0.012 N HCI (ACS grade in high purity water),
buffers for  pH meter  calibration, and solvents  for
cleaning the pump and  reservoir.


Test Organisms:

Test organisms, 135 or more,  of the same age and
species.


Procedure (Day 1):
The  first  step  (Figure  8-13)  in the  test involves
conditioning the solid phase extraction  column.
Column conditioning procedures  may  vary with  the
manufacturer  of   the  column.  The  procedures
described  in this document are modifications of the
                                                8-22

-------
Figure 8-13.  Step-wise diagram for C18 solid phase extraction column test samples.
Step
1
Step
2
Step
3


HCI ->,£
Prepare C18 SPE Columns (3)
• 25 mL methanol
• 25 mL high purity water
DO NOT LET SORBENT GO DRY


200 mL pH 3
Filtered Water
+
Prepared
Column
-*•
Collect 1 0 mL Sample
L — NaOH
	 _ _J. 	 .
Toxicity
Test

£~ HCI
200 mL pH 3
Filtered Water
•*•
Prepared Column
from Step 2*
±
Collect 30 mL Samples
After 25 mL & 150 mL
>

	 L DISCS
r Wate

Dilution Water
I I
4r
I

200 mL pH0
Filtered Water
+
Prepared Column
•*•
Collect 1 0 mL Sample
I
	 *


Toxicity Test
Effluent Sample
|
^
I

200 mL pH,
Filtered Water
4-

rd Methanol &
r After Rinses



•J
^ — NaOH
200 mL pH 9
Filtered Water
+
Prepared Column
•*•
Collect
10 mL Sample

i
4 — HCI
r
Toxicity Test

NaOH ^l
200 mL pH 9
Filtered Water
•*•
Prepared Column
from Step 2"
.^
Collect 30 mL Samples
After 25 mL& 150 mL
Prepared Column
from Step 2*
^
Collect 30 mL Samples
After 25 mL & 150 mL
4 — NaOH _M HCI
Toxicity Test
Toxicity Test
Toxicity Test
                                                                                                    Day
                                                                                                     1
                                                                                                   Day
                                                                                                    2 .
                                                                                                   Day
                                                                                                    1
                                                                                                   Day
                                                                                                    2
                                       "Use same column used with the dilution water.
conditioning  steps  used  with  Baker® 1 C-\Q  SPE
columns.

Using  a flow-rate of 10  mL/min, 25 mL of HPLC
grade  methanol is pumped through  the  column and
discarded.  Next 25  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
  1  J.T. Baker Chemical Company, Phillipsburg, NJ.
be prevented from going dry following the methanol'
wash.  The  amount  of  time  needed  between
introduction of solutions to prevent any column drying
will be unique to each investigator's apparatus.  This
timing  should  be determined before  performing this
procedure with actual effluent samples. If the column
dries at any time after introduction of the  methanol,
the column must be reconditioned (with methanol).

As  the last volume of pH 3  high purity water  is
entering the column,  the filtered pH 3 dilution water is
placed into the reservoir.  Again, the column must not
be  allowed to  dry  before  the pH 3 dilution water
enters  the column.  The  pH  3  high  purity water
                                                 8-23

-------
passing from  the  column  should  be measured to
determine the point at which the dilution water begins
to leave the column. The pH 3 high purity water used
to condition the column is discarded. The full 200 ml_
of the filtered pH 3 dilution water is collected,  and a
10 mL aliquot is taken for the toxicity blank to detect
toxicity leached from  the column. This 10 mL aliquot
is readjusted to the  initial  pH of the dilution  water
using 0.01 N NaOH  and 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.  Again, the  column  sorbent must not  be
allowed to dry between the dilution water blank  and
the  effluent.  Collect  a 30 mL  aliquot  of column
effluent after 25 mL of the wastewater passes through
the  system. A second 30 mL aliquot is collected after
a total of  150  mL of  the effluent passes through the
column. Collection of the  first post-column effluent
sample after 25 mL of effluent has passed the column
ensures that any dilution water left in the system will
not  be present  in  the  post-column effluent sample.
The second sample of post-column effluent provides
information on  column  overloading and  toxicant
breakthrough.  Both of  these 30  mL aliquots  are
readjusted to the initial  pH of the  effluent using the
drop-wise  addition of  0.01  N  NaOH.  The total
volume of NaOH necessary for pH adjustment should
be  recorded.  These aliquots are reserved for  day 2
toxicity testing. Columns are not re-used but should
be saved for subsequent elution.
If  a dilution  water  is  used,  for  example receiving
water,  in which  trace organic contaminants  may  be
present or in  which organics such as humic acid may
be present, the same column  should not be used  for
the  effluent.  Rather,  a  new  column should  be
conditioned and in place of the 200 mL dilution water
blank, 200 mL of a synthetic dilution  water should  be
used. It should be checked for toxicity in the same
way. The pH  should be adjusted to the same value as
that of the effluent sample.
The above procedure is repeated  using a  new
conditioned 3  mL C-\Q SPE column  and the filtered
pH| effluent (Figure 8-13). Prior to attaching a new
column to the apparatus, the  reservoir and  pump
must be cleaned with  acetone and high purity water.
After the samples are started through the column, the
pH of the aliquots of  dilution  water and effluent
collected should be checked. It is unlikely  however,
that pH readjustment  to pH0 and pH| (respectively)
will be  necessary. If pH adjustment is necessary it
should be performed  using 0.012 N HCI or 0.01  N
NaOH, recording all volumes added.
In the final CIB SPE test, pH 9 (not 11) dilution water
and  effluent  are  chromatographed  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 un-
ionized form, and therefore removal, the C-\Q column
will not withstand a pH above 10. For this reason, the
pH  11 filtered dilution water  and  effluent  aliquots
prepared in Sections 8.3 and 8.4 are readjusted to pH
9 with 0.12 N HCI and 0.012 N HCI before application
to the column. The 25 mL of high purity water used to
rinse the column following methanol conditioning  must
also  be adjusted to pH 9 with NaOH. The  single 10
mL aliquot of post-column pH  9 dilution water and
both 30  mL aliquots  of  post-column pH 9 effluent
are further adjusted to pH0 and pH| 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 if pH has  drifted over night. Toxicity  tests
are performed on a single 100% concentration  of all
three  of  the dilution water  blanks  collected. These
blanks will  provide  information  on  the  presence of
toxicity leached from the Cia column at different  pHs.

The  six  30 mL post-column  effluent aliquots are
tested for toxicity using an exposure series  based on
the  24  hour LC50  of  the  original  effluent.
Chromatographed  effluent  aliquots  are  tested at
concentrations of 4X-,  2X-,  1X-, and  0.5X-LC50
(cf., Section 8.2 for alternate series). The pH of  each
solution  tested should  be  measured daily   and
recorded along with organism survival. A sample data
sheet for the GTS SPE test is shown in Figure 8-14.


