United States            Office of Water         EPA 821-R-00-028A
          Environmental Protection        (4303)              October 2000
          Agency
&EPA  Preliminary Report:
         Interlaboratory Variability Study of
         EPA Short-term Chronic and
         Acute Whole Effluent Toxicity Test
         Methods,
         Vol.1
 K
 c

-------
Disclaimer

This preliminary report has been reviewed by the Analytical Methods Staff in the Engineering and Analysis
Division within the USEPA Office of Water and EPA's WET Interlaboratory Variability Workgroup.
Mention of company names, trade names, or commercial products in this report does not constitute
endorsement or recommendation for use.

Questions or comments regarding this report should be addressed to:

William A. Telliard, Director
Analytical Methods Staff
Engineering and Analysis Division (4303)
Office of Science and Technology
U.S. EPA Office of Water
Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, DC 20460

-------
TABLE OF CONTENTS

Acknowledgement	xii

Executive Summary	  xiii

1.0 INTRODUCTION AND BACKGROUND	1
       1.1     Regulatory Background  	 1
       1.2     The WET Study	2
       1.3     Other EPA Documents 	2

2.0 STUDY DESIGN AND OBJECTIVES  	3
       2.1     Objectives	3
       2.2     General Study Design	3
              2.2.1    Study Management	3
              2.2.2    Methods Evaluated	4
              2.2.3    Laboratories	5
              2.2.4    Samples	7
              2.2.5    Schedule  	7

3.0 LABORATORY  PROCUREMENT	9
       3.1     Identification and Solicitation of Potential Referee Laboratories	9
       3.2     Identification and Solicitation of Potential Participant Laboratories	9
       3.3     Prequalification Requirements for Participant Laboratories	10
              3.3.1    Prequalification Documentation 	 10
              3.3.2    Prequalification Determination	 12
       3.4     Prequalification and Selection of Referee Laboratories	 13
       3.5     Selection of Participant Laboratories	14
       3.6     Participant Laboratory Meeting	14

4.0 PRELIMINARY  TESTING	 17
       4.1     Preliminary Testing Plan	 17
              4.1.1    Part 1 - Background Testing	17
              4.1.2    Part 2 - Range-finding  	 19
              4.1.3    Part 3 - Holding Time Testing 	20
              4.1.4    Part 4 - Definitive Testing  	20
       4.2     Preliminary Testing Results  	21
              4.2.1    Preliminary Testing for Ceriodaphnia Acute and Chronic Test Methods	21
                      4.2.1.1 Reference toxicant sample type 	21

                                             iii

-------
                      4.2.1.2 Effluent sample type  	23
                      4.2.1.3 Receiving water sample type  	25
               4.2.2   Preliminary Testing for Fathead Minnow Acute and Chronic Test Methods ... 27
                      4.2.2.1 Reference toxicant sample type  	27
                      4.2.2.2 Effluent sample type	28
                      4.2.2.3 Receiving water sample type  	29
               4.2.3   Preliminary Testing for the Selenastrum Chronic Test Method 	29
                      4.2.3.1 Reference toxicant sample type  	29
                      4.2.3.2 Effluent sample type  	30
                      4.2.3.3 Receiving water sample type  	31
               4.2.4   Preliminary Testing for the Mysid Chronic Test Method	32
                      4.2.4.1 Reference toxicant sample type  	32
                      4.2.4.2 Effluent sample type  	32
                      4.2.4.3 Receiving water sample type  	33
               4.2.5   Preliminary Testing for Sheepshead Minnow Acute and Chronic  Test
                      Methods	34
                      4.2.5.1 Reference toxicant sample type  	34
                      4.2.5.2 Effluent sample type  	35
                      4.2.5.3 Receiving water sample type  	36
               4.2.6   Preliminary Testing for Silverside Minnow Acute and Chronic Test Methods  . 37
                      4.2.6.1 Reference toxicant sample type  	37
                      4.2.6.2 Effluent sample type  	37
                      4.2.6.3 Receiving water sample type  	40
               4.2.7   Preliminary Testing for the Champia Chronic Test Method	41
                      4.2.7.1 Reference toxicant sample type  	42
                      4.2.7.2 Effluent sample type  	43
                      4.2.7.3 Receiving water sample type  	43
               4.2.8   Preliminary Testing for the Holmesimysis Acute Test Method	45
                      4.2.8.1 Effluent sample type  	45
                      4.2.8.2 Receiving water sample type  	46
       4.3     Problems Encountered in Preliminary Testing	47

5.0 SAMPLE PREPARATION  	49
       5.1     Freshwater Methods	49
       5.2     Marine Methods	53
               5.2.1   Mysid Chronic and Sheepshead Acute and Chronic Test Methods	53
               5.2.2   Silverside Acute and Chronic Test Methods  	55
       5.3     Problems Encountered in Sample Preparation 	57
                                               IV

-------
6.0 PACKAGING AND DISTRIBUTION OF TEST SAMPLES	58
       6.1    Sample Distribution Scheme	58
       6.2    Packaging and Shipping of Samples  	58
       6.3    Sample Tracking	60
       6.4    Problems Encountered in Sample Distribution	64

7.0 INTERLABORATORY TESTING  	65
       7.1    General Testing Requirements	65
       7.2    Method-Specific Requirements  	68

8.0 DATA REPORTING AND EVALUATION	81
       8.1    Report Submission	81
       8.2    Data Review	84
              8.2.1   Data Package Receipt and Gross Completeness Check 	84
              8.2.2   Data Accuracy and Quality Check  	84
              8.2.3   Effect Concentration Re-calculation and Verification  	87

9.0 RESULTS  	91
       9.1    Analysis of Results	91
              9.1.1   Valid Tests	91
              9.1.2   Successful Test Completion Rate  	91
              9.1.3   False Positive  Rate	91
              9.1.4   Precision  	92
       9.2    Ceriodaphnia Acute Test Method Results	96
              9.2.1   Successful Test Completion Rate  	96
              9.2.2   False Positive  Rate	96
              9.2.3   Precision  	96
       9.3    Ceriodaphnia Chronic Test Method Results	  102
              9.3.1   Successful Test Completion Rate  .	102
              9.3.2   False Positive  Rate	103
              9.3.3   Precision  	  103
       9.4    Fathead Acute Test Method Results  	  113
              9.4.1   Successful Test Completion Rate  	113
              9.4.2   False Positive  Rate	113
              9.4.3   Precision  	  113
       9.5    Fathead Chronic Test Method Results	  119
              9.5.1   Successful Test Completion Rate  	  119
              9.5.2   False Positive  Rate	  119
              9.5.3   Precision  	  120

-------
       9.6     Selenastrum Chronic Test Method Results 	130
               9.6.1   Successful Test Completion Rate	130
               9.6.2   False Positive Rate	 131
               9.6.3   Precision  	 131
       9.7     Mysid Chronic Test Method Results	143
               9.7.1   Successful Test Completion Rate 	143
               9.7.2   False Positive Rate	 143
               9.7.3   Precision  	 144
       9.8     Sheepshead Acute Test Method Results	 150
               9.8.1   Successful Test Completion Rate 	150
               9.8.2   False Positive Rate	 150
               9.8.3   Precision  	 150
       9.9     Sheepshead Chronic Test Method Results	153
               9.9.1   Successful Test Completion Rate 	153
               9.9.2   False Positive Rate	 153
               9.9.3   Precision  	 154
       9.10    Silverside Acute Test Method Results	160
               9.10.1 Successful Test Completion Rate 	160
               9.10.2 False Positive Rate	160
               9.10.3 Precision  	 160
       9.11    Silverside Chronic Test Method Results  	 165
               9.11.1 Successful Test Completion Rate 	165
               9.11.2 False Positive Rate	165
               9.11.3 Precision  	 165
       9.12    Results Summary	 172
               9.12.1 Successful Test Completion Rate 	172
               9.12.2 False Positive Rate	173
               9.12.3 Precision  	 174

10.0 REFERENCES	 177

VOLUME 2: APPENDIX
       A. WET Study Plan
       B. Participant Laboratory Standard Operating Procedures
       C. List of Participant Laboratories
                                               VI

-------
LIST OF TABLES

Table 1.  Summarized test results from the WET Study	  xiv
Table 2.1. Four phases and specific objectives of the WET Study. .	4
Table 2.2. WET test methods included in the WET Study	5
Table 2.3. Endpoints evaluated for each test method in the WET Study	6
Table 2.4. Final schedule for interlaboratory testing during the WET Study	8
Table 3.1. Summary of referee laboratory solicitation	9
Table 3.2. Referee laboratories selected for the WET Study	14
Table 3.3. Results of participant laboratory prequalification and selection	16
Table 4.1. Summary of preliminary testing requirements	 18
Table 4.2. Results from Ceriodaphnia acute preliminary testing	22
Table 4.3. Results from Ceriodaphnia chronic preliminary testing	22
Table 4.4. Water chemistry of effluent sample source for freshwater test methods	24
Table 4.5. Water chemistry of receiving water sample source for freshwater test methods	26
Table 4.6. Results from fathead acute preliminary testing	27
Table 4.7. Results from fathead chronic preliminary testing	28
Table 4.8. Results from Selenastrum chronic preliminary testing	31
Table 4.9. Results from mysid chronic preliminary testing	33
Table 4.10. Water chemistry of receiving water sample source  for mysid chronic and sheepshead
            acute and chronic test methods	34
Table 4.11. Results from sheepshead acute preliminary testing	35
Table 4.12. Results from sheepshead chronic preliminary testing	36
Table 4.13. Results from silverside acute preliminary testing	38
Table 4.14. Results from silverside chronic preliminary testing	38
Table 4.15. Water chemistry of effluent sample source for silverside acute  and chronic test methods.  .. 39
Table 4.16. Water chemistry of the receiving water sample source for silverside acute and chronic test
            methods	41
Table 4.17. Results  from Champia chronic preliminary testing	42
Table 4.18. Water chemistry of the effluent sample source for the Champia chronic test method	43
Table 4.19. Water chemistry of the receiving water sample source for the Champia chronic test
            method	44
Table 4.20. Results from Holmesimysis acute preliminary testing	45
Table 4.21. Water chemistry of the effluent sample sources for the Holmesimysis acute test method.  .. 46
Table 4.22. Water chemistry of the receiving water sample source for the Holmesimysis acute test
             method	47
Table 5.1. Description of samples used for freshwater methods in the WET Study	50
Table 5.2. Spiking concentrations in reference toxicant ampules	51
Table 5.3. Effluent sample volumes collected for interlaboratory testing	52

                                               vii

-------
Table 5.4. Receiving water sample volumes collected for interlaboratory testing	52
Table 5.5. Description of samples used for marine methods in the WET Study	54
Table 6.1. Sample distribution scheme for the WET Study	59
Table 6.2. Number and volume of samples required for freshwater methods in the WET Study	61
Table 6.3. Number and volume of samples required for marine methods in the WET Study	62
Table 6.4. Episode numbers and sample numbers used in the WET Study	63
Table 7.1. Summary of test conditions and test acceptability criteria for the Ceriodaphnia acute test
            method	69
Table 7.2. Summary of test conditions and test acceptability criteria for the Ceriodaphnia chronic
            test method	70
Table 7.3. Summary of test conditions and test acceptability criteria for the fathead acute test
            method	71
Table 7.4. Summary of test conditions and test acceptability criteria for the fathead chronic test
            method	72
Table 7.5. Summary of test conditions and test acceptability criteria for the Selenastrum chronic test
            method	73
Table 7.6. Summary of test conditions and test acceptability criteria for the mysid chronic test
            method	74
Table 7.7. Summary of test conditions and test acceptability criteria for the sheepshead acute test
            method	75
Table 7.8. Summary of test conditions and test acceptability criteria for the sheepshead chronic test
            method	76
Table 7.9. Summary of test conditions and test acceptability criteria for the silverside acute test
            method	77
Table 7.10.  Summary of test conditions and test acceptability criteria for the silverside  chronic test
            method	78
Table 7.11.  Summary of test conditions and test acceptability criteria for the Champia  chronic test
            method	79
Table 7.12.  Summary of test conditions and test acceptability criteria for the Holmesimysis acute test
            method	80
Table 8.1. Report due dates	82
Table 8.2. Data reporting elements	82
Table 8.3. Test data qualifier flags	85
Table 8.4. Sample results affected by EPA guidance on concentration-response relationships
            (USEPA, 2000a)	88
Table 9.1. Test results identified as potential outliers by ASTM h statistics	95
Table 9.2. Results for Ceriodaphnia acute test method performed on blank samples	98
Table 9.3. Results for Ceriodaphnia acute test method performed on reference toxicant samples	99
Table 9.4. Results for Ceriodaphnia acute test method performed on effluent samples	100
                                               vm

-------
Table 9.5. Results for Ceriodaphnia acute test method performed on receiving water samples	101
Table 9.6. Precision of point estimates from the Ceriodaphnia acute test method	101
Table 9.7. Results for Ceriodaphnia chronic test method performed on blank samples	105
Table 9.8. Results for Ceriodaphnia chronic test method performed on reference toxicant samples. ..  107
Table 9.9. Results for Ceriodaphnia chronic test method performed on effluent samples	  109
Table 9.10. Results for Ceriodaphnia chronic test method performed on receiving water samples.  ...  Ill
Table 9.11. Precision of point estimates from the Ceriodaphnia chronic test method	  112
Table 9.12. Precision of NOEC values from the Ceriodaphnia  chronic test method	112
Table 9.13. Results for fathead acute test method performed on blank samples	114
Table 9.14. Results for fathead acute test method performed on reference toxicant samples	115
Table 9.15. Results for fathead acute test method performed on effluent samples	117
Table 9.16. Results for fathead acute test method performed on receiving water samples	118
Table 9.17. Precision of point estimates from the fathead acute  test method	118
Table 9.18. Results for fathead chronic test method performed on blank samples	122
Table 9.19. Results for fathead chronic test method performed on reference toxicant samples	124
Table 9.20. Results for fathead chronic test method performed on effluent samples	126
Table 9.21. Results for fathead chronic test method performed on receiving water samples	128
Table 9.22. Precision of point estimates from the fathead chronic test method	129
Table 9.23. Precision of NOEC values from the fathead chronic test method	  129
Table 9.24. Results for Selenastrum chronic test method performed on blank samples with EDTA. . .  133
Table 9.25. Results for Selenastrum chronic test method performed on blank samples without
            EDTA	134
Table 9.26. Results for Selenastrum chronic test method performed on reference toxicant samples
            with EDTA	  135
Table 9.27. Results for Selenastrum chronic test method performed on reference toxicant samples
            without EDTA	136
Table 9.28. Results for Selenastrum chronic test method performed on effluent samples with
            EDTA	137
Table 9.29. Results for Selenastrum chronic test method performed on effluent samples without
            EDTA	138
Table 9.30. Results for Selenastrum chronic test method performed on receiving water samples with
            EDTA	139
Table 9.31. Results for Selenastrum chronic test method performed on receiving water samples
            without EDTA	140
Table 9.32. Precision of point estimates from the Selenastrum chronic test method conducted with
            EDTA	141
Table 9.33. Precision of point estimates from the Selenastrum chronic test method conducted without
            EDTA	141
Table 9.34. Precision of NOEC values from the Selenastrum chronic test method	  142
                                               IX

-------
Table 9.35.  Results for mysid chronic test method performed on blank samples	145
Table 9.36.  Results for mysid chronic test method performed on reference toxicant samples	146
Table 9.37.  Results for mysid chronic test method performed on effluent samples	147
Table 9.38.  Results for mysid chronic test method performed on receiving water samples	148
Table 9.39.  Precision of point estimates from the mysid chronic test method	 149
Table 9.40.  Precision of NOEC values from the mysid chronic test method	 149
Table 9.41.  Results for sheepshead acute test method performed on blank samples	151
Table 9.42.  Results for sheepshead acute test method performed on reference toxicant samples	151
Table 9.43.  Results for sheepshead acute test method performed on effluent samples	 152
Table 9.44.  Results for sheepshead acute test method performed on receiving water samples	152
Table 9.45.  Precision of point estimates from the sheepshead acute test method	 153
Table 9.46.  Results for sheepshead chronic test method performed on blank samples	155
Table 9.47.  Results for sheepshead chronic test method performed on reference toxicant samples.  ... 156
Table 9.48.  Results for sheepshead chronic test method performed on effluent samples	157
Table 9.49.  Results for sheepshead chronic test method performed on receiving water samples	158
Table 9.50.  Precision of point estimates from the sheepshead chronic test method	 159
Table 9.51.  Precision of NOEC values from the sheepshead chronic test method	159
Table 9.52.  Results for silverside acute test method performed on blank samples	161
Table 9.53.  Results for silverside acute test method performed on reference toxicant samples	162
Table 9.54.  Results for silverside acute test method performed on effluent samples	 163
Table 9.55.  Results for silverside acute test method performed on receiving water samples	164
Table 9.56.  Precision of point estimates from the silverside acute test method	164
Table 9.57.  Results for silverside chronic test method performed on blank samples	167
Table 9.58.  Results for silverside chronic test method performed on reference toxicant samples	168
Table 9.59.  Results for silverside chronic test method performed on effluent samples	169
Table 9.60.  Results for silverside chronic test method performed on receiving water samples	170
Table 9.61.  Precision of point estimates from the silverside chronic test method	171
Table 9.62.  Precision of NOEC values from the silverside chronic test method	171
Table 9.63.  Successful test completion rates for test methods evaluated in the WET Study	172
Table 9.64.  False positive rates for test methods evaluated in the WET Study	173
Table 9.65.  Within-laboratory, between-laboratory, and total variability observed for test methods
            evaluated in the WET Study	174
Table 9.66.  Summarized precision estimates (CVs) for test methods evaluated in the WET Study. ... 176

-------
LIST OF FIGURES





Figure 3.1 Participant laboratory selection process	15
                                         XI

-------
Acknowledgement

This report was prepared by DynCorp Information and Enterprise Technology, Inc., under the direction of
William A. Telliard, Director of the Analytical Methods Staff of the Engineering and Analysis Division
within EPA's Office of Water.  EPA would like to acknowledge the technical input provided by members
of EPA's WET Intel-laboratory Variability Workgroup throughout the design and implementation of the
WET Study and throughout the preparation of this preliminary report.  EPA thanks the contribution of
contractor and referee laboratory support for their involvement in the implementation of the WET Study
and preparation of this preliminary report. EPA also thanks participant laboratories listed in Appendix C
for their involvement in the WET Study. In addition, EPA thanks the Association of Metropolitan
Sewerage Agencies (AMSA) for coordinating the participation of non-EPA sponsored laboratories.

EPA WET Interlaboratory Variability Workgroup
William Telliard, USEPA Office of Water, Washington, DC
Marion Kelly, USEPA Office of Water, Washington, DC
Teresa Norberg-King, USEPA Office of Research and Development, Duluth, MN
Florence Fulk, USEPA Office of Research and Development, Cincinnati, OH
John Fox, USEPA Office of Water, Washington, DC
Stephen Sweeney, USEPA Office of General Counsel, Washington, DC
Kathryn Greenwald, USEPA Office of Enforcement and Compliance Assurance,  Washington, DC

Contractor Support
DynCorp Information and Enterprise Technology, Inc., Alexandria, VA
William Peltier, Watkinsville, GA

Referee Laboratories
EA Engineering, Science and Technology, Inc., Sparks, MD
Ogden Environmental and Energy Services, Inc., San Diego, CA
MEC Analytical, Inc., Tiburon, CA
EnviroSystems, Inc., Hampton, NH
                                             Xll

-------
Executive Summary

This report presents the results of the U.S. Environmental Protection Agency's (EPA's) interlaboratory
variability study of EPA short-term chronic and acute whole effluent toxicity (WET) test methods (the
WET Study). In 1995, EPA promulgated 17 WET test methods for use in monitoring toxicity under the
National Pollutant Discharge Elimination System (NPDES) [60 FR 53529; October 16, 1995].  In the
WET Study, the following 12 approved WET test methods were evaluated: cladoceran, Ceriodaphnia
dubia, acute test; cladoceran,  Ceriodaphnia dubia, survival and reproduction test; fathead minnow,
Pimephales promelas, acute test; fathead minnow, Pimephales promelas, larval survival and growth test;
green alga, Selenastrum capricornutum, growth test; mysid, Mysidopsis bahia, survival, growth, and
fecundity test; sheepshead minnow, Cyprinodon variegatus, acute test; sheepshead minnow, Cyprinodon
variegalus, larval survival and growth test; inland silverside, Menidia beryllina, acute test; inland
silverside, Menidia beryllina, larval survival and growth test; red macroalga, Champia parvula,
reproduction test;  and mysid, Holmesimysis costata, acute test. The purpose of the WET Study was to
characterize interlaboratory variability, the rate of successful test completion, and the rate of false positive
incidence for the evaluated test methods.

A total of 4 referee laboratories and 55 participant laboratories were involved in the WET Study with 7 to
35 participant laboratories per method. Interlaboratory data were not obtained for the red macroalga,
Champia parvula, reproduction test, and the mysid, Holmesimysis costata, acute test, due to insufficient
participant laboratory support. For all other methods, referee laboratories prepared and distributed three or
four blind test samples to each participant laboratory. Laboratories received some combination of the
following test sample types: blank sample, reference toxicant sample, municipal or industrial effluent
sample, and receiving water sample. Participant laboratories were required to analyze each blind test
sample according  to the promulgated WET test method manuals and specific instructions provided in the
WET Study participant laboratory standard operating procedure (SOP) document.

Participant laboratories submitted all test data on analyzed samples to the Sample Control Center (SCC)
operated by DynCorp Information & Enterprise Technology, Inc., for independent review and calculation.
SCC biologists reviewed test data to verify that all pertinent information was provided, tests were
conducted in accordance with the WET method manuals and the WET Study SOP, and test results were
accurately calculated. Following test review, results were compiled and method performance
characteristics (interlaboratory variability, successful test completion rate, and false positive rate) were
calculated for each WET test method.

Table 1 displays summarized  results from the WET Study. Successful test completion rates were greater
than 90% for all WET test methods except the Ceriodaphnia chronic (82%) and Selenastrum chronic
(63.6% with Ethylenediaminetetraacetic acid [EDTA] and 65.9% without EDTA) test methods.  False
positive rates were less than 5% for all WET test methods except the Selenastrum chronic test method

                                               xiii

-------
conducted without EDTA (33.3%). Interlaboratory variability was described by the coefficient of variation
(CV) calculated for point estimates.  Interlaboratory CVs ranged from 20.0% to 38.5% for acute test
methods. CVs for chronic WET test methods ranged from 8.72% to 40.6% for survival endpoints and
from 10.6% to 58.5% for sublethal endpoints.

Table 1.  Summarized test results from the WET Study.
Test method
Ceriodaphnia acute
Ceriodaphnia
chronic
Fathead acute
Fathead chronic
Selenastrum
chronic
(with EDTA)
Selenastrum
chronic
(without EDTA)
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Successful
test
completion
rate
(%)
95.2
82.0
100
98.0
63.6
65.9
97.7
100
100
94.4
100
Survival endpoint
False positive
rate"
(%)
0.00
0.00
0.00
0.00
-
-
0.00
0.00
0.00
0.00
0.00
cvb
(%)
29.0
21.5
20.0
13.4
-
-
31.2
26.0
8.72
38.5
40.6
Growth endpoint
False positive
rate"
(%)
-
-
-
4.35
0.00
33.3
0.00
-
0.00
-
0.00
cvb
(%)
-
-
-
20.9
34.3C
58. 5C
41.3
-
10.6
-
43.8
Reproduction endpoint
False positive
rate"
(%)
-
3.70
-
-
-
-
0.00
-
-
-
-
cvb
(%)
-
35.1
-
-
-
-
-
-
-
-
-
a False positive rates reported for each method represent the higher of false positive rates observed for hypothesis testing or
point estimate endpoints.
b CVs reported for each method represent the CV based on total variance averaged across sample types.
c CVs reported for the Selenastrum  chronic test are based on IC25 values.
                                                  XIV

-------
1.0 INTRODUCTION AND BACKGROUND

Whole effluent toxicity (WET) is defined as "the aggregate toxic effect of an effluent measured directly by
an aquatic toxicity test" [54 FR 23686; June 2, 1989].  EPA has promulgated 17 aquatic toxicity test
methods designed specifically for measuring WET [60 FR 53529; October 16, 1995]. The EPA-approved
WET test methods resulted from many years of development and testing by EPA, States, municipalities,
academia, and the regulated community.  These WET test methods employ a suite of standardized
freshwater, marine and estuarine plants, invertebrates, and vertebrates to measure acute and short-term
chronic toxicity in effluents and receiving waters. WET test methods approved for use in National
Pollutant Discharge Elimination System (NPDES) monitoring are listed in 40 CFR § 136.3, Table IA.
Specific test procedures for conducting the approved WET tests are published in the following three test
method manuals (the WET method manuals):

•       USEPA, Methods for Measuring the Acute Toxicity of Effluents and Receiving Waters to
        Freshwater and Marine Organisms, Fourth Edition, EPA-600-4-90-027F, August 1993
•       USEPA, Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving
        Water to Freshwater Organisms, Third Edition, EPA-600-4-91-002, July 1994
•       USEPA, Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving
        Water to Marine and Estuarine Organisms, Second Edition, EPA-600-4-91-003, July 1994

1.1     Regulatory Background

The Clean Water Act (CWA) was enacted in 1972 with the objective of "restoring the chemical, physical,
and biological integrity of the Nation's waters." Along with other goals, CWA section 101(a)(3) states that
"it is the national policy that the discharge of toxic pollutants in toxic amounts be prohibited." The U.S.
Environmental Protection Agency (EPA)  has pursued this goal through implementation of the water quality
standards program and the NPDES permitting program. These programs have adopted an integrated
strategy of water quality-based toxics control that includes three approaches: chemical-specific control,
whole effluent toxicity (WET) control, and biological criteria/bioassessment.

To implement this strategy, States are encouraged to define numeric or narrative water quality standards
that include chemical-specific criteria, criteria for WET, and biological criteria. Some states have included
numeric criteria for WET, while others have relied on narrative  criteria. These water quality standards and
criteria are maintained by controlling the  discharge of pollutants through the NPDES permitting program.
When a discharge causes  or has a reasonable potential to cause or contribute to the excursion  of numeric or
narrative water quality standards, a NPDES permit limit will be issued to control the discharge. Permit
limits for WET are established if the discharge has a reasonable potential to cause or contribute to the
excursion of water quality standards for WET.

-------
1.2    The WET Study

In 1999 - 2000, EPA conducted an interlaboratory variability study of EPA short-term chronic and acute
whole effluent toxicity test methods (the WET Study).  EPA conducted the WET Study to quantify
interlaboratory precision and validate the current versions of 12 of the 17 approved WET test methods.
The study was precipitated by litigation over the rulemaking that standardized and approved the WET test
methods for use in NPDES monitoring [60 FR 53529; October 16, 1995]. In a settlement agreement with
the litigants, EPA agreed to conduct the WET Study and determine false positive rates, successful test
completion rates, and coefficients of variation (CVs) for each of 12 WET test methods. Based on the
results of the WET Study and peer review comments, EPA will ratify or withdraw each of the WET test
methods.

1.3    Other EPA Documents

The WET method manuals were incorporated by reference in the WET final rule [60 FR 53529; October
16, 1995]. EPA issued clarifications to the WET test methods  on April 10, 1996, via a memorandum from
Tudor Davies, Director of EPA's Office of Science and Technology.  This memorandum, titled
"Clarifications Regarding Flexibility in 40 CFR Part 136 Whole Effluent Toxicity (WET) Test Methods"
(USEPA, 1996), provided clarification on the following WET test issues: pH and ammonia control,
temperature, hardness, test dilution concentrations, and acceptance criteria for Champia parvula.

In February 1999, EPA published an errata to the WET method manuals (USEPA, 1999).  This errata
amended the approved WET test methods to correct typographical errors and omissions, provide technical
clarification, and establish consistency among the 1995 WET rule language and the WET method manuals.

In June 2000, EPA published a guidance document titled, "Understanding and Accounting for Method
Variability in Whole Effluent Toxicity Applications Under the National Pollutant Discharge Elimination
System Program" (USEPA, 2000b).  This guidance document was intended to provide regulatory
authorities with an understanding of WET test variability and provide guidance on accounting for and
minimizing WET test variability and its effects on the regulatory process.

In July 2000, EPA published a guidance document titled, "Method Guidance and Recommendations for
Whole Effluent Toxicity (WET) Testing (40 CFR Part 136)" (USEPA, 2000a).  This document included
guidance and recommendations  on nominal error rate adjustments, confidence intervals, concentration-
response relationships, dilution series selection, and selection of an acceptable dilution water for WET
testing.

-------
2.0 STUDY DESIGN AND OBJECTIVES

2.1    Objectives

In conducting the WET Study, EPA's primary objectives were to (1) generate data to characterize the
interlaboratory variability of the 12 WET test methods through the determination of CVs for LCSOs
(median lethal effect concentrations) and IC25s (25% inhibition  concentrations) and ranges for NOEC (No-
Observed-Effect-Concentration) values, (2) obtain data on the rate at which participating laboratories
successfully completed WET tests initiated (successful test completion rate), and (3) generate data on the
rate at which WET tests indicate "toxicity" is present when measuring non-toxic samples (false positive
rate).  EPA developed a preliminary study plan to meet these objectives, and made this study plan available
for public and peer review comment on October 9, 1998. EPA revised the study plan in response to public
and peer review comments; the study plan was finalized on June 11, 1999 (see Appendix A). The WET
Study was conducted according to the final study plan with the following exceptions:

•      The minimum number of participant laboratories was reduced from nine to seven for the
       sheepshead  acute and chronic test methods. Only  seven laboratories qualified to participate in
       these methods.  A minimum of seven participant laboratories still satisfied the data quality
       objective requiring a minimum of six complete and useable data sets for each WET test method.
•      No interlaboratory testing was conducted for the Champia chronic and Holmesimysis acute test
       methods. The interlaboratory testing phase was canceled for these test methods due to insufficient
       participant laboratory support.  Only one participant laboratory could be procured for the
       Champia chronic test method, and only two participant  laboratories could be procured for the
       Holmesimysis acute test method. This report presents only referee laboratory preliminary testing
       results for these two methods.

2.2    General Study Design

The WET Study was conducted in four phases designed to accomplish the overall study objectives. These
phases, and the specific objectives associated with each phase, are shown in Table 2.1.

2.2.1   Study Management

The WET Study was directed by EPA with contractor support provided by DynCorp Information &
Enterprise Technology, Inc., under the Sample Control Center (SCC) contract.  EPA's Office  of Water
Engineering and Analysis Division Analytical Methods Staff (AMS) and EPA's Office  of Research and
Development (ORD) staff provided overall management and technical oversight of the study.  Laboratory
procurement, day-to-day management, coordination of study activities, data review, and preparation of the
preliminary study report were performed by SCC under AMS and ORD guidance. Four referee laboratory

-------
contractors conducted preliminary testing and prepared and distributed blind test samples to 55 participant
laboratories.  The participant laboratories conducted the WET tests and submitted data reports to SCC.
Table 2.1.  Four phases and specific objectives of the WET Study.
            Phase
                      Objectives
  1 - Laboratory Procurement
Identify potential referee and participant laboratories to support the study
Prequalify and select referee laboratories for Phases 2, 3, and 4
Prequalify and select participant laboratories for Phase 4 of the study
 2 - Preliminary Testing
Determine the suitability of selected real-world sample matrices for use in
the study through characterization of physical, chemical, and toxicological
properties of the test sample
Determine the appropriate spiking concentrations for reference toxicant
samples to achieve the desired range of toxicity
Determine the persistence of toxicity in real-world test samples
Assess whether the desired range of sample toxicity will be maintained in
test samples following shipping and handling
 3 - Sample Preparation and
    Distribution
Prepare real-world and synthetic test samples for use by referee and
participant laboratories in Phase 4
Minimize variability between samples prepared for and distributed to each of
the Phase 4 laboratories
Distribute blind test samples to all qualified laboratories for initial use within
36 hours of individual sample shipment from the referee laboratories
 4 - Intel-laboratory Testing
Obtain interlaboratory test data for each WET method using four real-world
and synthetic test samples to evaluate precision of the test methods, the rate
at which laboratories successfully completed tests initiated, and the rate at
which the tests indicate "toxicity" is present when measuring non-toxic
samples	
2.2.2   Methods Evaluated
EPA evaluated 12 of the 17 promulgated WET test methods in the WET Study. These included two acute
freshwater methods, three chronic freshwater methods, three acute marine methods, and four chronic
marine methods.  The test methods and the test endpoints evaluated in the WET Study are displayed in
Table 2.2 and Table 2.3, respectively.  Each  test method was conducted in accordance with the WET
method manuals (USEPA, 1993; USEPA, 1994a; USEPA, 1994b), and as appropriate,  specific guidance
provided in the WET Study participant laboratory standard operating procedure (SOP) documents
(Appendix B), the WET method manuals errata (USEPA, 1999), and clarifications provided in an April 10,
1996, memorandum from Tudor Davies, Director of EPA's Office of Science and Technology (USEPA,
1996).

-------
2.2.3   Laboratories

A referee laboratory was selected to provide support for each test method evaluated in the WET Study. The
referee laboratory was responsible for conducting preliminary testing on sample types to ensure that
samples used in the WET Study were stable and provided the desired level of effect.  Referee laboratories
also collected, prepared, and distributed test samples for interlaboratory testing, and conducted testing
simultaneous with participant laboratory testing to provide further information on sample consistency.
Table 2.2.  WET test methods included in the WET Study.
Test method
Cladoceran, Ceriodaphnia dubia, acute test0
Cladoceran, Ceriodaphnia dubia, survival and reproduction testd
Fathead minnow, Pimephales promelas, acute test0
Fathead minnow, Pimephales promelas, larval survival and growth
testd
Green alga, Selenastrum capricormitum, growth testd
Mysid, Mysidopsis bahia, survival, growth, and fecundity test0
Sheepshead minnow, Cyprinodon variegatus, acute test0
Sheepshead minnow, Cyprinodon variegatus, larval survival and
growth test0
Inland silverside, Menidia beryllina, acute test0
Inland silverside, Menidia beryllina, larval survival and growth test0
Red macroalga, Champia parvula, reproduction test0
Mysid, Holmesimysis costata, acute tesf'f
Common test
method name"
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Champia chronic
Holmesimysis acute
Test method
numberb
-
1002.0
-
1000.0
1003.0
1007.0
-
1004.0
-
1006.0
1009.0
-
a Common test method names were used in this report to refer to the test methods in the WET Study.
b Test method numbers were not designated for acute test methods in USEPA, 1993.
0 USEPA, Methods for Measuring the Acute Toxicity of Effluents and Receiving Waters to Freshwater and Marine Organisms,
Fourth Edition, EPA-600-4-90-027F, August 1993
d USEPA, Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Water to Freshwater
Organisms, Third Edition, EPA-600-4-91-002, July 1994
0 USEPA, Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Water to Marine and Estuarine
Organisms, Second Edition, EPA-600-4-91-003, July 1994
f The EPA-approved acute test with Holmesimysis costata was performed using the acute test procedures for Mysidopsis bahia
and test conditions optimized for H. costata.