Interferences/Controls and Blanks:
Controls on test organism performance, dilution water
quality, and such are provided by the control from the
Baseline toxicity test.   The pH  Adjustment   and
Filtration tests  (Sections  8.3 and 8.4)  provide
information on the  effects  of  pH  adjustment  and
filtration on effluent toxicity apart from any  additional
changes  caused  by CIQ SPE.  Effluent and  blank
results from these two tests must be consulted  prior
to interpreting the results  of the Cis SPE test, both in
terms of identifying any  toxic artifacts added during
filtration and pH adjustment  and in  allocating toxicity
reduction  to  the three components  potentially
impacting effluent toxicants in the Cis SPE test.
The  Cis  technique  is,  of  those methods  so far
discussed,  the  most  dependent  on   more
manipulations.  More  problems  are  likely  to be
encountered with toxic blanks because in addition to
those associated with pH adjustment and filtering, the
                                                8-24

-------
                        Figure 8-14.   Example of data sheet for effluent solid phase extraction test with and without pH adjustment.
CD
ro
en
Test Type: Cia SPE
Test Initiation (Date & Time):
Investiga
Sample
Date of

%Etf.
cone.









tor:
Log #, Name:
Collection:


PH
3
3
3
3
3
3
3
3
3




vor
25
25
25
25
150
150
150
150
blank



Species//*
No. Anirrtc
Source of
Dilution to
Test Volu
ae:
Jls/No. R(
Animals:
teter/Con
me:
30S.

rol:



Other Information:








Survival Readings:
24 h
A pH









481)
A pH









72 h
A pH









96 h
A pH





























PH
PH|
PH,
PH,
PH|
pH,
PH|
PH|
PH|
PH|
vor
25
25
25
25
150
150
150
150
blank
24 h
A pH









48 h
A pH









72 h
A pH









96 h
A pH





























PH
11
11
11
11
11
11
11
11
11
vor
25
25
25
25
150
150
150
150
blank
24 h
A pH









48 h
A pH









72 h
A pH









96 h
A pH









                                                HCI
                                                                       NaOH
                                                                                                                HCI
                                                                                                                                       NaOH
                             25 ml pH 3
                             25 mL pH|
                             25 mL pH 9
                                "Volume through column.
150mLpH3
150 mL pH(
150mLpH 9

-------
     method  involves use of resin and methanol as
well. Blanks for toxicity must be checked in the same
manner as before  for acid  and  base addition, filter
toxicity, pH drift, as well as  toxicity  from  the  CIQ
column.  But in addition  to these,  some  effluents
behave in  a  peculiar way after passing through the
SPE  column.   They become  toxic  sometimes
apparently  due to "slime" growth or precipitate  and
other times for  no apparent  reason. Observations and
judgement must be  used to detect such problems and
only through experience  can  one  recognize  these
when they occur. Failure to  recognize them will result
in the conclusion that the C^Q column did not remove
toxicity when it  in fact may have.


Results/Subsequent Tests:
The above unique  properties of  some effluents  and
the  potential for  blank  toxicity  problems  make
interpretation of results  more subjective. If toxicity  is
not reduced  in  post-column effluent,  not too  much
credence  should be placed on  the  results. Going
back and sorting through the possible causes can be
very time consuming. If none  of the  other  Phase  I
treatments  have affected toxicity, a wise choice is to
elute the columns  with 100%  methanol. If a  2 ml_
volume of  methanol  is used,  and the sorption  and
elution efficiency are 100%, any  substances  retained
by the  columns will  be concentrated  100 times.  If 150
pL  of  the  methanol fraction is diluted to  10 ml_  in
dilution  water,  the  methanol concentration  is  1.5%
and below the LC50 for all species given in Table 8-
3.  This provides  a  concentration of  effluent
constituents  1.5X whole effluent concentration. This
small amount  of concentration over  whole effluent
allows  detection even if some loss occurred either  in
sorption or elution.  One should run a methanol blank
for  comparison  to the fraction test.

If toxicity is found in the fraction,  proceed to Phase II
for  identification of that component of toxicity.

If  the  post-column effluent  is  not toxic,   then
considerable  weight can be attached  to the results.
The next step is to  go on to Phase II CIQ procedures
for  that toxicity removed  by the column. Some
checking should be done  to  assure that  the  GIB
column has not just served as a smaller pore  size
filter than the  glass  filter used to  filter the sample
before passing  it through the column.  This caution  is
especially applicable to the samples  in which pH  was
adjusted.


8.7 Oxidant Reduction Test

Principles/General Discussion:
This test  is  designed to determine to  what extent
constituents  reduced  by  sodium  thiosulfate  are
responsible for  effluent toxicity.  Chlorine, a commonly
used biocide and  oxidant,  is frequently found at
Table 8-3.   Toxicity ol Methanol to Several Freshwater
           Species.
Species
Ceriodaphnia
dubia




Daphnia
magna
Daphnia pulex

Hyalella
azteca
Sa/mo
gairdneri
Pimephales
promelas


Lepomis
macrochirus
Life-
stage
<6 ha

<24 ha

<48ha

<24 hb

<14hd

juvenile0'

juvenile'

<24 ha

28-32 df

juvenile'

LC50 (%, v/v (95% Cl))
24 h
>3.0
(--)
2.7
(2.6-2.9)
2.4
(2.2-2.6)
NR

2.56
(2.3-2.8)
2.5e
(1.9-2.8)
2.5
(2.5-2.7)
4.0
(-)
3.8
(3.7-3.9)
2.4
(2.2-2.7)
48 h
>3.0
(--)
2.7
(2.6-2.9)
2.0
(1.9-2.2)
3.2C
(2.5-3.7)
NR

NR

2.5
(2.5-2.7)
4.0
(")
3.8
(3.7-3.9)
2.4
(2.2-2.7)
72 h 96 h










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)
a Data for C. dubia and fathead minnows were generated at ERL-
 Duluth. C. dubia were fed. Dilution water used was diluted mineral
 water.
 Randall and Knopp, 1980. (Tested in spring water; static &
 unmeasured.)
0 48 h EC50.
d Bowman et al., 1981. (Tested in well-water, static &
 unmeasured.)
6 18HLC50.
1 Poirier et al., 1986. (Tested in Lake Superior water, and all tests
 were measured and flow-throughs.)
 Note: (--) Confidence interval cannot be calculated as no partial
      mortality occurred.   NR = Not reported.

acutely  toxic concentrations in  municipal effluents.
Other chemicals used in disinfection (such as ozone,
and  chlorine dioxide),  formed  during chlorination,
(such as  mono and dichloramines), and compounds
such as bromine, iodine,  manganous ions  and some
electrophile organic chemicals are also neutralized in
this analysis.