-------
Interlaboratory testing was conducted by participant laboratories selected for each test method. The
number of participant laboratories conducting each method was dependent upon the number of laboratories
responding to the solicitation, the number of laboratories qualified to conduct the given test method (as
determined by prequalification requirements), and the number of laboratories that were sponsored by
parties external to EPA. The WET Study plan called for a minimum of 9 and a maximum of 20
participant laboratories to constitute the base study design for each method. Additional laboratories
(greater than 20) participated on a more limited basis as part of an extended study design.  In deviation
from this design, only seven qualified laboratories could be procured to participate in the sheepshead acute
and chronic test methods. In addition, interlaboratory testing for the Champia chronic and Holmesimysis
acute test methods was canceled due to insufficient participant laboratory support (see Section 2.1).  See
Section 3 for more detailed description of referee and participant laboratory selection.
Table 2.3. Endpoints evaluated for each test method in the WET Study.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Champia chronic
Holmesimysis acute
Acute tests
Survival
LC50
X
-
X
-
-
-
X
-
X
-
-
X
Test
duration
(hours)
48
-
96
-
-
-
96
-
96
-
-
96
Short-Term Chronic Tests
Survival
LC50
NOEC
-
X
-
X
-
X
-
X
-
X
-
-
Growth
IC25
NOEC
-
-
-
X
X
X
-
X
-
X
-
-
Reproduction
IC25
NOEC
-
X
-
-
-
X
-
-
-
-
X
-
Test
duration
(days)
-
8a
-
7
4
7
-
7
-
7
7b
-
a The C. dubia test acceptability criteria states that the test is complete when 60% of controls have 3 broods (approximately 7
days); for purposes of this study, all tests continued for 8 days and each laboratory was requested to carefully distinguish and
carefully record the number of broods.
 C. parvula were exposed to test samples for 2 days, followed by a 5-day recovery period in control water.

-------
2.2.4   Samples

For each test method, four test sample types were prepared in bulk by the referee laboratory, divided, and
distributed to participant laboratories for testing. The four sample types included (1) blank sample, (2)
reference toxicant sample, (3) effluent sample, and (4) receiving water sample.  Blank and reference
toxicant samples were distributed to participant laboratories as liquid ampules (to mix and dilute to the
required volume at the participant laboratory), while effluent and receiving water samples were distributed
as whole-volume samples (consisting of the full volume necessary to conduct the test).  The blank sample
was a non-toxic sample prepared as the typical synthetic control dilution water for each test method.
Testing of the blank sample provided a means of determining the false positive rate for each test method.
Interlaboratory precision was evaluated through testing of the reference  toxicant, effluent, and receiving
water sample types. These sample matrix types (with the exception of the effluent sample type for the
Champia chronic and Holmesimysis acute test methods) were spiked with a toxicant to achieve a desired
level of effect and facilitate the evaluation of precision.

Laboratories participating in the base study design received four blind test samples, which included some
combination of the four test sample types (blank, reference toxicant, effluent, and receiving water).
Laboratories that participated in the extended study design received three blind test samples as ampules,
which included some combination of blank and reference toxicant samples. The combination of blind test
samples received by individual laboratories included replicate (i.e., duplicate) test samples for each test
method except  for methods with less than nine participant laboratories.  Replicate samples were always
shipped together and tested simultaneously, with the exception of the Selenastrum  chronic test, where only
one sample was tested per week. See Sections 5 and 6 for a more detailed description of sample
preparation and distribution, respectively.

2.2.5    Schedule

Interlaboratory testing in the WET Study was conducted from September,  1999, through April, 2000. The
final schedule for the WET Study is provided in Table 2.4. For each method (with the exception of the
Selenastrum chronic method), testing was conducted in two testing periods. Each participant laboratory
simultaneously tested two samples during each testing period. For the Selenastrum chronic method, testing
was conducted  in four testing periods, with a single sample tested with and without EDTA during each
testing period.

Samples were shipped overnight to arrive at each participant laboratory on the day of scheduled testing.
Sample distribution was organized such that all effluent samples were tested on the same day in all
participant laboratories, and all receiving water samples were tested on the same day (with the  exception of
delays due to sample or organism shipments).  Since the synthetic matrix and ampule form of blank and
reference toxicant samples provided more sample stability, these  samples were distributed for testing in
either the first or second testing periods.  As a result, tests conducted on reference toxicant and blank

-------
samples were initiated on one of two testing dates. Replicate samples were always distributed together for
simultaneous testing.
Table 2.4. Final schedule for interlaboratory testing during the WET Study.
Activity
Fathead chronic testing period #1
Fathead chronic testing period #2
Ceriodaphnia chronic testing period #1
Silverside chronic testing period #1
Fathead acute testing period #1
Silverside chronic testing period #2
Ceriodaphnia chronic testing period #2
Silverside acute testing period #1
Fathead acute testing period #2
Ceriodaphnia acute testing period #1
Silverside acute testing period #2
Ceriodaphnia acute testing period #2
Mysid chronic testing period #1
Mysid chronic testing period #2
Sheepshead acute testing period #1
Selenastrum chronic testing period #1
Sheepshead acute testing period #2
Selenastrum chronic testing period #2
Sheepshead chronic testing period #1
Selenastrum chronic testing period #3
Sheepshead chronic testing period #2
Selenastrum chronic testing period #4
Testing
(start date -
9/28/99 -
10/5/99 -
10/12/99 -
10/19/99 -
10/21/99-
10/26/99
10/26/99
11/2/99-
11/4/99-
1 1/9/99 -
1 1/9/99 -
11/11/99-
2/22/00 -
2/29/00
3/7/00 -
3/9/00 -
3/14/00-
3/16/00-
3/21/00-
3/23/00 -
3/28/00
3/30/00
dates8
finish date)
10/5/99
10/12/99
10/20/99
10/26/99
10/25/99
- 11/2/99
- 11/3/99
11/6/99
11/8/99
11/11/99
11/13/99
11/13/99
2/29/00
- 3/7/00
3/11/00
3/13/00
3/18/00
3/20/00
3/28/00
3/27/00
- 4/4/00
- 4/3/00
a Samples were shipped to arrive at participant laboratories on the indicated start date. Tests were initiated on the same day as
sample receipt.

-------
3.0 LABORATORY PROCUREMENT
3.1
Identification and Solicitation of Potential Referee Laboratories
Since the responsibilities of the referee laboratory would be demanding and critical to successful
implementation of the WET Study, SCC solicited a select list of laboratories that possessed exceptional
qualifications, based on EPA technical staff recommendations.  Solicitation bid packages included the
following documents: (1) referee laboratory prequalification document; (2) referee laboratory statement of
work (SOW), including a preliminary study schedule; and (3) referee laboratory bid sheet.  Referee
laboratories were first solicited in December, 1998. Additional solicitations to an expanded number of
laboratories were necessary to fill referee laboratory positions for all of the test methods (Table 3.1).

Table 3.1.  Summary of referee laboratory solicitation.
Solicitation
date
12/29/98
1/22/99
2/19/99
Response
deadline
1/13/99
2/4/99
3/5/99
Methods solicited
All twelve methods
All seven marine methods
Silverside acute and chronic;
Holmesimysis acute
Laboratories
solicited
7
12
19
Number of
qualified responses
2
2
3
3.2     Identification and Solicitation of Potential Participant Laboratories

EPA attempted to maximize the number of qualified laboratories participating in the WET Study and select
laboratories that were representative of laboratories throughout the United States that routinely conduct
WET tests for permittees.  SCC biologists and EPA program staff identified a list of potential participant
laboratories from a variety of sources, including EPA and State environmental agencies, the Society of
Environmental Toxicology and Chemistry (SETAC), reviews of the public literature, the Directory of
Environmental Laboratories (DynCorp, 1996), and the list of laboratories conducting testing for EPA's
Discharge Monitoring Report Quality Assurance (DMRQA) program.  In addition, the petitioners provided
a list of laboratories interested in participating without EPA sponsorship and a list of potential non-EPA
sponsors.

On July 9,  1999, SCC solicited all 319 laboratories included in the compiled potential laboratory list.
Laboratories solicited included state, academic, municipal, industrial and private laboratories. Solicitation
bid packages included the following documents: (1) a detailed cover letter describing the solicitation; (2) a
participant laboratory prequalification document; (3) a participant laboratory SOW, including a
preliminary study schedule; and (4) a participant laboratory bid sheet.

-------
3.3     Prequalification Requirements for Participant Laboratories

To ensure that laboratories participating in the WET Study possessed the capacity and capabilities,
experience and proficiency, and quality assurance and quality control systems necessary to meet the needs
of the study, EPA required all laboratories to meet specific prequalification requirements.

3.3.1    Prequalification Documentation

To demonstrate its qualifications, each potential participant laboratory was required to provide the 16
prequalification items listed below. Prequalification was conducted independently for each test method,
and laboratories could submit prequalification materials for any or all methods evaluated in the WET
Study.

General information

(1)     A cover page with the laboratory name, address, telephone number, fax number, e-mail address,
        contact person, and additional contacts for day-to-day sample tracking and technical issues if
        different from primary contact.

(2)     A statement on the number of tests that the laboratory can conduct at one time with the proposed
        staff, including the number of tests using a single test method and the number of tests using
        multiple test methods (e.g., three  C. dubia survival and reproduction tests, three fathead minnow
        survival and growth tests, and two of each simultaneously).

Capacity and capabilities

(3)     A statement that the combination of facilities, equipment, staff, and laboratory capabilities are
        sufficient to meet study needs.

(4)     Detailed information on the type and size of laboratory and test equipment used for conducting
        each test method, including information on temperature control, sample storage,  water purification
        devices (i.e., Millipore Milli-Q® filtration and ion exchange), and dilution water sources.
        Laboratories were required to provide summaries of routine water quality monitoring data on
        dilution water and water used for culturing or maintaining each species (e.g., 3-4 months of pH,
        alkalinity, hardness, and salinity measurements on dilution and culture waters).

(5)     A statement that the laboratory can receive next day deliveries (including Saturday deliveries) via
        overnight carriers (i.e., Fed Ex, UPS, etc.) and initiate a test on the same day as receipt.

(6)     A list of laboratory staff able to participate in the study, including resumes and titles.
                                                10

-------
(7)     Information on the source of organisms, including whether organisms are cultured in-house or
        obtained externally.  If organisms are cultured in-house, the laboratory was required to provide
        standard operating procedures for organism culturing (as required in number 9 below), a summary
        of how culture performance is assessed, and data on culture performance (e.g., Ceriodaphnia
        dubia brood board monitoring data for the past six month or records of Pimephales promelas egg
        production). If organisms are obtained from an external source, the laboratory was required to
        specify the source, number of organisms that can be obtained from this source on a given day, age
        of obtained organisms, and organism holding and maintenance conditions.

Experience and proficiency

(8)     Copies of internal standard operating procedures (SOPs) for conducting each of the test  methods.
        Internal laboratory SOPs for each test method were required to be in place with dates of SOP
        origination.

(9)     Copies of supporting internal laboratory SOPs for organism culturing, food preparation, and
        dilution water preparation for each species and each method.

(10)    A statement on the number of effluent tests conducted in the last year using  each of the WET test
        methods. The laboratory was required to specify the frequency with which test acceptability
        criteria were met in these tests and the average control response measured in these tests.

(11)    A statement regarding State or regional certifications and documentation of current certifications
        (if applicable).

Quality assurance/quality control

(12)    Evidence that the laboratory maintains control (cusum) charts for reference  toxicant tests for each
        method. The laboratory was required to submit the most current control chart for each test
        method, covering at least 12-24 data points and showing control limits.  The raw data (actual data
        sheets and summarized data) for each data point also was required. Data charts with NOEC and/or
        IC25 for the same test values were requested as well as an explanation of why one is used rather
        than the other.  Explanations were required if methods used to develop reference toxicant control
        charts deviated from promulgated methods or from the previous edition of a relevant test protocol.

(13)    Evidence that reference toxicant tests are conducted at the appropriate frequency (e.g., monthly for
        tests that are routinely run for NPDES permits). Along with control chart information described
        above, the laboratory was required to provide a statement on the frequency of reference  toxicant
        testing. If control charts (particularly for less common test methods) were composed of fewer than
        12-24 data points, an explanation was required.
                                                11

-------
(14)    Copies of internal laboratory SOPs for conducting reference toxicant tests and constructing control
        charts.  This information had to include a narrative explanation of the width of the control limits
        for the  laboratory and a statement of corrective action for any toxicity test endpoint value that falls
        outside the control limits.

(15)    Results of the most recent DMRQA study, if the lab participated. The laboratory was required to
        provide data point(s) for each method performed for the previous year's DMRQA study. If the
        laboratory did not participate, a narrative statement to that effect was required.

(16)    A signed statement of accuracy and completeness of submitted prequalification materials.  The
        following statement, signed and dated by an authorized representative of the laboratory, was
        required: "I certify that the information provided in this prequalification package is complete and
        accurate to the best of my knowledge."

3.3.2   Prequalification Determination

SCC evaluated  prequalification materials and recommended laboratory rejection based on the following
criteria:

(1)     Combination of facilities, equipment, staff and lab capacity and capabilities were insufficient to
        meet study needs.

(2)     Organism source information was not provided, culture and/or collection information was severely
        lacking, or source information was inadequate to assess the health of the organisms routinely used.

(3)     Internal laboratory SOPs for each method were vague and could not be discerned and/or were
        generally insufficient to support performance of the methods in accordance with specific
        instructions provided by EPA.

(4)     Statements regarding the number of effluent tests conducted per year, test acceptability rates,
        average control response, and/or State certifications were not provided, did not adequately
        demonstrate proficiency in the test method, or did not adequately demonstrate that the laboratory is
        representative of laboratories throughout the United States that routinely conduct WET testing for
        permittees.

(5)     Control charts were not adequately maintained for reference toxicant tests, or data were not
        provided (cusum chart for each endpoint and raw data for each data point). Control charts did not
        cover 12-24 data points for each species and test method, and an acceptable explanation was not
        provided.
                                                12

-------
(6)     Reference toxicant tests were not conducted at the appropriate frequency (monthly for tests that are
        routinely run for permits) and a satisfactory explanation was not provided.

(7)     No acceptable explanation or evidence of corrective action was provided for any control chart
        value falling outside the control limits.

(8)     The laboratory did not provide the most recent DMRQA study results, or an acceptable
        explanation for non-passing results was not provided. If the laboratory did not participate in the
        DMRQA study, the laboratory did not include an acceptable explanation as to why they did not
        participate.

(9)     No signed statement of accuracy and completeness of prequaliflcation materials was included.

Laboratories that failed to meet the prequaliflcation criteria due to an incomplete submission were notified
via fax of their deficiency and allowed an opportunity to submit the missing information. Laboratories that
did not respond to this notification or were unable to provide the missing information failed to prequalify
for the WET Study. Only eight laboratories failed to prequalify for the WET Study. Two of these
laboratories failed to prequalify for more than one test method.

3.4     Prequalification  and Selection of Referee Laboratories

Referee laboratories were required to  submit the same prequaliflcation materials listed in Section 3.3.1  for
participant laboratories. In addition, referee laboratories were required to submit three client references
and provide background information on potential real-world effluent and receiving water sample sources.
Referee laboratory prequaliflcation materials were evaluated based on the rejection criteria listed in  Section
3.3.2 and the additional reference and sample source requirements.  The capacity and capabilities of
potential referee laboratories were highly scrutinized to ensure that the laboratory could meet the sample
collection,  preparation, distribution, and testing requirements of the study.  All potential referee
laboratories that met prequaliflcation criteria and were determined to possess the capacity and capabilities
to meet the needs of the study were considered for the referee laboratory position.  From this pool of
qualified laboratories, the lowest bidder was selected as the referee laboratory for each bid lot of test
methods. The referee laboratories selected for each test method are listed in Table 3.2.
                                                 13

-------
Table 3.2. Referee laboratories selected for the WET Study.
Referee laboratory
EA Engineering, Science and
Technology, Inc.
EA Engineering, Science and
Technology, Inc.
Ogden Environmental and Energy
Services, Inc.
MEC Analytical, Inc
EnviroSystems, Inc.
Method(s) supported
Ceriodaphnia acute and chronic; Fathead acute and
chronic; Selenastrum chronic
Mysid chronic; Sheepshead acute and chronic
Silverside acute and chronic
Holmesimysis acute
Champia chronic
Date awarded
03/10/1999
04/01/1999
06/08/1999
06/02/1999
04/14/1999
3.5     Selection of Participant Laboratories

Laboratories that met all prequalification criteria (Section 3.3) were eligible for participation in the WET
Study. From the pool of prequalified laboratories, those laboratories with the nine lowest bids per method
were selected for EPA sponsorship and participation in the base study design. If a tie for the ninth lowest
bid was encountered, selection among equal bid prices from prequalified laboratories was determined based
on business classification, with preferences granted in the order of small company, minority owned, woman
owned, or large company.  From the remaining prequalified laboratories, those that had identified a willing
sponsor outside of EPA (non-EPA sponsorship) were also selected for participation in the WET Study.
From this group of prequalified, non-EPA sponsored laboratories, a maximum of 11 laboratories (for each
WET test method) were randomly selected for participation in the base study  design.  The 9 EPA-
sponsored laboratories and the 11 randomly chosen non-EPA-sponsored laboratories constituted the 20
laboratories included in the base study  design for each WET test method. All remaining prequalified
laboratories with non-EPA sponsorship were selected for participation in the extended study design. Figure
3.1  displays the process of participant laboratory selection.  SCC formally notified all laboratories on
September 8, 1999, of their selection, sponsor for the methods (EPA or non-EPA), and level of
participation (base or extended design). The results of laboratory prequalification and selection are
displayed in Table 3.3 for each test method. A total of 55 laboratories were selected to participate in the
WET Study, with many laboratories participating in multiple methods.  See Appendix C for an
alphabetical list of the participant laboratories.