This test  does  not simply  affect chlorine  toxicity.
Chlorine  is  unstable  in  aqueous  solutions   and
decomposition  is  more  rapid  in solutions  when
chlorine is present at  low concentrations. Phase  I
Initial and  pH 3 Aeration tests will provide information
on  chlorine toxicity as will  the Oxidant  Reduction
Test.

In  this test, varying quantities of sodium  thiosulfate
(Na2S2O3), are added  to aliquots of  the  effluent  to
produce  increasing ratios of  reducing  agent/total
thiosulfate reducible  constituents.  Frequently,  the
reduced  form  of  the  toxicant  has  a much lower
toxicity.

Data available  for  Ceriodaphnia dubia,  Daphnia
magna and fathead minnows (Table 8-4)  show  that
sodium  thiosulfate  has a low  toxicity. Other  data
                                                 8-26

-------
generated  at  ERL-D  shows that  for  Ceriodaphnia,
both feeding and reduced hardness lowers the LC50
value. The LC50 for Ceriodaphnia ranges from 300 to
3,300 mg/L in reconstituted  water of 10 to 300 mg/L
(as CaCOs) hardness in fed  tests. Even at the lowest
value,  300 mg/L is nearly  equal  to  2  M  sodium
thiosulfate.  Some  of the  added  thiosulfate  will
combine with certain oxidants present in effluent, thus
lowering  the  concentration  of  reactive,  toxic
thiosulfate. The LC50 values  shown  are therefore
lower than might be expected in effluents  because in
reconstituted  water,  the thiosulfate  is  not  likely to
react.  More  importantly,   since  an effluent
concentration of 4X the LC50 is tested, toxic oxidant
levels should not be excessively high, as might be the
case if 100% effluent were always tested.  As a result
there should not be a need to add very large amounts
of thiosulfate to neutralize toxic oxidants  in the  test
solution.

 Table 8-4.   Toxicity of Sodium Thiosulfate to
           Ceriodaphnia, Daphnia, and Fathead
           Minnows.

                       LC50 (g/L) (95% Cl)
Species
Ceriodaphnia
dubia3
Daphnia
magnab
Pimephales
promelasa
24 h
2.5
(--)
2.2
(NR)
8.4
(7.6-9.3)
48 h
0.85
(0.72-1.0)
1.3
(NR)
8.4
(7.6-9.3)
72 h




79
(7.4-8.5)
96 h




7.3
(6.4-8.3)
 a Data for C. dubia and fathead minnows were generated at
  ERL-Duluth. Both species were < 24 h old at test initiation and
  C. dubia were fed. Dilution water used was diluted mineral
  water.
 b Dowden and Bennett, 1965; dilution water was an artificial water.
 Note:  (--) Confidence interval cannot be calculated as no partial
      mortality occurred.   (NR) Not reported.
The  approach  is  to  add  a  concentration  (at  the
highest  addition) that is approximately equal to  the
^28203  LC50 of the test species. If the test species
is  not listed  in Table  8-4,  an N32S203  LC50  will
have to  be determined.

For cases where oxidants account for only part of the
toxicity,  sodium thiosulfate  may  only reduce  the
toxicity as opposed to completely eliminating it. Time
to  mortality must be  measured  rather than observing
mortality  at  a fixed   time.  Time  to   mortality
measurements  are necessitated as there  are  no
dilutions from  which to  calculate an LC50. In other
sections of Phase  I,  except for Section 8.7 and 8.8,
we fix the time at which mortality is measured and
vary the exposure concentrations. The endpoint is the
estimated concentration  that will kill  half  the
organisms in a specified time.  In  time to  mortality
tests,  the concentration  is fixed  and  the time required
to  kill  50% of the organisms is  measured. Obviously,
to  do so requires frequent  mortality  observation.
Since toxicity is a function of time and concentration,
both LCSOs and LTSOs (lethal time to 50%  mortality)
are measures of the same property.

This test is useful even when chlorine appears to be
absent in the effluent.  Oxidants other  than chlorine
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.


Volume Required:
A maximum volume of 100 mL effluent is required for
the  Oxidant Reduction  Test.  The  exact  volume
required  will depend on the 24 hour Initial LC50.  The
test  requires  10  effluent aliquots at 4X-LC50 or
100%.


Apparatus:
Glass stirring  rods,  1  mL glass pipettes,  automatic
pipette (10 mL),  disposable pipette tips (10 mL), 10
and 100 microliter syringes, eye dropper or wide bore
pipette, light box and/or microscope (optional).


Reagents:
The sodium thiosulfate stock  concentration  should be
10X  the  N32S2O3 LC50 concentration for the  test
species being  used.  The  stock  solution  should  be
prepared in freshly boiled water.


Test Organisms:
Test organisms, 50 or more, of the  same age  and
species.


Procedure (Day 2):
To perform  the test,  put 10 mL aliquots of effluent
diluted to 4X-LC50 (or  100%) in  10  test chambers.
Add  1.0,  0.8,  0.6, 0.4,  0.2,  0.1,  0.05, 0.025,  and
0.012 mL to nine aliquots (mix), and do not add  any
to  the tenth. The treatment  receiving 1 mL  should
contain  the approximate concentration 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 observing time to
mortality is shown on the data form.

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.  By
comparing the  time to  mortality in  the various
Na2S2Os  exposure  concentrations with that  in  the
treatment without thiosulfate, one can  determine
whether the addition of N32S2O3 increased time to
mortality at some thiosulfate concentration. If, in all of
                                                 8-27

-------
Figure 8-15.  Example of data sheet for the oxidant reduction test.
  Test Type:  OXIDANT REDUCTION




  Test Initiation (Date & Time):
  Investigatory
  Sample Log #, Name:_




  Date of Collection:
                               Species/Age: _
                               No. Animals/No. Reps:




                               Source of Animals:	




                               Dilution Water/Control:,




                               Test Volume:	




                               Other Info:      	
            Stock = 	




            Comments:
                                                                4X-LC50:_




                                                                TRC:	
                                                    or 100%
mL Stock
added
1.0
0.8
0.6
0.4
0.2
0.1
0.05
0.025
0.012
0.0
Survival Readings:
2h










4 h










8h










10 h










24 h










48 h










72 h










96 h










g/L Na2S2O3
                                                   8-28

-------
the effluent exposures, time to mortality decreases as
the volume of sodium thiosulfate added increases the
test  should  be repeated  with  a  10  times weaker
Na2S2Os  solution.  If a significant  loss  in effluent
toxicity is apparent over the first 24  hour  period after
sample  arrival in  the laboratory (i.e.,  Initial LC50 <
Baseline LC50),  it  may be  necessary  to conduct
future Oxidant Reduction tests immediately upon the
sample's arrival in the laboratory.


Results/Subsequent Tests:

If  oxidants are causing toxicity,  time to mortality
should increase somewhere  in the range of tested
thiosulfate additions. 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.

If oxidant toxicity  is evident, a  measurement  of free
chlorine  should  be made  and  the concentration
compared to the  chlorine  toxicity value  for the  test
species used. Mono and dichloramine should also be
measured since they have  different toxicities than free
chlorine (see Phase III for  confirming mixtures  as
toxicants). A comparison  of Aeration  and  Cis  SPE
Test results  to the Oxidant  Reduction Test  results
may  provide  even  more  information on   the
physical/chemical  nature of the oxidants.

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) are added to aliquots ot
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  may be more important in
aquatic toxicity than  the total quantity of the metal.

Addition of  the ethylenediaminetetraacetate  ligand
(EDTA),  a  strong  chelating  agent, will produce
relatively non-toxic  complexes 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, strontium
and  zinc.  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 frequently
reduced.

Since EDTA will  complex relatively non-toxic metals
(e.g., calcium, magnesium)  as  well  as more  toxic
heavy metals, fairly concentrated  solutions  are
needed. The mass of chelating agent required should
be approximated  because  EDTA can  become  toxic
when present above a  certain  concentration.  The
range of  EDTA concentrations  that will  adequately
bind the metals but not be toxic appears to be smaller
than that for sodium thiosulfate and oxidants.

Table 8-5 contains  LC50s  for  disodium  EDTA at
various  hardness  and salinity values to  Ceriodaphnia
and fathead minnows. Note that  the concentration of
EDTA tolerated by organisms increases directly  with
both water hardness  and salinity.  By  measuring the
hardness  and salinity 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  prediction  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  does
sodium chloride. The usual methods for measurement
of salinity  (conductivity  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 only 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  toxicity
test water is likely to be much lower  than  the LC50
for  EDTA in effluent.  For  example,  the  values
contained in Table 8-5  represent  worst  case
conditions presented by  EDTA in relatively  pure
water. Likewise,  the  toxic concentration of EDTA in
one effluent  will 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 effluent
aliquots  should bracket the  expected LC50 based on
clean water with a similar salinity and hardness value
as per the above discussion.

As in the pH Adjustment and Oxidant Reduction tests,
the effluent  itself  is used as  a control rather than a
blank based on dilution  water. By  adding increasing
concentrations of EDTA to each effluent aliquot, at
                                                8-29

-------
    Table 8-5.    Toxicity of Disodium EDTA to Ceriodaphnia dubia and Fathead Minnows in Water of Various
               Hardnesses and Salinities.
Species
Ceriodaphnia
dubia








Pimephales
promelas








Water Type
Very Hard
Hard
Mod. Hard
Soft
Very Soft
Soft
Soft
Soft
Soft
Very Hard
Hard
Mod. Hard
Son
Very Soft
Sort
Soft
Soft
Soft
Water Hardness Salmitya
(mg/L as CaCO3) (ppt)
280-320
160-180
80-100
40-48
10-13
40-48 3
40-48 2
40-48 1
40-48 0.5
280-320
160-180
80-100
40-48
10-13
40-48 3
40-48 2
40-48 1
40-48 0.5
LC50 (g/L) (95% Cl)
24 h
0.71
(.S8-.87)
0.50
(.42-.60)
0.23
(.21 -.27)
0.12
(.10-.13)
0.04
(.03-.04)
0.44
0.33
(.27-.41)
0.12
(.10-.13)
0.05
0.81
(.6S-.97)
0.54
(.43-.66)
0.29
(.23-.3S)
0.14
(.12-.18)
0.04
(.03-.04)




48 h
0.41
(.36-.47)
0.44
0.22
0.11
0.03
0.32
(.23-.4S)
0.23
(.21-27)
0.11
0.05
0.81
( 68-. 97)
0.50
(.40-. 62)
0.27
(.22- 33)
0.14
(.12-.18)
0.03
(.03-.04)




72 h 96 h









0.81 0.81
(.6S-.97) (.S2-.83)
0.47 0.44
(.36- 60) (.34:56)
0.27 0.25
C22-.33) (.20-.31)
0.11 0.08
(.08- 14) (.07-.09)
0.03 0.03
(.03-.04) ( 02-.04)
0.37
(.2S-.48)
0.23
(.17-.32)
0.17
(-13-.21)
0.11
    a Brine from evaporated seawater used as source of salinity. All data generated at ERL-Duluth using synthetic waters described in
     Horning and Weber, 1985. All C. dubia were <24 h old and the fathead minnows were all  <36 h old at test initiation.
     Ceriodaphnia were fed during exposure.
    (--)  Confidence interval cannot be calculated as no partial mortality occurred.
some  addition  the metals  will  be  chelated  but
unbound  EDTA  will  not  be  present at  toxic
concentrations. The goal of this test is to add enough
EDTA  to  reduce metal toxicity. This  is most  easily
achieved  by  adding insufficient  EDTA  in the lowest
EDTA  additions so metal toxicity is  not removed.  In
the midrange of 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 capability  of the added  EDTA  is lessened,
especially for very toxic effluents {LC50 < 10%).
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 hour  Initial LC50  and  the
particular  option chosen  to determine  the EDTA
concentration series required.


Apparatus;
Glass  stirring  rods, burettes  for  EDTA  addition,
automatic pipette (10 mL), disposable pipette tips (10
mL), 10  and 100 pL syringe, eye dropper  or  wide
bore pipette, light box and/or microscope (optional).
                                                 8-30

-------
Reagents:
EDTA stock  solution  (see  discussion  under
"Procedure"), reagents for determination of effluent
hardness and salinity (see APHA, 1980; Methods 314
and 210).


Test Organisms:
Test organisms, 50 or more,  of the same age and
species.


Procedure (Day 1):
There are four options  to determine the concentration
of EDTA to add. The most accurate approach, when
it can be used, is to  measure  the hardness of the
4X-LC50 effluent concentration  (or  100% when  the
LC50 is  >25%)  using  the  standard  method  for
measuring hardness (APHA, 1980). The concentration
of EDTA that produced the endpoint for  the effluent
sample is the concentration of EDTA needed at the
midrange of EDTA additions  in  the toxicity test. An
example  will  illustrate the 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). This  amount  corrected  for a  10  ml_ sample
would be 10/100 X 5 mL = 0.5 mL. Therefore 0.5 mL
of 0.01  M EDTA added to a 10 mL  sample of 36%
effluent  provides the EDTA concentration desired at
the midrange. To provide this  EDTA concentration at
the  0.2 mL  addition,  increase  the  0.01  M
concentration by 0.5/0.2 or 2.5X =  0.025 M EDTA
stock. (Molecular weight (MW) of N32EDTA is 372.3
g.) When the hardness measurement endpoint cannot
be discerned because  of interferences, other options,
described below, can be used.