3.6     Participant Laboratory Meeting

EPA invited all laboratories that submitted prequalification materials to attend a participant laboratory
meeting held in Chicago, Illinois, on September 16, 1999.  At the participant laboratory meeting, EPA and
SCC staff presented the study plan for  the WET Study and highlighted participant laboratory tasks and
requirements.  Participant laboratories  had the opportunity to meet EPA and SCC staff and to ask
                                                14

-------
questions regarding the WET Study and their responsibilities. At the meeting, EPA and SCC staff did not
release any information regarding sample contents, sample descriptions, or sample distribution schemes
that would jeopardize the blind aspect of the study. Following the meeting, SCC distributed meeting notes
to all participant laboratories that were unable to attend.


Figure 3.1  Participant laboratory selection process.
 PREQUALIFICATION
     Did lab pass
prequalification criteria?
    BID COST
   EVALUATION
                                                           No
                                                                                 No involvement
                                Yes
  Was bid among 9
lowest for the method?
     RANDOM
   SELECTION
        LAB
    INVOLVEMENT
    ASSIGNMENT
No
    Was external
sponsorship available?
                                Yes
    Base Design
                                     No involvement
                                                               Yes
                            Was lab randomly selected for
                                  base design?
                                              Yes
                                                              No
       (  Extended Design  j
                                                15

-------
Table 3.3.  Results of participant laboratory prequalification and selection.
Test method
Ceriodaphnia acute
Ceriodaphnia
chronic
Fathead acute
Fathead chronic
Selenastrum chronic
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Champia chronic*
Holmesimysis acute3
No. of labs
solicited
319
319
319
319
319
319
319
319
319
319
319
319
No. of labs
responding
48
55
50
49
16
16
7
9
13
15
2
2
No. of labs
failing to
prequalify
2
0
1
0
2
3
0
2
1
2
1
0
No. of labs selected for base study
design
EPA sponsored
labs
9
9
9
9
9
9
7
7
9
9
1
2
Non-EPA
sponsored labs
11
11
11
11
2
2
0
0
0
1
1
2
No. of labs
selected for
extended study
design
Non-EPA
sponsored labs
8
15
9
7
0
0
0
0
0
0
0
0
Total no. of
participant labs
28
35
29
27
11
11
7
7
9
10
1
2
1 Intel-laboratory testing was canceled for this method due to insufficient participant laboratory support.  Tests were performed by the referee laboratory only.
                                                                           16

-------
4.0 PRELIMINARY TESTING

4.1     Preliminary Testing Plan

Referee laboratories contracted to support each test method during the WET Study were responsible for
conducting preliminary testing.  A four-part preliminary testing plan was developed and instituted to
confirm the appropriate selection of samples for use in interlaboratory testing. The four parts of the
preliminary testing plan consisted of the tests listed in Table 4.1 and accomplished the following objectives.

•       Part 1 - Background testing: determine the suitability of real-world sample matrices (i.e.,
        effluent, receiving water) for use in the WET Study through characterization of physical, chemical,
        and toxicological properties
•       Part 2 - Range-finding:  determine the appropriate spiking concentrations  for the preparation of
        spiked effluent, receiving water, and reference  toxicant samples types
•       Part 3 - Holding time testing: determine the persistence of toxicity in spiked effluent and
        receiving water samples
•       Part 4 - Definitive testing: assess whether test samples provided the desired range of toxicity
        following sample preparation,  shipping, and handling

4.1.1    Part 1 - Background Testing

Part 1 of preliminary testing verified that selected real-world (effluent and receiving water) sample matrix
sources were acceptable for study use by assessing the  physical, chemical, and toxicological characteristics
of the samples.  Referee laboratories were required to submit information on potential real-world effluent
and receiving water sample sources as part of their prequalification materials. EPA and SCC reviewed
these materials including historical information from  the source, and made a preliminary selection of the
effluent sample source and the receiving water source for each test method.  Following this determination,
the referee laboratory initiated Part 1 of preliminary testing.

The referee laboratories collected preliminary test samples according to Section 8 of the method manuals
(USEPA,  1994a; USEPA, 1994b). Following  sample collection, the referee laboratories conducted
physical and chemical analyses of both the effluent sample and the receiving water sample, including
alkalinity, hardness, pH, temperature, total residual chlorine, total ammonia, dissolved oxygen,  total
dissolved solids, total suspended solids, total organic  carbon, biological oxygen demand, and chemical
oxygen demand.  For samples that were to be used in marine tests, salinity and copper also were measured.
                                                17

-------
Table 4.1. Summary of preliminary testing requirements.
Test method
Ceriodaphnia acute
Ceriodaphnia
chronic
Fathead acute
Fathead chronic
Selenastrum
chronic
Mysid chronic
Sheepshead acute
Sheepshead chronic
Sheepshead acute
Sheepshead chronic
Champia chronic
Holmesimvsis acute
Part 1 - Background
testing
Unspiked
effluent

•

•
•a
•

•

•
•
•
Unspiked
receiving
water

•

•
•a
•

•

•
•
•
Part 2 - Range-finding
Spiked
effluent
•
•
•
•
•a
•
•
•
•
•
•c
•c
Spiked
receiving
water
•
•
•
•
•a
•
•
•
•
•
•
•
Reference
toxicant
•
•
•
•
•a
•
•
•
•
•
•
•
Part 3 - Holding time
testing
Spiked
effluent
•

•

•a
•
•

•

•c
•c
Spiked
receiving
water
•

•

•a
•
•

•

•
•
Part 4 - Definitive testing
Spiked
effluent
•
•
•
•
•b
•
•
•
•
•
•c
•c
Spiked
receiving
water
•
•
•
•
•b
•
•
•
•
•
•
•
Reference
toxicant
•
•
•
•
•b
•
•
•
•
•
•
•
a Test conducted with EDTA.
b Tests conducted with and without EDTA.
c Tests were conducted only on unspiked effluent samples.
                                                                     18

-------
Following chemical and physical characterization of the sample matrices, a single background definitive
test using each of the test species was conducted on a sample from each real-world source.  If historical
information (chemical analysis or toxicological analysis) on the real-world matrix source was available,
this information was evaluated along with results of background testing. Following completion of analysis
and historical data gathering, a final determination of real-world sample sources was made. The selection
criteria for real-world sample sources included the following elements:
•       Accessibility - Selected real-world sample sources were readily accessible to the referee
        laboratory. This included logistical accessibility as well as permission from the source provider to
        collect, test, and use the sample source in the WET Study
•       Historic testing and experience - It was important for referee laboratories to have significant
        experience in collecting and testing the selected real-world samples. This experience and
        knowledge of historic testing allowed referee laboratories to identify conditions or characteristics of
        the source that could potentially pose problems in the WET Study
•       Characterization - Selected real-world sample sources were well characterized by the referee
        laboratory through historic physical, chemical, and toxicological testing, as well as Part 1 of
        preliminary testing
•       Consistency - Selected real-world sample sources either provided a consistent level of toxicity or
        consistently produced no toxicity (in which case the sample could be spiked to achieve the desired
        effect level)

4.1.2   Part  2 - Range-finding

Part 2 of preliminary testing determined the range  of spiking concentrations necessary to achieve target
effect levels  for each  sample matrix.  This determination was critical to insuring that test concentrations
used in interlaboratory testing bracketed the effect concentrations (LC50,  IC25, IC50, and NOEC)
evaluated in  the WET Study.  Spiking of effluent and receiving water samples was targeted to produce
effect concentrations (LC50s for acute test methods and IC25s for chronic test methods) of 25% sample
during interlaboratory testing. Spiking of the reference toxicant sample was targeted to produce effect
concentrations of 50% sample. Spiking levels for the Selenastrum chronic test method were targeted to
produce  an IC50 of 38% sample during interlaboratory testing (see Section 4.3). Effluents and receiving
water samples were not spiked if they possessed persistent toxicity and produced effect concentrations near
the target effect level.  During Part 2 of preliminary testing, each matrix (effluent for the effluent sample,
receiving water for the receiving water sample, and synthetic dilution water for the reference toxicant
sample) was  spiked at a range of concentrations estimated to encompass the desired effect concentrations
(LCSOs for acute test methods and IC25s for chronic test methods). KC1 was used as  the spiking agent for
freshwater methods, sheepshead acute and chronic methods, and the mysid chronic test method; CuSO4 was
used as the spiking agent for silverside acute and chronic test methods, Champia chronic, and
Holmesimysis acute test methods. Preliminary spiking levels for Part 2 testing were determined from
referee laboratory reference toxicant testing databases, literature values, or range-finding tests conducted
just prior to Part 2 testing.
                                                 19

-------
Part 2 preliminary tests on spiked matrices were conducted as definitive tests according to the WET method
manuals and specific requirements of the WET Study plan (see Section 7). Part 2 testing was conducted
for each test method and each sample matrix. If the results of Part 2 testing were not conclusive or if they
differed greatly from historic reference toxicant testing conducted in the referee laboratory, Part 2 tests
were repeated.  Following Part 2 testing, appropriate spiking concentrations for interlaboratory testing were
estimated.  These spiking levels were estimated as the effect concentration determined in Part 2 testing
divided by 25% (0.25) for effluent and receiving water samples and divided by 50% (0.5) for the reference
toxicant sample. For example, if Part 2  testing for a given method determined an IC25 of 100 mg/L KC1 in
the effluent matrix, the final spiking concentration for the effluent sample type should be 400 mg/L KC1.
This final spiking concentration was determined by dividing the Part 2 IC25 value (100 mg/L) by the target
effect level of 25% (0.25) sample, or in  effect multiplying the Part 2 IC25 value by 4 (i.e., 1/0.25) to obtain
400 mg/L KC1. When the final sample spiked at 400 mg/L KC1 is diluted during interlaboratory testing
using the standard 0.5 dilution factor (i.e., 100%, 50%, 25%, 12.5%, and 6.25% sample), the 25% sample
test concentration should contain 100 mg/L KC1, and the test IC25 should be near the targeted 25% sample
range.

4.1.3    Part 3 - Holding Time Testing

Part 3 of preliminary testing determined the persistence of toxicity in the real-world samples. Excess
volume of the spiked real-world samples was retained from Part 2 testing and stored in the dark at 4°C.
Following storage for seven days, a second test was conducted using the stored  sample, and results were
compared to that of the initial test.  If effluent or receiving water samples were not spiked, Part 3 testing
was conducted on excess unspiked real-world sample collected in Part 1. Part 3 testing was  conducted on
effluent and receiving water samples for each test method. For acute and chronic test methods using the
same species, Part 3 testing was conducted using only the acute test method. The results of holding time
testing provided valuable information on the persistence of sample toxicity.  This information was useful in
the timing and scheduling of referee laboratory sample preparation for interlaboratory testing. This
information also supported the assumption that sample toxicity remained constant and test results were
valid even if participant  laboratories were unable to conduct tests on the scheduled day of testing due to
sample shipment problems or laboratory scheduling or culture problems.

4.1.4    Part 4 - Definitive Testing

Part 4 of preliminary testing validated that the samples and spiking concentrations (if applicable) selected
for the WET Study achieved the desired range  of effect concentrations following sample preparation,
shipping, and handling.  Part 4 testing also served as a trial run for sample collection, preparation,
packaging, and shipment in the interlaboratory testing phase. Each sample type that was used in
interlaboratory testing was collected, prepared, packaged, and shipped exactly as described for
interlaboratory testing (see Section 6). The samples were shipped by the referee laboratory round-trip back
to the referee laboratory. Upon receipt,  the referee laboratory conducted definitive WET tests as described
for interlaboratory testing (see Section 7).  Spiking concentrations for Part 4 testing were determined from
                                                20

-------
the results of Part 2 testing with necessary adjustments to meet the target effect levels.  Part 4 testing was
used to determine if the selected spiking concentrations achieved the targeted effect levels following sample
preparation and shipment.  Spiking levels that produced effect concentrations within the range of 10-35%
for effluent and receiving water samples or 35-60% for reference toxicant samples were considered    *
appropriate for use in the interlaboratory testing phase.  During interlaboratory testing, referee laboratories
again shipped samples round-trip back to themselves and conducted testing simultaneously with participant
laboratories.

4.2     Preliminary Testing Results

4.2.1    Preliminary Testing for Ceriodaphnia Acute and Chronic Test Methods

4.2.1.1  Reference toxicant sample type

The reference toxicant sample type was composed of moderately hard synthetic freshwater prepared
according to Section 7 of the WET method manuals (USEPA, 1994a) and spiked with potassium chloride
(KCI). The spiking level for this sample type was targeted to produce a test result (LC50 for
Ceriodaphnia acute test and IC25 for Ceriodaphnia chronic test) of 50% sample. Spiking levels for Part 2
preliminary testing were selected based on historical testing at the referee laboratory for the Ceriodaphnia
acute test method and from an initial Part 2 range-finding test for the Ceriodaphnia chronic test method.
Tables 4.2 and 4.3 show the results from preliminary testing for Ceriodaphnia acute and chronic test
methods, respectively. For the Ceriodaphnia acute test method,  Part 2 testing resulted in an LC50 of 424
mg/L KCI. Based on this result, the spiking level for Part 4 testing was increased to 850 mg/L KCI
(approximately 424 / the target result of 50% sample).  Part 4 Ceriodaphnia acute testing produced an
LC50 of 574 mg/L or 67.6% sample.  The spiking level for the interlaboratory testing phase (1000 mg/L
KCI) was based on an average of the LC50 results obtained in Part 2 and Part 4 testing (approximately the
average result of 500 / the target result of 50% sample). Referee laboratory testing on this interlaboratory
sample yielded LC50s of 40.6% and 34.4% sample.

Part 2 preliminary testing for the Ceriodaphnia  chronic test method produced IC25 values of 323 mg/L
KCI and 138 mg/L KCI in two separate tests. The referee laboratory selected a Part 4 spiking level of 650
mg/L so that the IC25 would potentially fall between 50% (if the true IC25 was closer to 323 mg/L KCI)
and 25% sample (if the true IC25 was closer to  138 mg/L KCI).  Results of this test revealed an IC25  of
132 mg/L KCI. The referee laboratory then repeated the Part 4 test with a slightly lower spiking level.
This test produced an IC25 of 134 mg/L KCI. Since three consecutive tests produced IC25 values between
132 and 138 mg/L KCL, the referee laboratory selected a final spiking level of 270 mg/L (the average
result of 135 / the target result of 50% sample) to achieve the target result of 50% sample in
interlaboratory testing.  Unfortunately, this spiking level did not produce the desired effect during
interlaboratory testing.  The resulting sample was marginally toxic and produced toxic results in only some
laboratories (see Section 5.3).
                                               21

-------
Table 4.2.  Results from Ceriodaphnia acute preliminary testing.
Sample type

Reference
toxicant


Effluent


Receiving
water

Part'
2
4
1L
2
3
4
IL
2
3
4
IT
Concentrations tested
(mg/L KC1)
37.5,75, 150,300,600
53, 106,212.5,425,850
62.5, 125,250,500, 1000
131,262.5,525, 1050,2100
131,262.5,525, 1050,2100
167.5, 335, 670, 1340, 2680
167.5,335,670, 1340,2680
106,212,424,848, 1696
106,212,424,848, 1696
125, 250, 500, 1000, 2000
1 1? 5, 29.5, 450, 900. 1800
NOAEC
(mg/L KC1)
300
425
250
525
525
670
335
424
424
250
450
LC50
(mg/L KC1)
424
574
406
651
689
948
670
526
487
365
554
LC50
(% sample)
70.7
67.6
40.6 & 34.4b
31.0
32.8
35.4
25.0
31.0
28.7
18.3
ms
a Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL). Part 1 preliminary
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute method.
b The referee laboratory tested two reference toxicant samples due to a sample distribution error (see Section 6.4).
Table 4.3.  Results from Ceriodaphnia chronic preliminary testing.
Sample
type
Reference
toxicant
Effluent
Receiving
water
Part"
2b
2
4
4C
IL
1
2
4
IL
1
2
4
4C
IT,
Concentrations tested
(mg/L KC1)
56, 100, 180,320,560
43.75,87.5, 175,350,700
40.5,81, 162.5,325,650
31.25,62.5, 125,250,500
17,34,67.5, 135,270
6.25, 12.5,25,50, 100d
87.5, 175, 350, 700, 1400
106.3,212.5,425,850, 1700
131,263,525, 1050,2100
6.25, 12.5,25,50, 100d
75, 150,300,600, 1200
92.5, 185,370,740, 1480
62.5, 125,250,500, 1000
75 150 300 600 1900
Survival
NOEC
(mg/L KC1)
320
175
325
125
270
100
350
425
263
100
300
185
250
SOO
Reproduction
NOEC
(mg/L KC1)
320
87.5
81
125
270
100
350
425
263
100
300
92.5
250
300
IC25
(mg/L KC1)
323
138
132
134
>270
-
424
538
389
-
368
114
342
37?
IC25
(% sample)
57.7
19.8
20.3
26.9
>100
>100
30.3
31.6
18.5
>100
30.7
7.7
34.2
31 3
 Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute
b An initial Part 2 range-finding test  was conducted due to a lack of historical referee laboratory data
c Part 4 testing was repeated to confirm spiking levels.
d The units for test concentrations in Part 1 were percent sample.
(IL). Part 1 preliminary
method.
for this method and toxicant.
                                                         22

-------
4.2.1.2 Effluent sample type

The effluent sample type was composed of a municipal wastewater treatment plant effluent spiked with KCI.
The effluent source was selected based on historical consistency in chemical and toxicological testing
conducted by the referee laboratory.  This same effluent source was used for all freshwater methods, the mysid
chronic test method, and the sheepshead acute and chronic test methods.  The municipal wastewater treatment
plant is designed to treat 180 mgd, is able to handle peak flows of 400 mgd, and currently treats 140 to 150
mgd. The facility employs tertiary treatment for biological nutrient removal including single-stage
nitrification/denitrification, sand filtration, chlorination/dechlorination, and anaerobic digestion. Water
chemistry of the effluent on each sample collection date is shown in Table 4.4. More detailed chemical
analyses (including total dissolved solids, total suspended solids, total organic carbon, biological oxygen
demand, and chemical oxygen demand) was performed at the beginning and end of preliminary testing.
Historically, the effluent has demonstrated low to no acute or chronic toxicity to freshwater organisms. Part 1
preliminary testing confirmed this consistency (Table 4.3).

For Part 2 of preliminary testing, the effluent sample was spiked with KCI to provide a consistently toxic
sample. The spiking level for this sample type was targeted to produce a test result (LC50 for Ceriodaphnia
acute test and IC25 for Ceriodaphnia chronic test) of 25% sample. Part 2 preliminary testing for the
Ceriodaphnia acute test method produced an LC50 of 651 mg/L KCI. This sample was held for 7 days at
4°C and then retested for Part 3 preliminary testing.  After holding, the sample produced an LC50 of 689
mg/L, which represents only a 5.8% change from the original Part 2 test result.  This indicated that the
toxicity of the sample was persistent and appropriate for use in the  WET Study. The spiking level for Part 4
preliminary testing (2680 mg/L KCI) was based on an average of the LC50 results obtained in Part 2 and Part
3 testing  (the average result of 670 / the target result of 25% sample).  Part 4 testing produced an LC50 of
948 mg/L or 35.4%  sample. The same spiking level was used for the interlaboratory sample. Referee
laboratory testing of the interlaboratoryt sample produced an LC50 of 25.0% sample.

For the Ceriodaphnia chronic test method, Part  1 testing on the unspiked effluent resulted in no toxicity (IC25
> 100% sample).  Spiking of the effluent sample in Part 2 testing produced an IC25 of 424 mg/L KCI. Based
on this result, the spiking level was increased to 1700 mg/L KCI (approximately 424 / the target result of 25%
sample) for Part 4 testing.  Part 4 testing produced an IC25 of 538 mg/L or 31.6% sample. Based on this
result, the spiking level used for interlaboratory testing was further increased to 2100 mg/L (approximately
538 / the target  result of 25% sample). The interlaboratory sample yielded an IC25 of 18.5% sample in
referee laboratory testing.
                                                23

-------
Table 4.4.  Water chemistry of effluent sample source for freshwater test methods.
Parameters
Alkalinity
(mg/L as CaCO3)
Hardness (mg/L as
CaC03)
Conductivity (,uS/cm)
pH
Temperature (°C)
Total residual chlorine
(mg/L)
Dissolved oxygen
(mg/L)
Total ammonia (mg/L)
Total dissolved solids
(mg/L)
Total suspended solids
(mg/L)
Total organic carbon
(mg/L)
Biological oxygen
demand (mg/L)
Chemical oxygen
demand (mg/L)
Sampling date"
08/11/99
76
148
828
7.4
28.1
O.01
7.5
1.17
420
7.0
21.4
3.8
59.9
08/19/99
62
160
933
7.1
26
<0.01
7.2
0.284
-
-
-
-
-
08/24/99
68
160
894
7.2
25.7
<0.01
7.8
0.122
-
-
-
-
-
09/07/99
78
172
909
7.2
23.9
<0.01
7.6
0.191
-
-
-
-
-
09/20/99
68
156
728
7.5
23.3
<0.01
8.1
-
-
-
-
-
-
11/29/99
54
152
695
6.8
12.9
O.01
8.9
0.0371
-
-
-
-
-
12/07/99
38
148
675
6.6
17.6
-
7.7
0.024
-
-
-
-
-
12/10/99
70
128
766
7.3
17.0
-
8.9
0.079
-
-
-
-
-
01/06/00
-
-
798
7.1
15.6
O.01
9.1
-
-
-
-
-
-
01/24/00
76
156
959
7.7
11.9
-
9.5
-
-
-
-
-
-
02/08/00
82
172
933
7.4
10.9
<0.01
10.3
1.88
626
<2.5
26.6
4.8
77.2
 "-" indicates that the parameter was not tested on the given sampling date.
                                                                     24

-------
4.2.1.3 Receiving water sample type

The receiving water sample type was composed of a natural surface water spiked with KC1.  The receiving
water was collected from the Gunpowder River, in Baltimore County, Maryland. Initial samples were
collected from a location near Bunker Hill Road.  Results of preliminary testing indicated that these unspiked
samples occasionally showed toxicity to fathead minnows and Selenastrum.  To avoid the potential problems
associated with intermittent ambient toxicity, subsequent samples were collected from a new location (near
Falls Road), upstream from potential sources of non-point source runoff. The results from toxicity tests
conducted with the Falls Road samples indicated no acute or chronic toxicity to Ceriodaphnia (Table 4.3) or
other freshwater species. Water chemistry of the receiving water is shown in Table 4.5 for each sample
collection date.

The receiving water sample was spiked with KG to provide a consistently toxic sample.  The spiking level for
this sample type was targeted to provide a test result (LC50 for Ceriodaphnia acute test and IC25 for
Ceriodaphnia  chronic test) of 25% sample.  Part 2 preliminary testing for the Ceriodaphnia acute test method
produced an LC50 of 526 mg/L KC1.  This sample was held for 7 days at 4°C and then retested for Part 3
preliminary testing. After holding, the sample produced an LC50 of 487 mg/L, which represents only a 7.4%
change from the original Part 2 test result. This indicated that the toxicity of the sample was persistent and
appropriate for use in the WET Study.  For Part 4 preliminary testing, the referee laboratory increased the
spiking level to 2000 mg/L to better achieve the target result of 25% sample. Part 4 testing produced an LC50
of 365 mg/L or 18.3% sample.  The spiking level for the interlaboratory sample (1800 mg/L KC1), was based
on an average  of the LC50 results obtained in Part 2, 3, and 4 testing (approximately the average result of 460
/ the target result of 25% sample). Testing of this sample produced an LC50 of 30.8% sample in referee
laboratory testing.

For the Ceriodaphnia chronic test method, Part 1  testing on the unspiked receiving water resulted in no
toxicity (IC25  > 100% sample).  Spiking of the receiving water sample in Part 2 testing produced an IC25 of
368 mg/L KC1. Based on this result, the spiking level was increased to 1480 mg/L KC1 for Part 4 testing
(approximately 368 / the target result of 25% sample).  Part 4 testing produced an IC25 of 114 mg/L or 7.7%
sample.  Due to the discrepancy between the results of Part 2 and Part 4 testing, Part 4 testing was repeated.
Repeated Part  4 testing produced an IC25 of 342 mg/L KC1, which was consistent with Part 2 testing results.
Final spiking levels were based on these tests and set at 1200 mg/L for the interlaboratory testing phase.
Referee laboratory testing of the interlaboratory sample produced an IC25 of 31.3% sample.
                                                25

-------
Table 4.5. Water chemistry of receiving water sample source for freshwater test methods.
Parameters
Alkalinity
(mg/L as CaCO3)
Hardness (mg/L as CaCO3)
Conductivity (^S/cm)
pH
Temperature (°C)
Total residual chlorine
(mg/L)
Dissolved oxygen (mg/L)
Total ammonia (mg/L)
Total dissolved solids
(mg/L)
Total suspended solids
(mg/L)
Total organic carbon
(mg/L)
Biological oxygen demand
(mg/L)
Chemical oxygen demand
(mg/L)
Sampling date"
06/24/99
30
80
183
7.7
14.1
<0.01
10.4
0.056
94.5
5.5
1.3
<1.0
12.8
07/07/99
28
60
154
7.9
15.0
O.01
9.6
0.096
-
-
-
-
-
07/14/99
30
60
148
6.8
16.3
<0.01
10.3
-
-
-
-
-
-
07/28/99
24
60
159
7.9
18.3
<0.01
9.8
-
-
-
-
-
-
08/05/99
28
48
161
7.9
16.6
O.01
10.7
0.092
-
-
-
-
-
08/20/99
34
44
155
7.3
16.5
<0.01
10.9
0.307
-
-
-
-
-
09/21/99
34
60
157
7.5
14.6
<0.01
10.0
-
-
-
-
-
-
12/21/99
26
36
139
8.3
7.8
-
12.6
-
-
-
-
-
-
01/31/00
28
56
124
7.8
2.1
-
14.6
-
-
-
-
-
-
 "-" indicates that the parameter was not tested on the given sampling date.
                                                                     26

-------
4.2.2    Preliminary Testing for Fathead Minnow Acute and Chronic Test Methods

4.2.2.1  Reference toxicant sample type

The reference toxicant sample type was composed of moderately hard synthetic freshwater prepared according
to Section 7 of the WET method manuals (USEPA, 1994a) and spiked with KCI.  The spiking level for this
sample type was targeted to produce a test result (LC50 for fathead acute test and IC25 for fathead chronic
test) of 50% sample. Tables 4.6 and 4.7 show the results from preliminary testing using fathead acute and
chronic  test methods, respectively. Spiking levels selected for Part 2 preliminary testing resulted in an LC50
of 915 mg/L KCI for the fathead acute test method. Based on this result, the spiking level for Part 4 testing
was increased to 1830 mg/L KCI (915 / the target result of 50% sample). Part 4 fathead acute testing
produced an LC50 of 1167 mg/L or 63.8% sample. Spiking levels were further increased to 2200 mg/L KCI
for the interlaboratory testing phase to better achieve the target result of 50% sample.  Referee laboratory
testing of the interlaboratory sample produced an LC50 of 42.4% sample.

Part 2 preliminary testing using the fathead chronic test method produced an IC25 of 545 mg/L KCI. Based on
this result, the spiking level for Part 4 testing was decreased to 1090 mg/L KCI (545 / the target result of 50%
sample). Part 4 testing yielded an IC25 of 610 mg/L KCI.  Based on Part 4 testing results, the  spiking level
for the interlaboratory sample was further increased to 1220 mg/L KCI  (610 / the target result  of 50%
sample). Referee laboratory testing of the interlaboratory sample yielded an IC25 of 63.3% sample.

Table 4.6.  Results from fathead acute preliminary testing.
Sample type
Reference toxicant
Effluent
Receiving water
Part"
2
4
IL
2
3
4
IL
2
3
4
IL
Concentrations tested
(mg/L KCI)
104.9,209.8,419.5,839, 1678
114.5,229,457.5,915, 1830
138,275,550, 1100,2200
300,600, 1200,2400,4800
300, 600, 1200, 2400, 4800
350, 700, 1400, 2800, 5600
333, 666, 1332, 2664, 5328
250, 500, 1000, 2000, 4000
250,500, 1000,2000,4000
300, 600, 1200, 2400, 4800
313,625, 1250,2500,5000
NOAEC
(mg/L KCI)
419.5
915
550
600
600
700
666
1000
500
600
625
LC50
(mg/L KCI)
915
1167
924
1308
1356
990
1028
1270
1168
1256
985
LC50
(% sample)
54.5
63.8
42.4
27.3
28.3
17.7
19.3
31.7
29.2
26.2
19.7
a Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL). Part 1 preliminary
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute method.
                                                 27

-------
Table 4.7. Results from fathead chronic preliminary testing.
Sample type

Reference
toxicant


Fffln^nt




Receiving
water


Part"
2
4
IL
1
2
4
IL
lc
1"
2

4
IL
Concentrations tested
(mg/L KC1)
88.4, 176.8,353.5,707, 1414
68, 136, 272.5, 545, 1090
76, 153,305,610, 1220
6.25, 12.5, 25, 50, 100"
144,287.5,575, 1150,2300
181.3,362.5,725, 1450,2900
225,450,900, 1800,3600
6.25, 12.5,25,50, 100b
6.25, 12.5,25,50, 100b
136,272.5,545, 1090,2180

140,280,560, 1120,2240
150,300,600, 1200,2400
Survival
NOEC
(mg/L KC1)
353.5
545
610
100
575
725
900
100
100
545

560
600
Reproduction
NOEC
(mg/L KC1)
353.5
545
610
100
575
725
900
50
100
545

560
600
IC25
(mg/L KC1)
545
610
772
-
721
901
968
-
-
566

606
708
IC25
(% sample)
38.5
56.0
63.3
>100
31.4
31.1
26.9
5.6
>100
26.0

27.0
29.5
 Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL). Part 1 preliminary
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute method.
b The units for test concentrations in Part 1 were percent sample.
c Receiving water sample was collected from Bunker Hill Road site.
d Receiving water sample was collected from Falls Road site.
4.2.2.2  Effluent sample type

The effluent sample source used for the fathead acute and chronic test methods was the same as described in
Section 4.2.1.2.  This effluent sample was spiked with KC1 to provide a consistently toxic sample.  The
spiking level for this sample type was targeted to produce a test result (LC50 for fathead acute test and IC25
for fathead chronic test) of 25% sample. Part 2 preliminary testing for the fathead acute test method produced
an LC50 of 1308 mg/L KC1.  This sample was held for 7 days at 4°C and then retested for Part 3 preliminary
testing.  After holding, the sample produced an LC50 of  1356 mg/L, which represents only a 3.7% change
from the original Part 2 test result.  This indicated that the toxicity of the sample was persistent and
appropriate for use in the WET Study.  For Part 4 preliminary testing, the referee laboratory increased the
spiking level to 5600 mg/L KC1 to better achieve the target effect level of 25% sample. This test produced an
LC50 of 990 mg/L or 17.7% sample, however, less than  90% survival was  experienced in the control. Since
Part 4 testing was unreliable, the spiking level for interlaboratory testing was based on an average of Part 2
and Part 3 testing results and set at 5328 mg/L KC1 (the  average result of 1332 / the target result of 25%
sample). This interlaboratory sample yielded an LC50 of 19.3% sample in referee laboratory testing.

For the  fathead chronic test method, Part 1 testing of the  unspiked effluent resulted in no  toxicity (IC25 >
100% sample). Spiking of the effluent sample in Part 2 testing produced an IC25 of 721  mg/L KC1.  Based
                                                  28

-------
on this result, the spiking level was increased to 2900 mg/L KC1 for Part 4 testing (approximately 721 / the
target result of 25% sample). Part 4 testing produced an IC25 of 901 mg/L or 31.1% sample. Based on Part
4 testing, the spiking level for interlaboratory testing was further increased to 3600 mg/L (approximately 901 /
the target result of 25% sample).  This interlaboratory sample yielded an IC25 of 26.9% sample in referee
laboratory testing.

4.2.2.3 Receiving water sample type

The receiving water sample source used for the fathead acute and chronic test methods was the same as
described in Section 4.2.1.3.  The receiving water sample was spiked with KC1 to provide a consistently toxic
sample. The spiking level for this sample type was targeted to provide a test result (LC50 for fathead acute
test and IC25 for fathead chronic  test) of 25% sample.  Part 2 preliminary testing for the fathead  acute test
method produced an LC50 of 1270 mg/L KC1. This sample was held for 7 days at 4°C and then  retested for
Part 3 preliminary testing.  After holding, the sample produced an LC50 of 1168 mg/L, which represented
only a 8.7% change from the original Part 2 test result.  This indicated that the toxicity of the sample was
persistent and appropriate for use in the WET Study. The spiking level for Part 4 preliminary testing (4800
mg/L KC1) was based on an average of the  LC50 results obtained in Part 2 and Part 3 testing (approximately
the average result of 1219 / the target result of 25% sample). Part 4 testing produced an LC50 of 1256 mg/L
or 26.2% sample.  Based on this result, the spiking level was further increased to 5000 mg/L (approximately
1256 / the target result of 25% sample) for the interlaboratory testing phase. Referee laboratory testing of the
interlaboratory sample yielded an LC50 of 19.7% sample.