A second option  is  to measure the calcium and
magnesium of the sample, and calculate  the amount
of  EDTA needed  to  chelate  the calcium  and
magnesium.  EDTA binds with  both Ca2+ and Mg2 +
on a  1:1 basis.  The total number of  moles of Ca2 +
(MW = 40.08) and  Mg2+ (MW = 24.305) in 10 mL of
effluent  at 4X-LC50 must equal the number of moles
of EDTA added to the same effluent sample.  This
calculated concentration should be the one added at
the midrange of EDTA additions. The calcium and
magnesium should be  measured at 100% if the LC50
is greater than 25%.

The third option is to  increase the  range of EDTA
concentrations tested rather than attempt to measure
the calcium and magnesium and calculate the needed
EDTA. Our experience is that setting up the toxicity
test with more EDTA  concentrations  may be easier
than  specifically  measuring  the  calcium  and
magnesium.

A fourth  option  is to set the EDTA concentration at
the midrange, equal to the EDTA LC50 concentration.
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 salinity, and hardness and test
organism used.

(Day 2): To perform the  test,  10 aliquots (or more
depending on the option chosen)  of the effluent  are
prepared at a concentration  equal  to 4X-LC50,  or
100% effluent where the  LC50 is greater than 25%
(where the LC50 value is calculated from the 24 hour
Initial toxicity test results). Next, 1.0 mL of the EDTA
stock  is added to  the first 10  mL  aliquot of  the
effluent. To the second 10 mL sample of effluent,  0.8
mL is added,  to the  third,  0.6 mL,  and so on until  the
fifth  10 mL effluent sample has  received  0.2 mL.
Continue the procedure with four more 10 mL aliquots
of the  4X-LC50 (or  100%)  effluent using 0.1, 0.05,
0.025, and  0.012 mL of EDTA stock. The tenth is a
blank  used  to compare treatment  effects on time to
mortality. A microliter syringe will  be needed for  the
smaller  additions. If more  than  1 mL of EDTA is
required, a stronger stock  concentration  of  EDTA
should be  used.  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. Special care must be
taken  to mix  the  contents  of each chamber  before
introducing the  test  organisms. A  sample data
collection sheet is shown  in Figure 8-16.


Interferences/Controls and Blanks:
Controls  prepared  for the  Baseline  Toxicity test
provide  quality  control for  test organisms,  dilution
water and test conditions. The  untreated aliquot acts
as a blank for  use  in  determining the presence  of
EDTA  toxicity.  Time to  mortality  at  each  EDTA
addition is compared to the untreated aliquot.  If time
to mortality  is shorter in all treatments, repeat the test
using  a lower EDTA range.  If time to mortality is  not
reduced in any treatment, repeat using a higher  range
of EDTA concentrations. Erratic patterns in  mortality
cannot be used. If  this occurs  it  suggests  that this
test is not useful  for  the  particular  effluent  being
studied.
Results/Subsequent Tests:
If  the  appropriate  EDTA  concentration  range  is
utilized, the time to mortality will not change from that
seen in the exposure 4X-LC50 of unaltered  effluent
at low additions of EDTA. In the midrange, toxicity will
be reduced and at high additions of EDTA, toxicity will
be as high or higher due to unbound EDTA toxicity
and effluent toxicants other than  chelatable metals if
present. Time to  mortality  must  be  used to detect
partial toxicity removal.
                                               8-31

-------
Figure 8-16. Example of data sheet for the EDTA chelation test.
  Test Type:  EDTA CHELATION




  Test Initiation (Date & Time):
  Investigatory
  Sample Log #, Name:_




  Date of Collection:	
            Stock = 	




            Comments:
                               Species/Age:_
                               No. Animals/No. Reps:




                               Source of Animals:	




                               Dilution Water/Control:




                               Test Volume:	




                               Other Info:	
                                                                4X-LC50:
g/L EDTA
                                                    or 100%
mL Stock
added
1.0
0.8
0.6
0.4
0.2
0.1
0.05
0.025
0.012
0.0
Survival Readings:
2h










4 h










8h










10 h










24 h










48 h










72 h










96 h










                                                    8-32

-------
Toxicity may be  removed at all exposures  if  the
lowest addition  of  EDTA removes metal toxicity  and
the highest addition does not cause EDTA 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 untreated
effluent suggests  EDTA toxicity  and a  lower  EDTA
range should  be tested.

If toxicity is reduced in a systematic  manner, proceed
to Phase II methods for specific identification  of the
metal(s).
8.9  Graduated pH Test

Principles/General Discussion:
This test is designed to determine whether effluent
toxicity can be attributed to ammonia. The test will not
confirm ammonia as the  toxicant  (cf., Phases  II  and
III)  but  will indicate whether its  presence  in  the
effluent should be further  investigated.

Ammonia is specifically addressed in Phase I because
of its frequent  presence in  municipal and  industrial
effluents. Total  ammonia  concentration in an effluent
sample  does  not  relate very  well  to  the toxicity
because the toxicity of ammonia is a function of DO,
pH, and temperature. The pH has a very large effect
on ammonia toxicity and for many effluents, especially
POTW effluents, pH rises upon contact with  air such
as in a toxicity test. Literature data can be used only
as a general guide because of the large effect of very
small pH changes. These are usually not  reported
fully enough to be useful.

One might expect ammonia to be removed during the
pH  11/Aeration test.   Based  on the authors'
experience,  ammonia is  not readily reduced  below
concentrations  toxic to  aquatic  organisms.  Other
techniques  which can be used to remove ammonia-
related toxicity may also displace other toxicants with
completely  different   physical  and  chemical
characteristics.  For example,  ion  exchange resins
(e.g., zeolite) in addition  to removing ammonia, may
also remove  toxic  organic  compounds   through
adsorption.  For  this reason, a specific test to address
toxicity related to this common pollutant is useful.

Ammonia acts  as a basic compound in  water. The
un-ionized,  more  toxic  form,  NHa,  predominates
above pH 9.3 and the ionized, essentially non-toxic
form, NH4 + , is most abundant  below this pH  at
25°C.  Through the  pH range  from  7.0-8.0,  the
percent of total ammonia in  the toxic form increases
rapidly.  It is important to  note that the toxicity of the
toxic form decreases with increasing pH (EPA, 1985).
Therefore,  the  increase in the concentration of  the
toxic form is partially compensated by the decrease in
toxicity.  As  pH increases, the percentage  in the toxic
form is greater but the toxicity  of  the  toxic form is
less.  As  pH  decreases,  the percentage of  total
ammonia as NHs decreases, but the toxicity of NH3
increases.  Temperature also  affects dissociation  of
ammonia  but  since  it is  usually held  constant  in
toxicity  tests,  it  can be  ignored  for  purposes here.
The increase  in  the percent  of  total  ammonia
occurring  in  the  toxic form  with increasing  pH  is
greater  than  the decrease in its toxicity.  The  net
result is  an  increase in  toxicity,  based  on  total
ammonia in the sample.

These opposing effects of pH change can be used to
detect the presence  of ammonia toxicity. By diluting
aliquots of effluent to  the  24 hour  Initial LC50 and
adjusting each aliquot to a different pH value, effluent
toxicity  can  be  greatly  enhanced  or completely
eliminated. For example, at pH 6 and 25°C,  0.0568%
of the total ammonia  in solution is present in the toxic
form. At  pH  7  and  25°C,  0.566%  of the  total
ammonia is present  as NHs  and at pH  8, 5.38% is
present in the un-ionized  form.  Similar changes in
percent NHs  for pHs 6,  7,  and 8  occur  at other
temperatures,  if a  test temperature different  from
25°C is needed. This difference in  the percent  of
un-ionized  ammonia  is enough  to make the  same
amount of total ammonia  about 3 times  more toxic at
pH 8 as  at pH 6. Whether  or  not  toxicity will be
eliminated at pH 6 and the extent  to which toxicity will
increase at pH 8 will depend on the total ammonia
concentration.  But if a dilution  of effluent equal to the
LC50  is  used  for  pH   adjustment,  this  should
"normalize" the total ammonia concentration and one
should see toxicity differences between pH  6  and 8.
Either  pH  8  will have a  higher toxicity than the
unadjusted effluent LC50 or pH  6 will have a lower
toxicity depending on the pH of the unaltered effluent.

Perhaps the greatest challenge faced in  this  test is
that of maintaining a constant solution pH. This is a
necessity  if the concentration of un-ionized ammonia
is  to remain constant and the test  results are  to be
valid.  In conducting toxicity tests on effluent, it  is not
unusual to see the pH of the test solutions with high
effluent  concentration drift 1 to 2  units greater over a
48 to 96 hour period (see Procedure for suggestions).


Volume Required:
Depending on  the method chosen for pH stabilization,
120 mL to 1,500 mL of  effluent  is required for this
test. The volume used is split into three  aliquots and
tested  at  the  24  hour Initial LC50  concentration,
without dilution or replication.

Apparatus:

Burettes  for   acid  and  base addition, 3-50 mL
beakers1,  Parafilm®!  or  3-600  mL  beakers2,  wire
mesh test chambers2 (described below), magnetic
stirrers,  and perfluorocarbon stir bars, eye dropper or
                                                8-33

-------
wide  bore pipette,  light box  and/or microscope
(optional).

Reagents:
1.0 and 0.1  N NaOH2 or 0.1  N and 0.01 N NaOHl,
1.2 N and 0.12 N HCI2 or 0.12 N and 0.012 N HCI1.

Test Organisms:

Test organisms, 15 or more, of the same age and
species.


Procedure (Day 2):
If  the  dissolved oxygen  content  of  the  effluent is
adequate  and BOD is  not high, the following small
volume pH stabilization  procedure should be tried. An
aliquot of wastewater is diluted  to the 24 hour Initial
LC50 concentration using dilution  water  for  a final
volume of 120 ml. A 40 ml_ portion of this solution is
adjusted  to pH  7 using 0.12 and 0.012 N HCI or 0.1
and 0.01  N NaOH.  Caution must be taken to minimize
the volume of acid  or base added to the test solution.
To ensure this, as  the pH nears 7.0, titration with the
stronger  acid or base solution should  be  subrogated
with the  addition  of  the  weaker solution of acid or
base. Approximately half of this  volume is transferred
into a one ounce disposable transparent test chamber
and  five  test organisms  are  added. The remaining
volume is carefully transferred into the test chamber
until the  volume of pH  7 effluent rises slightly above
the edge of  the container. A section of Parafilm®,
slightly larger in  area  than the mouth of the test
chamber, is  laid on top of the effluent. The Parafilm®
is  carefully pulled down around the edges of the test
chamber, caution being  taken  to  prevent  test
organism loss at this time. Care  must also be used to
exclude any  air bubbles during the sealing  process.

This procedure is repeated with the other two 33 mL
aliquots of the  effluent adjusted to pH 6 and pH 8.
Test  organisms   can  be observed  through  the
transparent test chamber. The Parafilm® cover should
only be removed at the end of the test period in order
to measure test solution pH and  dissolved  oxygen.

If the above method does not hold pH or if oxygen is
too low,  the following  large volume  pH  stabilization
procedure may be  tried.

An aliquot of wastewater is  diluted to the  24  hour
initial LC50 concentration for  a final volume of 1.5 L.
A  500 mL portion of this  solution in a 600 ml beaker
is  adjusted to pH 7 using  1.2 N and 0.12 N HCI or 1.0
N  and 0.1   N  NaOH.  Caution  must  be  taken to
minimize the volume of acid or base added to the test
solution.  To  ensure  this,  as  the pH of the solution
  1 Required for the small volume pH stabilization method.
  2 Required for the large volume pH stabilization method
nears 7.0, titration with the stronger acid 01 base
should be subrogated  by  titration  with  the weaker
solutions.  This  procedure is repeated with a second
500 mL aliquot, adjusted to pH 6, and a third adjusted
to pH 8.

A 2 inch  high  stainless steel  mesh  cylinder  (60
mesh/inch) closed  at  one  end  with a  "Petri  dish
type,"  water-tight  bottom is suspended in each of
the  500  mL  test  solutions. Test organisms  are
carefully transferred into these cages.  This cylinder
allows  organism exposure  to  the  static  pH solution
while providing  a  mechanism for  maintaining  the
organisms in a small area  for observation  purposes.
The  cylinders -can be  removed  for test  organism
observation without  incurring  harm to  the animals.
They can  also  be removed  as  necessary for  pH
readjustment.  Solution  pH  should  be checked  (and
readjusted as  necessary) every half hour during the
first few hours and as needed for  the rest of the test.
DO  should also  be  measured  to  be sure it is
adequate.  Records of pH drift and  DO  levels  should
be  kept.  An  example  data collection sheet for  the
Graduated pH test is contained in Figure 8-17.