For the fathead chronic test method, Part 1  testing on the unspiked receiving water collected from the Bunker
Hill Road site indicated toxicity (IC25 of 5.6% sample). Following this test, the referee  laboratory moved the
receiving water collection site farther upstream to the Falls Road site. Part 1 testing on receiving water from
the new location revealed no toxicity (IC25 >100% sample). Part 2 testing on the spiked receiving water
produced an IC25 of 566 mg/L KC1. Based on this result, the spiking level was increased to 2240 mg/L KC1
for Part 4 testing (approximately  566 / the target result of 25% sample).  Part 4 testing produced an IC25  of
606 mg/L or 27.0% sample.  Based on this result, the final spiking  level for interlaboratory testing was
increased to 2400 mg/L (approximately 606 / the target result of 25% sample). This interlaboratory sample
yielded an IC25 of 29.5% sample in referee laboratory testing.

4.2.3   Preliminary Testing for the Selenastrum Chronic Test Method

4.2.3.1 Reference toxicant sample type

The reference toxicant sample type was composed of deionized water spiked with KC1.  The spiking level for
this sample type was targeted to produce an IC50 of 38% sample (see Section 4.3).  Table 4.8 shows the
results from preliminary testing for the Selenastrum chronic test method.  Spiking levels for Part  2 preliminary
testing were based on the results of an initial Part 2 range-finding test.  The range-finding test resulted in an
IC50 of 925 mg/L KC1.  Spiking levels selected for Part 2 testing resulted in an IC50 of 2925 mg/L KC1 for
                                                29

-------
the Selenastrum chronic test method. Based on this result, the spiking level for Part 4 testing was increased
to 7888 mg/L KC1. Part 4 testing with EDTA produced an IC50 of 1713 mg/L (or 57.1% sample) and Part 4
testing without EDTA produced an IC50 of 2943 mg/L (or 98.1% sample). Due to the discrepancy between
Part 4 results with EDTA and Part 2 results with EDTA, Part 4 testing was repeated. In additional Part 4
testing, IC50s of 1808 mg/L (or 22.9% sample) and 462 mg/L (or 5.9% sample) were produced with EDTA
and without EDTA, respectively. The spiking level for interlaboratory testing was set at 5655 mg/L (average
result of 2149 / the target result of 38% sample) based on an average of Part 2 and Part 4 testing with EDTA.
Referee laboratory testing of the interlaboratory sample yielded IC50s of 35.5% sample and 37.6% sample
with  and without EDTA, respectively.

4.2.3.2 Effluent sample type

The effluent sample source used for the Selenastrum chronic test method was the same as described in Section
4.2.1.2. Part 1 testing on the unspiked effluent resulted in no toxicity (IC50 > 100% sample). The effluent
sample was then spiked with KC1 to provide a consistently toxic sample. The spiking level for this sample
was targeted to produce a test result (IC50) of 38% sample.  Part 2 preliminary testing for the Selenastrum
chronic test method produced an IC50 of 4383 mg/L KC1. This sample was held for 7 days at 4°C and then
retested for Part 3 preliminary testing. After holding, the sample produced an IC50  of 5143 mg/L, which
represents a 17% difference from the initial Part 2 test result. The toxicity of the sample was determined to be
reasonably persistent and appropriate for use in the interlaboratory study. Based on Part 2 testing results, the
spiking level for Part 4 testing was set at 11540 mg/L KC1 (approximately 4383 / the target result of 38%
sample).  Part 4 testing resulted in an IC50 of 4609 mg/L or 39.9% sample with EDTA and an IC50 of 4821
mg/L or 41.8% sample without EDTA. Since Part 4 testing results were very close to the 38% sample target,
the same spiking level was maintained for  interlaboratory testing.  Referee laboratory results on the
interlaboratory sample were 29.9% sample with EDTA and  19.8% sample  without EDTA.
                                               30

-------
Table 4.8.  Results from Selenastrum chronic preliminary testing.
Sample type
Reference toxicant
Effluent
Receiving water
Part"
2b
2
4w/EDTAc
4 w/EDTA
4w/oEDTA
4 w/o EDTAC
IL w/ EDTA
IL w/o EDTA
1
2 w/ EDTA
3 w/ EDTA
4 w/ EDTA
4 w/o EDTA
IL w/ EDTA
IL w/o EDTA
lc
lf
2 w/ EDTA
2 w/ EDTA6
2 w/ EDTA£
3 w/ EDTA
4 w/ EDTA
4 w/o EDTA
IL w/ EDTA
IL w/o EDTA
Concentrations tested
(mg/L KC1)
100, 1000, 10000, 30000, 60000
375,750, 1500,3000,6000
188,375,750, 1500,3000
493,986, 1972,3944,7888
188,375,750, 1500,3000
493, 986, 1972, 3944, 7888
353, 707, 1414, 2828, 5655
353,707, 1414,2828,5655
6.25, 12.5,25,50, 100d
938, 1875,3750,7500, 15000
938, 1875,3750,7500, 15000
721, 1443,2885,5770, 11540
721, 1443,2885,5770, 11540
721, 1443,2885,5770, 11540
721, 1443,2885,5770, 11540
6.25, 12.5, 25, 50, 100"
50, 100d
225,450,900, 1800,3600
1000, 2000, 4000, 8000, 16000
500, 1000, 2000, 4000, 8000
500, 1000,2000,4000,8000
807, 1614,3228,6456, 12912
807, 1614,3228,6456, 12912
732, 1464,2928,5857, 11713
732, 1464,2928,5857, 11713
NOEC
(mg/L KC1)
100
375
375
986
187.5
<493
1414
1414
100
937.5
937.5
2885
2885
2885
1443
100
100
3600
1000
<500
1000
3228
1624
2928
2928
IC50
(mg/L KC1)
925
2925
1713
1808
2943
462
2007
2126
-
4383
5143
4609
4821
3502
2319
-
-
>3600
4885
4399
4928
4069
2098
3093
4201
IC50
(% sample)
1.5
48.7
57.1
22.9
98.1
5.9
35.5
37.6
>100
29.2
34.3
39.9
41.8
29.9
19.8
97
>100
>100
30.5
55.0
61.6
31.5
16.2
26.8
36.4
 Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL).
b An initial Part 2 rangefinding test was conducted due to a lack of historical referee laboratory data for this method and toxicant.
c Part 4 testing was repeated to confirm spiking levels.
d The units for test concentrations in Part 1 were percent sample.
0 Receiving water sample was collected from Bunker Hill Road site.
 Receiving water sample was collected from Falls Road site.
8 Part 2 testing was repeated to confirm spiking levels.
4.2.3.3 Receiving water sample type


The receiving water sample source used for the Selenastrum chronic test method was the same as described in
Section 4.2.1.3.  Part 1 testing on the unspiked receiving water collected from the Bunker Hill Road site
indicated toxicity (IC50 of 97% sample). Following this test, the referee laboratory moved the receiving water
collection site farther upstream to the Falls Road site. Part 1 testing on receiving water from this site revealed
no toxicity (IC50 >100% sample).  The receiving water sample was then spiked with KC1 to provide a
                                                     31

-------
consistently toxic sample. The spiking level for this sample was targeted to produce a test result (IC50) of
38% sample.  Part 2 preliminary testing for the Selenastrum  chronic test method produced an IC50 of >3600
mg/L KCI. Since the test result was outside of the concentration range tested, Part 2 testing was repeated.
Two additional Part 2 tests produced IC50s of 4885 and 4399 mg/L. The sample from the latter Part 2 test
was held for 7 days at 4°C and then retested for Part 3 preliminary testing. After holding, the sample
produced an IC50 of 4928 mg/L, which represents only a 12% difference from the original Part 2 test result.
The toxicity of the sample was determined to be reasonably persistent and appropriate for use in the WET
Study. Based on an average of results from the second Part 2 test and Part 3 testing, the spiking level for Part
4 testing was set at 12912 mg/L  KCI (average result of 4906 / the target result of 38% sample).  Part 4 testing
produced ICSOs of 4069 (or 31.5% sample) and 2098 mg/L KCI (or 16.2% sample) with and without EDTA,
respectively.  The spiking level for interlaboratory testing was based on an average of results from Part 2 and
Part 4 testing and set at 11713 mg/L KCI (average result of 4451 / the target result of 38% sample). Referee
laboratory results on the interlaboratory sample yielded ICSOs of 26.8% sample with EDTA and 36.4%
sample without EDTA.

4.2.4    Preliminary Testing for the Mysid Chronic Test  Method

4.2.4.1  Reference toxicant sample type

The reference toxicant sample type was composed of synthetic seawater (prepared using Forty Fathoms
artificial sea salts added to deionized water) spiked with  KCI.  The spiking level for this sample was targeted
to produce a test result (IC25) of 50% sample. Spiking levels for Part 2 preliminary testing were selected
based on historical testing at the  referee laboratory for the mysid chronic test method. Table 4.9 shows the
results from preliminary testing for the mysid chronic test method.  Part 2 preliminary testing resulted in an
IC25 of 426 mg/L KCI.  Based on this result, the spiking level for Part 4 testing was increased to 900 mg/L
KCI (approximately 426 / the target result of 50% sample). Part 4 testing resulted in an IC25 of 530 mg/L
(58.9% sample).  Based on this result, the spiking level for interlaboratory testing was further increased to
1200 mg/L KCI. Referee laboratory testing of the interlaboratory sample resulted in an IC25 of 36.4%
sample.

4.2.4.2  Effluent sample type

The effluent sample source used  for the mysid chronic test method  was the same as described in Section
4.2.1.2.  The salinity of the effluent was adjusted to 25 ppt prior to the conduct of marine tests. Part 1 testing
on the unspiked effluent resulted in an IC25 of 64.4% sample. The effluent sample was then spiked with KCI
to provide a consistently toxic sample. The spiking level for this sample type was targeted to produce a test
result (IC25) of 25% sample.  Part 2 preliminary testing  for the mysid chronic test method produced an IC25
of 486 mg/L KCI. This sample was held for 7 days at 4°C and then retested for Part 3 preliminary testing.
After holding, the sample produced an IC25 of 420 mg/L, which represented a  14% difference from the
original  Part 2 test result. The toxicity of the sample was determined to be reasonably persistent and
appropriate for use  in the WET Study. Based on the result of Part 2 testing, the spiking level for Part 4
                                                32

-------
testing was increased to 1960 mg/L (approximately 486 / the target result of 25% sample).  Part 4 testing
produced an IC25 of 521 mg/L KC1 (26.6% sample).  Based on an average of results from Part 2 and Part 4
testing, the spiking level for interlaboratory testing was increased to 2000 mg/L KC1 (approximately the
average result of 504 / the target result of 25% sample). Referee laboratory testing of the interlaboratory
sample resulted in an IC25 of 29.9% sample.

Table 4.9.  Results from mysid chronic preliminary testing.
Sample type
Reference
toxicant
Effluent
Receiving
water
Part"
2
4
IL
1
2
3
4
IL
1
2
3
4
IL
Concentrations tested
(mg/L KC1)
45.25,90.5, 181,362,724
56.3, 113,225,450,900
75, 150,300,600, 1200
6.25, 12.5,25,50, 100b
90, 180,360,720, 1440
90, 180,360,720, 1440
122.5,245,490,980,1960
125,250,500, 1000,2000
6.25, 12.5, 25, 50, 100b
90, 180,360,720, 1440
90, 180,360,720, 1440
122.5,245,490,980,1960
150, 300, 600, 1200, 2400
Survival
NOEC
(mg/L KC1)
362
450
300
100
360
360
490
500
100
360
720
490
300
Growth
NOEC
(mg/L KC1)
362
450
300
50
360
360
490
500
50
360
360
490
300
IC25
(mg/L KC1)
426
530
437
-
486
420
521
598
-
486
634
608
564
IC25
(% sample)
58.8
58.9
36.4
64.4
33.7
29.2
26.6
29.9
88
33.8
44.0
31.0
23.5
a Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL).
b The units for test concentrations in Part 1 were percent sample.
4.2.4.3  Receiving water sample type

The receiving water sample type was composed of natural seawater spiked with KC1. Receiving water was
collected from Manasquan Inlet, in Manasquan, Monmouth County, New Jersey.  Seawater from this location
has historically been non-toxic to the test species, and is currently used by EPA's Division of Environmental
Science and Assessment as dilution water for toxicity testing.  Water chemistry of this receiving water on each
sample collection date is shown in Table 4.10.

The receiving water sample was filtered and adjusted to a salinity of 25 ppt prior to toxicity testing. Part 1
testing on the unspiked receiving water sample indicated moderate toxicity (IC25 of 88% sample).  The
receiving water sample was then spiked with KC1 to provide a consistently toxic sample. The spiking level for
this sample type was targeted to produce a test result (IC25) of 25% sample.  Part 2 preliminary testing for the
mysid chronic test method produced an IC25 of 486 mg/L KC1.  This sample was held for 7 days at 4°C and
                                                 33

-------
then retested for Part 3 preliminary testing. After holding, the sample produced an IC25 of 634 mg/L, which
represented a 30% change from the initial Part 2 test result. For Part 4 preliminary testing, the spiking level
was increased to 1960 mg/L KC1 (approximately 486 / the target result of 25% sample) based on Part 2
testing results. Part 4 testing resulted in an IC25 of 608 mg/L. Based on this result, the spiking level was
further increased to  2400 mg/L KC1 (approximately 608 / the target result of 25% sample) for the
interlaboratory testing phase. Referee laboratory testing of the interlaboratory sample resulted in an IC25 of
23.5% sample.

Table 4.10.  Water chemistry  of receiving water sample source for mysid chronic and sheepshead acute
and chronic test methods.
Parameters
Alkalinity (mg/L as CaCO3)
pH
Temperature (°C)
Total residual chlorine (mg/L)
Dissolved oxygen (mg/L)
Salinity (ppt)
Copper 0/g/L)
Total ammonia (mg/L)
Total dissolved solids (mg/L)
Total suspended solids (mg/L)
Total organic carbon (mg/L)
Biological oxygen demand (mg/L)
Chemical oxygen demand (mg/L)
Sampling date8
11/03/99
98
8.0
3.5
<0.01
-
28.3
<10
0.980
33,300
10.5
<1.0
2.1
854
11/23/99
110
8.2
3.9
-
10.9
31.7
-
-
-
-
-
-
-
12/29/99
-
8.3
2.5
-
9.2
35
-
-
-
-
-
-
-
01/13/00
-
7.5
10.6
-
10.1
31.9
-
-
-
-
-
-
-
a "-" indicates that the parameter was not tested on the given sampling date.

4.2.5    Preliminary Testing for Sheepshead Minnow Acute and Chronic Test Methods

4.2.5.1  Reference toxicant sample type

The reference toxicant sample type was composed of synthetic seawater (prepared using Forty Fathoms
artificial sea salts added to deionized water) spiked with KC1. The spiking level for this sample type was
targeted to produce a test result (LC50 for sheepshead acute test and IC25 for sheepshead chronic test) of
50% sample.  Tables 4.11 and 4.12 show the preliminary testing results for sheepshead acute and chronic test
methods, respectively.  Spiking levels selected for Part 2 testing resulted in an LC50 of 1580 mg/L KC1 for the
sheepshead acute test method. Based on this result, the spiking level for Part 4 testing was increased to 3160
mg/L KC1 (1580 / the target result of 50% sample). Part 4 sheepshead acute testing produced an LC50 of
                                                34

-------
1157 mg/L or 36.6% sample.  The same spiking level was used for the interlaboratory sample, which
produced an LC50 of 40.6% sample in referee laboratory testing.

Part 2 preliminary testing for the sheepshead chronic method produced an IC25 of 1257 mg/L KC1.  Based on
this result, the spiking level for Part 4 testing was increased to 2600 mg/L KC1 (approximately 1257 / the
target result of 50% sample).  Results of this test revealed an IC25 of 1528 mg/L KC1.  Based on Part 4
testing results, the spiking level for interlaboratory testing was further increased to 3000 mg/L KC1
(approximately  1528 / the target result of 50% sample).  Referee laboratory testing of the interlaboratory
sample yielded an IC25 of 54.3% sample.