The "cage"  method  described  above  is  useful
because  the  cages  can  be  placed  under  a
microscope for observation. Remember that the  DO
need is   not  as  high  as  is usually  required in
conventional toxicity tests. If the DO is not at or below
a lethal level for the duration of the test, the DO  can
be  considered  adequate. More importantly, the  DO
should be  comparable in each test vessel.

If pH drifts from nominal more than 0.1 or at most 0.2
pH  units,   the  results  may be unusable.  Short of
flow-through testing other  approaches  have  to be
used.

An alternative is to use higher  effluent concentrations,
perhaps 4X-LC50,  adjusting pH initially and every 10
or 15 minutes  as  needed  to  maintain it at nominal.
Then observe the  animals every  15-30 minutes  and
note onset of  symptoms. Ammonia does not  have  a
long lag  period before  symptoms develop,  so  the
observations  can be done in  a  work day.  Care will
have to be taken  to  avoid  trauma  to  the animals
during pH adjustment.

Interferences/Controls and Blanks:
Blanks prepared for the Baseline test act as a check
on the general health of test organisms,  dilution water
quality and test conditions. The Baseline test acts as
a control  for the detection  of problems  (especially in
the  smalt volume  pH  stabilization method)   and
problems  associated with acid or base addition. If the
Baseline effluent pH is  close to the pH  of any of the
pH  adjusted test  solutions,  the  toxicity  should be
similar. Significantly greater toxicity suggests  that an
interference from  other factors (cf.,  Section  8.3 for
discussion on ionic strength related  toxicity).
                                                8-34

-------
Figure 8-17.  Example of data sheet for the graduated pH test.
Test Type: GRADUATED pH TEST
Test Initiation (Date & Time):
Investigator:
Sample Log #,
Date of Collect

Name:
on:

                                                                 Species/Age:_
                                                                 No. Animals/No. Reps:_




                                                                 Source of Animals:	
                                                                 Dilution Water/Control:,




                                                                 Test Volume:	




                                                                 Other Info:	
                                                                 1X-LC50:
or 100%

pH
6.0
7.0
8.0
Survival Readings:
2h'
A pH DO



4h*
A pH DO



6h"
A pH DO



24 h
A pH DO



48 h
A pH DO



72 h"
A pH DO



96 h'
A pH DO



            Not performed in small volume pH stabilization method.




         Comments:
                                                    8-35

-------
Other compounds with toxicities that increase directly
with pH may lead to confounding results or may give
results  similar to ammonia.  Phase  II  contains a
suggested test to more specifically identify ammonia
as the cause of toxicity.

Results/Subsequent Tests:

When  ammonia is  present in  the effluent at toxic
levels, the pH  6 test solution should be less toxic than
the pH 7 solution, which, in turn, should be less toxic
 than the pH 8 solution. However, ammonia is not the
only possible cause. 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. For example, if, for the  Baseline effluent,
the average pH  in the 100% concentration  in which
no organisms  survived was  8.0 and the  average pi-
was  7.5 for the 50%  concentration in which  al
organisms  survived,  the  estimated  pH at the  LC5C
(71%)  could be approximated as  7.7.  One  would
expect greater than  50% mortality  in the pH  8 test
solution and significantly  less lethality in the  pH  7
solution.  One  should then  proceed to  Phase  II for
identification. If ammonia is one of several toxicants in
an effluent this  procedure may  pose problems.  For
this  reason,  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., Section 9). Methods for further identifying
ammonia as the  toxicant can be found in Phase It.
                                                 8-36

-------
                                             Section 9
                                Time Frame and Additional Tests
9.0  Time Frame for  Phase I Studies

The  amount  of time  necessary to  adequately
characterize the physical/chemical  nature of,  and
variability in, an effluent's toxicants will be discharge
specific. Among the  factors affecting  the length  of
Phase I  studies 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
consistently containing  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
impacted 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 adequate.

There  are  no  clearly defined boundaries between
Phase  I  and  Phase  II.  The "subsequent  tests"
described for the characterization methods may be
thought of as intermediate studies between Phases I
and II.  In terms of guidance on the TIE'S time frame,
several samples should be  subjected to the Phase I
Characterization  Test  battery.  The  decision  to do
subsequent tests on these samples  to  confirm or
further delineate initial results is a judgement call and
will depend on  how clear-cut  the results  of Phase  I
turn  out  to  be. 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 organized  and experienced the performing lab is
at doing TIEs.

If  Phase  I  tests  needed  to  remove  or  neutralize
effluent toxicity vary with  the sample, the number of
tested samples must be increased. The frequency of
testing should  be sufficient  to  include  all  major
variability. While true, this  statement is of no help.
Again,  judgement  will  have  to  be used  but the
differences  seen among  samples  can be  used to
decide when further differences are not being  found.
Phase  I  toxicant characterization  testing  should
continue until there is reasonable certainty 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  is  unique.   The  LC50  of
samples can be very different but the  same screening
tier tests  must  be successful  in  removing  and/or
neutralizing effluent toxicity.

The individual  Phase I tests 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 constituents
accompanying the toxicants.  These efforts may reveal
more  toxicants than suggested by Phase  I testing. In
Phase II  one may  discover  that toxicants of a  quite
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
GIB  SPE,  this  does not prove  the  existence  of  a
single  toxicant.  Several  non-polar  organic
compounds may in fact be  causing  effluent  toxicity
over time, but, the CIQ SPE technique only detects
the presence of these compounds as  a group. This is
very important during Phase II toxicant identification.

9.1  When Phase I Tests are Inadequate:

For some effluents, the Phase I tests described above
will provide little or  no clue as to the characteristics of
the toxicants.  For  such effluents, other  approaches
must be tried.  Some additional approaches are given
below  with much less specificity  because our
experience with them is minimal. In addition to these,
one  should not  hesitate  to  use   originality  and
innovation to  develop other  approaches.  So long  as
toxicity is  used to  track the changes, any approach
may be helpful.

Activated Carbon

Chemists are reluctant to use carbon because it is
much less selective  than   ion  exchange or  SPE
columns  and extraction is  less  precise  and  more
difficult. Carbon's non-selectivity  is an advantage  in
                                                9- 1

-------
some situations. A rather wide array of more specific
methods  have failed if Phase I tests above  have  not
changed  toxicity, and therefore a "chemical sponge"
may be useful. In order to  start, one must be able to
alter toxicity somehow in order  to tell  what changes
are occurring. A second objective in early work is a
way to remove the toxicants  from the sample (i.e., to
concentrate them). Carbon has a high capability to do
both.  Furthermore, the  knowledge about carbon
sorption and extraction is large and help can be found
in the literature.  True,  carbon  may alter  some
chemicals,  but there are many  that it does not.  We
must recognize that other conventional methods such
as ion exchange are also  not specific. Ion exchange
columns  can sorb non-polar  organics  and SPE
columns  can  sorb metals.  Carbon  may  be very
helpful.

Other Specific Ion Columns

Many other types of resin  columns  are  available
through commercial  sources.  Many of  these have
"insurmountable"  blank  toxicity problems  but  some
show promise. Mixed  bed  ion exchange  columns
appear  promising because  pH  is  not  drastically
altered as the sample passes through the resin bed
and blank problems appear tolerable. Of course with
any of these lessser used methods, the  organism's
tolerance must be determined before  any chemicals
can be used.

Other LJgands

EDTA reduces toxicity for only part of  the cationic
metals. Other ligands such as citrate, nitrilotriacetate,
and cysteine may hold promise.

Clustering Phase I Tests

Our experience suggests independent action and less
than  additivity are much  more common than  we
realize, at least in effluents. When these  interactions
occur, interpreting Phase I data may be very difficult
and in some instances, especially with independent
action, no  apparent  effect on toxicity  will be  seen
unless Phase I tests are clustered.

The pH  of effluents plays an amazingly powerful role
in  affecting both form of toxicants and their toxicity.
Including pH adjustments to different values  than
suggested  in  Phase I  may be  helpful.  Combining
EDTA addition  with  post  SPE column sample may
reveal  additional  information.  We  have  used
aeration/filtration/pH adjustment/Ci8  SPE  in  various
combinations to decipher changes  occurring.  The
presence of more than one toxicant may often require
such combinations.

A  special effect occurs when an effluent, having two
toxicants at very  different concentrations, is diluted.
Suppose toxicant A would produce an  LC50 at 50%
effluent  and toxicant B causes an  effluent  LC50 at
5%.  In  most  cases, only toxicant  B  will materially
affect toxicity because the effect of A will be "diluted
out"  long before the LC50 of B is reached. In Phase
I,  tests for such an effluent would be done  near the
LC50 of B (20% and down) so that  the toxicity of A
will not be  noticed. If  one finds toxicity  at effluent
concentrations in  the very low range (such as  10%)
additional  Phase  I testing  at higher  effluent
concentrations  should  be done  subsequently.  Such
cases  should be  caught in  Phase III,  but earlier
detection will be much more  cost-effective.

The  two objectives which usually must be achieved
before the identity of the toxicant can be made is that
the toxicant(s) must be separated and concentrated.
Anything that can  be  done  in  Phase I to achieve
these goals will speed the process.
                                                 9-2

-------
                                           Sect/on 10
                                           fleferences
Adelman, I.R., LL Smith Jr., and G.D.  Siesennop.
  1976.  Acute  Toxicity  of Sodium  Chloride,
  Pentachlorophenol, Guthion®,  and  Hexavalent
  Chromium to  Fathead  Minnows (Pimephales
  promelas)  and Goldfish (Carassius  auratus).J.
  Fish. Res.  Board Can.  33:203-208.
American Public Health Association, American Water
  Works  Association  Water  Pollutin  Control
  Federation. 1980.  Standard  Methods  for the
  Examination  of Water and  Wastewater.  15th
  Edition. American  Public  Health  Association,
  Washington, DC.
Berg, E.  1982. Handbook for  Sampling and Sample
  Presentation of  Water and Wastewater. U.S. EPA,
  Cincinnati, Ohio.  EPA-600/4-82-019.
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.
Department of Health, Education and Welfare.  1977.
  Carcinogens - Working  With Carcinogens. Public
  Health Service, Center for  Disease  Control,
  National  Institute  of Occupational Safety  and
  Health. 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):1308-1316.
EPA. 1985.   Ambient Water  Quality Criteria for
  Ammonia.  EPA-440/5085-001.  Environmental
  Protection Agency, Environmental  Research
  Laboratory-Duluth,  Duluth, MN, and the  Criteria
  and Standards Division,  Washington, DC.
EPA.  1988A. Draft  Toxicity  Reduction  Evaluation
  Protocol  for  Municipal Wastewater  Treatment
  Plants. EPA  Water  Engineering   Research
  Laboratory, Cincinnati,  Ohio. Draft EPA Research
  Series Report.
EPA.  1988B.  Draft  Generalized  Methodology for
  Conducting Industrial  Toxicity Reduction
  Evaluations (TREs).   EPA Water  Engineering
  Research  Laboratory, Cincinnati, Ohio.  Draft EPA
  Research Series Report.
Federal Register.  1984. U.S.  EPA: Development of
  Water Quality Based Permit Limitations for  Toxic
  Pollutants;  National Policy. U.S. EPA, Volume 49,
  No. 48, Friday, March 9, 1984.
Horning,  W.  and C.I. Weber,  (Eds.).  1985.  Short-
   Term Methods for Estimating the Chronic Toxicity
   of Effluents and Receiving  Waters to Freshwater
   Organisms. U.S.  EPA,  Cincinnati,  Ohio.  EPA-
   600/4-85-014.
Magnuson, V.R.,  D.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:635-656.
Mount,  D.I.   1988.  Methods  for  Aquatic  Toxicity
   Identification  Evaluations:  Phase  III  Toxicity
   Confirmation  Procedures. Draft  EPA  Research
   Series Report.  Environmental Research Laboratory,
   Duluth,  MN.EPA/600/3-88/036.
Mount,  D.I.  and  L.  Anderson-Carnahan.  1988.
   Methods  for  Aquatic   Toxicity  Identification
   Evaluations:   Phase  II  Toxicity  Identification
   Procedures. Draft  EPA  Research Series Report.
   Environmental  Research  Laboratory,  Duluth,  MN.
   EPA/600/3-88/035.
Occupational Safety and  Health Administration.  1976.
   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.:137-140.
Peltier, W. and C.I. Weber, (Eds). 1985. Methods for
   Measuring  the  Acute Toxicity of  Effluents to
   Freshwater and  Marine  Organisms. 3rd Edition.
   U.S. EPA,  Cincinnati, Ohio. March  1985.  EPA-
   600/4-85-013.
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.
Stumm,  W.  and  J.J.  Morgan.  1981.  Aquatic
   Chemistry  -  An  Introduction  Emphasizing
   Chemical Equilibria in  Natural Waters. John  Wiley
   & Sons, Inc., New York, NY. 583 p.
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.
Walters, C.I. and C.W. Jameson. 1984. Health and
   Safety  for  Toxicity Testing.  Butterworth  Publ.,
   Woburn, MA.
                                              1 0-1
                                                                 * US GOVERNMENT PRINTING OFFICE 1988- 548-156/87041

-------