Table 4.11. Results from sheepshead acute preliminary testing.
Sample
type
Reference
toxicant
Effluent
Receiving
water
Part8
2
4
IL
2
3
4
IL
2
3
4
IL
Concentrations tested
(mg/L KC1)
155,310,620,1240,2480
198,395,790,1580,3160
198,395,790,1580,3160
310,620, 1240,2480,4960
310,620, 1240,2480,4960
325,650, 1300,2600,5200
325,650, 1300,2600,5200
310,620, 1240,2480,4960
310,620, 1240,2480,4960
398,795, 1590,3180,6360
350,700, 1400,2800,5600
NOAEC
(mg/L KC1)
1240
790
790
620
1240
650
1300
1240
620
795
700
LC50
(mg/L KC1)
1580
1157
1283
1329
1694
1172
1840
1580
1488
1247
1450
LC50
(% survival)
63.7
36.6
40.6
26.8
34.2
22.5
35.4
31.9
30.0
19.6
25.9
a Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL). Part 1 preliminary
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute method.
4.2.5.2  Effluent sample type

The effluent sample source used for the sheepshead acute and chronic test methods was the same as described
in Section 4.2.1.2.  This effluent sample was adjusted to a salinity of 25 ppt and spiked with KC1 to provide a
consistently toxic sample. The spiking level for this sample type was targeted to produce a test result (LC50
for sheepshead acute test and IC25 for sheepshead chronic test) of 25% sample. Part 2 preliminary testing for
the sheepshead acute test method produced an LC50 of 1329 mg/L KC1. This sample was held for 7 days at
4°C and then retested for Part 3 preliminary testing.  After holding, the sample produced an LC50 of 1694
mg/L, which represents a 27% change from the original Part 2 test result.  For Part 4 preliminary testing, the
referee laboratory increased the spiking level to 5200 mg/L KC1. This test produced an LC50 of 1172 mg/L
or 22.5% sample. The same spiking level was used for the interlaboratory sample, which produced an LC50
of 35.4% sample in referee laboratory testing.
                                                 35

-------
Table 4.12. Results from sheepshead chronic preliminary testing.
Sample
type

Reference
toxicant







Receiving
water

Part"
2
4
IL
1
2
4
IL
1

2
4
IL
Concentrations tested
(mg/L KC1)
98, 195,390,780, 1560
163,325,650, 1300,2600
188,375,750, 1500,3000
6.25, 12.5, 25, 50, 100b
250,500, 1000,2000,4000
275,550, 1100,2200,4400
300,600, 1200,2400,4800
6.25, 12.5,25,50, 100b

195,390,780, 1560,3120
253,505, 1010,2020,4040
275,550, 1100,2200,4400
Survival
NOEC
(mg/L KC1)
780
1300
1500
100
1000
1100
1200
100

780
1010
1100
Growth
NOEC
(mg/L KC1)
780
1300
1500
100
500
1100
1200
100

780
505
1100
IC25
(mg/L KC1)
1257
1528
1629
-
1158
1276
1406
.

1028
1172
1210
IC25
(% sample)
80.6
58.8
54.3
>100
28.9
29.0
29.3
>100

33.0
29.0
27.5
d Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL). Part 1 preliminary
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute method.
b The units for test concentrations in Part 1 were percent sample.
For the sheepshead chronic test method, Part 1 testing on the unspiked effluent resulted in no toxicity (IC25 >
100% sample). Spiking of the effluent sample in Part 2 testing produced an IC25 of 1158 mg/L KC1. Based
on this result, the spiking level was increased to 4400 mg/L KC1 for Part 4 preliminary testing (approximately
1158 / the target result of 25% sample).  Part 4 testing produced an IC25 of 1276 mg/L or 29.0% sample.
Based on the average of results from Part 2 and Part 4 testing, the spiking level for interlaboratory testing was
further increased to 4800 mg/L (approximately the average result of 1217 / the target result of 25% sample).
Referee laboratory testing of the interlaboratory sample yielded an IC25 of 29.3% sample.

4.2.5.3  Receiving water sample type

The receiving water sample source used for the sheepshead acute and chronic test methods was the same as
described in Section 4.2.4.3. The receiving water sample was filtered, adjusted to a salinity of 25 ppt, and
spiked with KC1 to provide a consistently toxic sample. The spiking level for this sample type was targeted to
provide a test result (LC50 for sheepshead acute test and IC25 for sheepshead chronic test) of 25% sample.
Part 2 preliminary testing for the sheepshead acute test method produced an LC50 of 1580 mg/L KC1. This
sample was held for 7 days at 4°C and then retested for Part 3 preliminary testing.  After holding, the sample
produced an LC50 of 1488 mg/L,  which represented  only a 5.8% change from the original Part 2 test result.
This indicated that the toxicity of the sample was persistent and appropriate for use in the WET Study. Based
on the results of Part 2 testing, the spiking level for Part 4 testing was set at 6360 mg/L KC1 (approximately
1580 / the target result of 25% sample).  Part 4 testing resulted in an LC50 of 1247 mg/L or 19.6% sample.
                                                 36

-------
Based on an average of Part 2 and Part 4 preliminary testing results, the spiking level for the interlaboratory
sample was reduced to 5600 mg/L (approximately the average result of 1414 / the target result of 25%
sample) for the interlaboratory testing phase. Referee laboratory testing of the interlaboratory sample yielded
an LC50 of 25.9% sample.

For the sheepshead chronic test method, Part 1 testing on the unspiked receiving water revealed no toxicity
(IC25 >100% sample). Part 2 testing on the spiked receiving water produced an IC25 of 1028 mg/L KC1.
Based on this result, the spiking level was increased to 4040 mg/L KC1 for Part 4 testing (approximately 1028
/ the target result of 25% sample). Part 4 testing produced an IC25 of 1172 mg/L or 29.0% sample. Based
on an average of test results from Part 2 and Part 4 testing, the spiking level for the interlaboratory sample
was set at 4400 mg/L (the average result of 1100 / the target result of 25% sample).  Referee laboratory
testing of the interlaboratory sample yielded an IC25 of 27.5% sample.

4.2.6  Preliminary Testing for Silverside Minnow Acute and Chronic Test Methods

4.2.6.1 Reference toxicant sample type

The reference toxicant sample type was composed of synthetic seawater (prepared using Forty Fathoms
artificial sea salts  added to deionized water) spiked with copper sulfate (CuSO4).  The spiking level for this
sample type was targeted to produce a test result (LC50 for silverside acute  test and IC25 for silverside
chronic test) of 50% sample.  Tables 4.13 and 4.14 show the results from preliminary testing for silverside
acute and chronic test methods, respectively. Spiking levels selected for Part 2 testing resulted in an LC50 of
>0.25 mg/L  Cu for the silverside acute test method.  Based on this result, the spiking level for Part 4 testing
was increased to 1 mg/L Cu.  Part 4 sheepshead acute testing produced an LC50 of 0.29 mg/L or 29%
sample.  The same spiking level was used for the interlaboratory sample, however test results on the
interlaboratory sample were >100% sample due to an error in the preparation of this sample type (see Section
5.3).

Part 2 preliminary testing for the silverside chronic method produced an IC25 of >0.3 mg/L Cu.  Based on this
result, the spiking level for Part 4 testing was increased to 1 mg/L Cu. Part  4 testing produced an IC25 of
0.189 mg/L Cu or 18.9% sample. The same spiking level was used for the interlaboratory sample, and referee
laboratory testing of this sample yielded an IC25 of 15.2% sample.

4.2.6.2 Effluent sample type

The effluent sample type was composed of an industrial wastewater effluent spiked with CuSO4. The effluent
source was selected based on historical consistency in chemical and toxicological testing conducted by the
referee laboratory. The industrial wastewater treatment facility that was used as an effluent source for the
silverside acute and chronic test methods was designed to treat wastes from  an oil refinery.  Water chemistry
from the effluent source is listed in Table 4.15.
                                                 37

-------
Table 4.13. Results from silverside acute preliminary testing.
Sample type
Reference toxicant
Effluent
Receiving water
Part"
2
4
IL
2
3
4
IL
2
3
4
IL
Concentrations tested
(mg/L Cu)
0.016,0.031,0.063,0.125,0.25
0.063,0.125,0.25,0.5, 1.000
0.063,0.125,0.25,0.5, 1.000
0.031,0.063,0.125,0.25,0.5
0.031,0.063,0.125,0.25,0.5
0.058,0.115,0.231,0.461,0.922
0.058,0.115,0.231,0.461,0.922
0.031,0.063,0.125,0.25,0.5
0.031,0.063,0.125,0.25,0.5
0.035,0.071,0.141,0.283,0.565
0.035,0.071,0.141,0.283,0.565
LC50
(mg/L Cu)
>0.25
0.29
>1
0.234
0.347
0.171
0.296
0.141
0.146
0.154
0.276
LC50
(% sample)
>100
29.0
>100
46.9
69.5
18.5
32.1
28.2
29.2
27.3
48.9
d Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL). Part 1 preliminary
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute method.


Table 4.14. Results from silverside chronic preliminary testing.
Sample type
Reference
toxicant
Effluent
Receiving water
Part"
2
4
IL
1
2
4
IL
1
2
4
IL
Concentrations tested
(mg/L Cu)
0.019,0.038,0.075,0.15,0.3
0.063,0.125,0.25,0.5, 1.0
0.063,0.125,0.25,0.5, 1.0
6.25, 12.5,25,50, 100b
0.038,0.075,0.15,0.3,0.6
0.075,0.15,0.3,0.6, 1.2
0.05,0.1,0.2,0.4,0.8
6.25, 12.5,25,50, 100b
0.038,0.075,0.15,0.3,0.6
0.031, 0.062, 0.124, 0.247, 0.494
0.031, 0.062, 0.124, 0.247, 0.494
Survival
NOEC
(mg/L Cu)
0.15
0.125
0.125
50
0.15
0.15
0.1
100
0.75
0.124
0.062
Growth
NOEC
(mg/L Cu)
0.15
0.125
0.125
25
0.15
0.15
0.1
100
0.75
0.124
0.062
IC25
(mg/L Cu)
>0.3
0.189
0.152
-
0.227
0.171
0.23
-
0.155
0.149
0.103
IC25
(% sample)
>100
18.9
15.2
43.9
37.8
14.3
28.8
>100
25.9
30.2
20.9
 Preliminary testing Parts 1-4 and referee laboratory testing during the interlaboratory testing phase (IL).  Part 1 preliminary
testing was conducted using only the chronic method, and Part 3 was conducted using only the acute method.
b The units for test concentrations in Part 1 were percent sample.


The effluent sample was adjusted to a salinity of 25 ppt and then spiked with CuSO4 to provide a consistently
toxic sample that was appropriate for the silverside acute and chronic test method.  The spiking level for this
sample type was targeted to produce a test result (LC50 for silverside acute test and IC25 for silverside
                                                      38

-------
chronic test) of 25% sample. Part 2 preliminary testing for the silverside acute test method produced an LC50
of 0.234 mg/L Cu. This sample was held for 7 days at 4°C and then retested for Part 3 preliminary testing.
After holding, the sample produced an LC50 of 0.347 mg/L Cu, which represents a 48% change from the
original Part 2 test result. Based on Part 2 results, the referee laboratory increased the spiking level in Part 4
testing to 0.922 mg/L Cu (approximately 0.234 / the target result of 25% sample). Part 4 testing resulted in
an LC50 of 0.171 mg/L or 18.5% sample.  The same spiking level was used for the interlaboratory sample,
and referee laboratory testing of this sample yielded an LC50 of 32.1% sample.

For the silverside chronic test method, Part 1 testing on the unspiked effluent resulted in an IC25  of 43.9%
sample. Spiking of the effluent sample in Part 2 testing produced an IC25 of 0.227 mg/L Cu. Based on this
result,  the spiking level was increased to 1.2 mg/L Cu for Part 4 testing.  Part 4 testing produced an IC25 of
0.171 mg/L or 14.3% sample.  Based on an average of Part 2 and Part 4 testing results, the interlaboratory
sample was spiked at 0.8 mg/L Cu (approximately the average result of 0.199 / the target result of 25%
sample).  Referee laboratory testing of the interlaboratory sample  produced an IC25 of 28.8% sample.

Table 4.15. Water chemistry of effluent sample source for silverside acute and chronic test methods.
Parameters
Alkalinity (mg/L as CaCO3)
Hardness (mg/L as CaCO3)
Conductivity (/j.S/cm)
pH
Temperature (°C)
Total residual chlorine (mg/L)
Dissolved oxygen (mg/L)
Salinity (ppt)
Copper G/g/L)
Total ammonia (mg/L)
Total dissolved solids (mg/L)
Total suspended solids (mg/L)
Total organic carbon (mg/L)
Biological oxygen demand (mg/L)
Chemical oxygen demand (mg/L)
Sampling date"
08/02/99
217
287
1457
7.37
16.7
0.02
7.9
0
22
14.0
975
4
13.2
15
41
08/23/99
202
289
1424
7.08
23.5
0.03
4.2
0
-
-
-
-
-
-
-
08/26/99
214
282
1476
7.31
18.5
0.02
3.9
0
-
-
-
-
-
-
-
08/28/99
208
288
1486
7.35
22.0
0.02
8.9
0
-
-
-
-
-
-
-
09/17/99
182
292
1320
7.16
19.2
<0.01
8.7
0
-
-
-
-
-
-
-
a "-" indicates that the parameter was not tested on the given sampling date.
                                                 39

-------
4.2.6.3  Receiving water sample type

The receiving water sample type was composed of a natural seawater spiked with CuSO4. Natural seawater
was obtained from the Scripps Institution of Oceanography (Scripps) seawater system in La Jolla, CA.
Scripps pumps seawater from a fixed collection site 320 m offshore of La Jolla, CA, filters the seawater
through a sand filter, and incorporates the seawater into a flow-through system for use in supplying aquariums
housed at Scripps. The referee laboratory routinely uses natural seawater from Scripps' seawater system for
in-house organism culturing and dilution water. The referee laboratory transported water from the Scripps
Institution to the laboratory where it was incorporated into the laboratory's flow-through natural seawater
system that includes two 2,200-gallon storage tanks,  an in-line 20-um filter, and an in-line heater/chiller unit.
Table 4.16 shows the water chemistry of the receiving water sample collected for preliminary testing. Prior to
testing, receiving water was filtered through a 0.2-um filter and adjusted to a salinity of 25 ppt with the
addition of deionized water.

The receiving water sample was then spiked with CuSO4 to provide a consistently toxic sample.  The spiking
level for this sample type was targeted to produce a test result (LC50 for silverside acute test and IC25  for
silverside chronic) of 25% sample.  Part 2 preliminary testing for the silverside acute test method produced an
LC50 of 0.141 mg/L Cu. This sample was held for 7 days at 4°C and then retested for Part 3 preliminary
testing.  After holding, the sample produced an LC50 of 0.146 mg/L Cu, which represented only a 3.5%
change from the initial Part 2 test result. This indicated that the toxicity of the sample was persistent and
appropriate for use in the WET Study. Based on Part 2 testing results, the  spiking level for Part 4
preliminary testing was increased to 0.565 mg/L Cu  (approximately 0.141 /the target result of 25% sample).
Part 4 testing produced an  LC50 of 0.154 mg/L or 27.3% sample.  The same spiking level was used for the
interlaboratory sample, and referee laboratory testing of this sample yielded an LC50 of 48.9% sample.

For the silverside chronic test method, Part  1 testing on the unspiked effluent resulted in no toxicity (IC25 >
100% sample).  Spiking of the receiving water sample in Part 2 testing produced an IC25 of 0.155 mg/L Cu.
Based on this result, the spiking level was decreased slightly to 0.494 mg/L Cu for Part 4 testing. Part 4
testing produced an IC25 of 0.149 mg/L or  30.2% sample.  The same spiking  level was used for the
interlaboratory sample, and referee laboratory testing of this sample yielded an IC25 of 20.9% sample.
                                                40

-------
Table 4.16. Water chemistry of the receiving water sample source for silverside acute and chronic test
methods.
Parameters
Alkalinity (mg/L as CaCO3)
Hardness (mg/L as CaCO3)
Conductivity (^uS/cm)
pH
Temperature (°C)
Total residual Chlorine (mg/L)
Dissolved oxygen (mg/L)
Salinity (ppt)
Copper (Mg/L)
Total ammonia (mg/L)
Total dissolved solids (mg/L)
Total suspended solids (mg/L)
Total organic carbon (mg/L)
Biological oxygen demand (mg/L)
Chemical oxygen demand (mg/L)
Sampling date
08/02/99
75
>2000
53,100
8.08
9.9
<0.01
7.2
34
5.4
<0.1
28,000
<1.0
<0.5
2
26
4.2.7   Preliminary Testing for the Champia Chronic Test Method

The referee laboratory supporting the Champia chronic test method originally was instructed to conduct
preliminary testing as described in Section 4.1. On January 28, 2000, interlaboratory testing for the Champia
chronic test method was canceled due to a lack of participant laboratory support (see Section 2.1).  With this
cancellation, the objectives of any uncompleted preliminary tests were adjusted to better direct the use of
preliminary test data toward single-laboratory validation rather than preparation for interlaboratory testing.
As a result, preliminary testing occurred during two time periods; testing from July to September 1999 was
conducted in preparation for interlaboratory testing, and testing from March to May 2000 was conducted to
provide additional single-laboratory data for the Champia chronic test method.
                                                 41

-------
4.2.7.1  Reference toxicant sample type

The reference toxicant sample type was composed of natural seawater spiked with CuSO4. Since natural
seawater is recommended for the Champia chronic test method, the same natural seawater source was used as
the matrix for the reference toxicant sample, dilution water in all tests, and the receiving water sample matrix.
This natural seawater source is described in more detail in Section 4.2.7.3.  For use as the reference toxicant
sample matrix and dilution water, the natural seawater was filtered through a 0.45-um membrane filter and
steam sterilized at 150°C for 30 minutes. Seawater used for the receiving water sample matrix was unfiltered
and unsterilized.

Table 4.17 shows the results from preliminary testing for the Champia chronic test method.  Spiking levels
selected for Part 2 testing on the reference toxicant sample type resulted in an IC25 of 0.162 /zg/L Cu. Part 4
preliminary testing using the same spiking levels produced an IC25 of 0.264/^g/L Cu.  When the reference
toxicant sample was retested in the Spring of 2000, this sample type resulted in an IC25 of O.SjUg/L Cu.  For
the three tests conducted on the reference toxicant sample type, a mean IC25 of 0.309/^g/L Cu and a CV of
56% was calculated.

Table 4.17. Results from Champia chronic preliminary testing.
Sample
type
Reference
toxicant
Effluent
Receiving
water
Part"
2
4
A
1
1
1
3
4
A
A
1
2
3
4
A
A
A
Test
date
7/27/99
8/17/99
5/16/00
7/28/99
8/3/99
8/10/99
8/4/99
9/14/99
5/9/00
5/9/00
7/28/99
7/27/99
8/3/99
8/18/99
5/31/00
5/23/00
5/23/00
Sample description
filtered, sterilized
natural seawater
spiked with Cu
unspiked municipal
effluent adjusted to
salinity of 30ppt
unspiked natural
seawater (unfiltered
and unsterilized)
spiked natural
seawater (unfiltered
and unsterilized)
unspiked natural
seawater (unfiltered
and unsterilized)
Units
Aig/L Cu
Afg/L Cu
Aig/L Cu
percent
percent
percent
percent
percent
percent
percent
percent
Aig/L Cu
Aig/L Cu
Aig/L Cu
Aig/L Cu
percent
percent
Concentrations tested
0.5, 1.0,5, 10, 15
0.5, 1,5, 10, 15
0.15,0.5, 1,5, 10
0.2,0.7,2.0,7.0, 10.0
0.156,0.312,0.625, 1.25,
2.5,5, 10
0.156,0.312,0.625, 1.25,
2.5,5, 10
0.2,0.7,2.0,7.0, 10.0
0.05,0.1,0.2,0.7,2,7, 10
0.05,0.1,0.2,0.7,2,7, 10
0.05,0.1,0.2,0.7,2,7, 10
100"
0.625, 1.25,2.5,5, 10,20
0.625, 1.25,2.5,5, 10,20
0.625, 1.25,2.5,5, 10,20,40
0.15,0.5, 1,5, 10
6.25, 12.5,25,50, 100
6.25, 12.5,25,50, 100
NOEC
(units)
<0.5
<0.5
0.15
<0.2
0.156
O.156
<0.2
0.050
2.0
0.70
NA
0.625
O.625
O.625
1.0
6.25
100
IC25
(units)
0.162
0.264
0.50
0.190
0.232
0.128
0.163
0.064
0.46
0.59
NA
0.671
0.522
0.720
1.48
9.02
64.54
a Preliminary testing Parts 1-4 and additional testing (A) requested following cancellation of interlaboratory testing.
b Tested as a single concentration (100%) receiving water. No toxicity was indicated.
                                                  42

-------
4.2.7.2  Effluent sample type

The effluent sample type was composed of a municipal wastewater treatment plant effluent. This effluent
source was selected based on historical testing by the referee laboratory that demonstrated relatively consistent
levels of toxicity. No spiking of this effluent was necessary; the unspiked effluent sample produced IC25
values in the range of 0.1% to 0.6% sample.  All tests were performed on unspiked effluent adjusted to a
salinity of 30 ppt. Water chemistry of the effluent on each sample collection date is shown in Table 4.18.

Table 4.18. Water chemistry of the effluent sample source for the Champia chronic test method.
Parameters
Salinity (ppt)
Conductivity (juS/cm)
pH
Total residual chlorine
(mg/L)
Total ammonia (mg/L)
Sampling date8
07/28/99
3
2250
7.56
0.15
22.6
08/03/99
1
2740
6.93
0.22
26.0
08/04/99"
1
2300
7.67
<0.05
-
08/10/99
2
2550
7.24
0.17
15.8
09/14/99
5
2980
6.99
0.38
14.4
05/09/00
1
2156
6.94
0.84
-
a "-" indicates that the parameter was not tested on the given sampling date.
b Sample collected 07/28/99 and held for 7 days at 4°C prior to testing.
Part 1 preliminary testing confirmed that the effluent was relatively consistent in toxicity to Champia parvula.
Results of three separate effluent samples collected on three separate days ranged from 0.128% sample to
0.232% sample with a mean of 0.183% sample and a CV of 28.5%.  The effluent sample collected on 7/28/99
was held for 7 days at 4°C and tested on 8/4/99 for Part 3 preliminary testing.  This test resulted in an IC25 of
0.163% sample, which represents a 14% change from the initial test conducted on that sample.  The effluent
was tested again on 9/14/99 for Part 4 testing, and resulted in an IC25 of 0.064% sample. In the Spring of
2000, the referee laboratory conducted duplicate testing of the effluent. The resulting IC25s were 0.46% and
0.59% effluent sample, a relative difference of 25% between split samples.

4.2.7.3  Receiving water sample type

The receiving water sample type was composed of a natural surface water spiked with CuSO4.  The receiving
water was collected from Rye Harbor, Rye, New Hampshire. The harbor provides anchorage for small
pleasure craft and a limited number of small commercial fishing vessels. The harbor receives no direct
discharges of treated or untreated wastewater. The water in the harbor is classified as SA-1 and has been used
by the referee laboratory since 1991 to maintain Champia cultures. Receiving water was collected from a boat
offshore away from other boat traffic or potential contamination.  The physical and chemical characteristics of
                                                 43

-------
the receiving water are listed in Table 4.19 for each sample collection date.  The same water source was used
for the reference toxicant sample matrix and for dilution water in all tests, however, water was filtered and
sterilized for these uses.  The receiving water  sample type was tested without filtration or sterilization.

Part  1 preliminary testing initially was conducted using a single concentration (100%) test. Results from this
test showed no toxicity in the receiving water. In the Spring of 2000, the referee laboratory conducted
duplicate testing on the receiving water. IC25 results from a split sample collected on 5/23/00 were 9.02%
and 64.54% sample.  No explanation was provided by the referee laboratory on the discrepancy between
results of these duplicate samples.
In Part 2 preliminary testing, the receiving water sample was spiked with CuSO4 to provide a consistently
toxic sample. Part 2 testing for the Champia chronic method produced an IC25 of 0.671 Mg/L CuSO4. This
sample was held for 7 days at 4°C and then retested for Part 3 preliminary testing.  After holding, the sample
produced an IC25 of 0.522 yUg/L CuSO4, which represents a 22% change from the original Part 2 test result.
Part 4 testing resulted in an IC25 of 0.72 /ug/L CuSO4.  An additional test of the spiked receiving water in the
spring of 2000 produced an IC25 of 1 .48 /ug/L CuSO4.  For the four tests conducted on the spiked receiving
water sample type, a mean IC25 of 0.848 //g/L Cu and a CV of 51% was calculated.
Table 4.19. Water chemistry of the receiving water sample source for the Champia chronic test method.
Parameters
Salinity (ppt)
Conductivity
(/zS/cm)
PH
Total residual
chlorine (mg/L)
Total ammonia
(mg/L)
Sampling date"
07/28/99
33
48100
7.92
O.05
O.10
08/03/99
33
48600
8.29
<0.05
<0.10
08/04/99
33
49100
8.54
<0.05
<0.10
08/10/99
33
49300
8.04
<0.05
<0.10
09/14/99
32
45400
7.70
<0.05
0.22
05/16/00
32
42500
7.89
O.05
-
05/23/00
33
42700
8.00
<0.05
-
05/23/00
33
43000
8.07
<0.05
-
05/31/00
33
43010
8.01
O.05
-
1 "-" indicates that the parameter was not tested on the given sampling date.
                                                 44

-------
4.2.8   Preliminary Testing for the Holmesimysis Acute Test Method

The referee laboratory supporting the Holmesimysis acute test method was originally instructed to conduct
preliminary testing according to Section 4.1.  Due to difficulties encountered in obtaining test organisms, a
limited number of preliminary tests were conducted by the referee laboratory from July to September 1999.
On January 28, 2000, interlaboratory testing for the Holmesimysis acute test method was canceled due to a
lack of participant laboratory support (see Section 2.1). With this cancellation, the objectives of any
uncompleted preliminary tests were adjusted to better direct the use of preliminary test data toward single-
laboratory validation rather than preparation for interlaboratory testing. The referee laboratory attempted to
complete preliminary testing from April through June 2000, however, the laboratory was unable to obtain test
organisms during this period. On June 29, 2000, any further preliminary testing at the referee laboratory was
canceled due to the unavailability of test organisms.

In all, only five preliminary Holmesimysis acute tests were performed (Table 4.20); one test was conducted on
receiving water, and two tests were conducted on each of two effluent sources. Of the five tests conducted,
only two met test acceptability criteria for survival, and these two tests  were not conducted according to the
WET method manual test requirement for test organism age. Neonates for these tests were collected directly
from the field rather than hatched in the laboratory from field collected adults. For this reason, exact ages of
neonates used for testing could not be determined (see Section 4.3).

Table 4.20. Results from Holmesimysis acute preliminary testing.
Sample type
Effluent 1
Effluent 1
Effluent 2
Effluent 2
Receiving water
Test date
7/21/99
7/27/99
8/31/99b
9/7/99b
7/21/99
Control survival (%)
80a
80a
92.5
90
80a
NOEC (% sample)
12.5
12.5
25
12.5
100
LC50 (% sample)
25.3
15.3
35.1
16.6
>100
* Failed to meet test acceptability criteria of >90% control survival.
b Tests were conducted on field collected neonates.
4.2.8.1  Effluent sample type

Two effluent sources were tested and considered for use as the effluent sample source.  Both effluent sources
were from municipal wastewater treatment facilities, and water chemistry for both is listed in Table 4.21.
Two Holmesimysis acute tests were conducted on effluent 1; these tests produced LCSOs of 23.9% and 12.4%
sample (Table 4.20). Control survival in each of these tests was only 80%, which is below the test
acceptability criteria for the method. Artificial seawater was used for dilution in these tests rather than natural
seawater as stated in the SOP. The dilution water may have been a factor in the poor control survival.  Due to
                                                  45

-------
difficulties in obtaining test organisms, tests conducted on effluent 2 used neonates directly collected from the
field, rather than neonates hatched in the laboratory from field collected adults.  These two tests resulted in
LCSOs of 28.2% and 12.6% sample.

Table 4.21. Water chemistry of the effluent sample sources for the Holmesimysis acute test method.
Parameters
Alkalinity (mg/L as CaCO3)
Hardness (mg/L as CaCO3)
pH
Temperature (°C)
Total residual chlorine (mg/L)
Dissolved oxygen (mg/L)
Salinity (ppt)
Copper (Mg/L)
Total ammonia (mg/L)
Total dissolved solids (mg/L)
Total suspended solids (mg/L)
Biological oxygen demand (mg/L)
Chemical oxygen demand (mg/L)
Effluent 1
Sampling date"
07/21/99
-
-
7.28
19.2
0.07
3.3
0.6
-
22.3
-
-
-
-
07/26/99
250
180
7.26
18.0
0.07
6.0
0.5
10
18.9
562
13.5
32.2
98
Effluent 2
Sampling date"
08/30/99
-
-
7.09
23.3
-
6.2
0
-
18.7
-
-
-
-
09/07/99
-
-
7.88
24.0
0.09
6.9
0
-
14.7
-
-
-
-
" "-" indicates that the parameter was not tested on the given sampling date.

4.2.8.2  Receiving water sample type

The receiving water sample type consisted of natural seawater collected from San Francisco Bay off of Point
Chauncey on the Tiburon Peninsula.  Samples were collected away from direct discharges of treated or
untreated wastewater. Water from this source has been used by the referee laboratory as dilution water for
several years without exhibiting toxicity.  The physical and chemical characteristics of the receiving water
collected from the bay are listed in Table 4.22. A single Part 1 preliminary test was conducted on the
receiving water resulting in an LC50 of > 100% sample. This test failed to meet test acceptability criteria due
to excessive control mortality. As mentioned above, use of artificial seawater for dilution may have
contributed to low control survival.
                                                  46

-------
Table 4.22. Water chemistry of the receiving water sample source for the Holmesimysis acute test
method.
Parameters
Alkalinity (mg/L as CaCO3)
Hardness (mg/L as CaCO,)
pH
Temperature (°C)
Total residual chlorine (mg/L)
Dissolved oxygen (mg/L)
Salinity (ppt)
Copper (,wg/L)
Total ammonia (mg/L)
Total dissolved solids (mg/L)
Total suspended solids (mg/L)
Biological oxygen demand (mg/L)
Chemical oxygen demand (mg/L)
Sampling date
07/20/99
122a
4558s
7.85
17.9
0.03
7.6
24
NDa
<0.10
27,200a
1.25a
NDa
1930"
• Analyses were conducted on the sample collected 7/27/99, but this sample was not used for Holmesimysis testing due to
insufficient test organisms. ND = not detected.
4.3     Problems Encountered in Preliminary Testing

For the Selenastrum chronic test method, spiking levels were originally targeted to produce an IC25 of 50%
for the reference toxicant sample type and 25% for the effluent and receiving water sample type. In Part 2
preliminary testing, IC25 values in repeated tests were variable and made it difficult for the referee laboratory
to isolate the correct spiking level. The referee laboratory observed that IC50 values were less variable than
IC25 values in repeated tests. For this reason, the target spiking levels were based on IC50 values rather than
IC25 values.  Target spiking levels were set to produce an IC50 of 38% for reference toxicant, effluent, and
receiving water sample types. The 38% level was selected in an attempt to allow IC25 and IC50 results from
interlaboratory testing to fall within the test concentration range.

For the mysid chronic and sheepshead acute and chronic test methods, CuSO4 was  originally selected as the
spiking agent.  During preliminary testing for these methods, results on spiked samples were highly variable.
Upon further investigation, the referee laboratory determined that this variability was due to precipitation of
copper in the spiked saltwater samples. A combination of factors including spiking concentrations, salinity of
                                                  47

-------
the sample, pH, other dissolved ions in the sample matrix, and storage of the spiked sample at <4°C
contributed to the precipitation of copper in spiked samples. Due to this precipitation, the spiking agent for
these marine methods was changed to KC1. The referee laboratory had experience in the use of KC1 as the
spiking agent for freshwater methods in the WET Study and had experience in the use of KC1 as a reference
toxicant for these methods. The same problem was encountered for the silverside acute reference toxicant
sample (see  Section 5.3), but the problem was not identified in time to change the spiking agent. Precipitation
of copper did not appear to affect the other sample types for the silverside acute and  chronic test methods,
possibly due to the lower spiking concentrations used for these methods.

For the Holmesimysis acute test method, problems were encountered in obtaining test organisms. Organisms
for this test method are generally field collected from kelp beds  off the coast of California, but they are not
present in sufficient numbers during the winter months.  From April through June 2000, the referee laboratory
attempted to collect organisms to complete preliminary testing, but Holmesimysis populations were still not at
sufficient densities at potential sites in San Diego or Santa Cruz, CA. Even when field-collected adult
Holmesimysis were available (July through September 1999), obtaining sufficient neonates within the required
age range was difficult.  Field-collected gravid females were held in the laboratory and culled daily to obtain
neonates within a 24-hour age range. This required maintaining a large number of gravid females to produce
the necessary neonates.  In addition, survival of newly hatched neonates was poor, which added to difficulties
in obtaining a sufficient number of test organisms. To avoid these difficulties, the referee laboratory collected
neonates directly from the field  for use in two preliminary tests.   The smallest of collected organisms were
used in these tests. This technique of obtaining test organisms did not allow an exact determination of the age
of test  organisms, so the age of test organisms could have been outside of the required range.

A second problem encountered in Holmesimysis acute preliminary testing was poor survival of neonates in test
controls.   Three of the five preliminary tests conducted failed to meet the test acceptability criteria of 90%
survival.   The two tests that met this criteria were tested using field-collected neonates, so it could not be
confirmed that test organism ages were within the range required in the methods manual.
                                                 48

-------
5.0 SAMPLE PREPARATION

For each method, four test sample types were prepared in bulk by the referee laboratory, divided, and
distributed to participant laboratories for testing. The four sample types included a blank sample, a reference
toxicant sample, an effluent sample, and a receiving water sample. Test samples were selected and/or
prepared to achieve a targeted range of effects in interlaboratory testing.  Preliminary testing was conducted to
validate the selection of real-world samples and spiking concentrations and confirm that each sample produced
the targeted effect and was appropriate for use in the WET Study.

5.1     Freshwater  Methods

For the freshwater methods, Table 5.1  describes each of the sample types that were prepared and distributed
for interlaboratory testing. The blank sample type for all freshwater test methods, with the exception of the
Selenastrum chronic method, consisted of moderately hard synthetic freshwater prepared according to Section
7 of the WET method manual.  The blank sample type was prepared by adding the appropriate amounts of
reagents (Section 7 of the WET method manual) to deionized water in cleaned and rinsed 5-gallon or 3-gallon
(depending on the volume needed for interlaboratory testing) plastic carboys. Following preparation, the bulk
blank sample was properly mixed and  aerated for at least 24 hours prior to removing aliquots for packaging
and distribution to participant laboratories. For the Selenastrum chronic test method, the blank sample was
prepared as deionized water. The bulk blank sample for each  test method was stored in the dark at <4°C prior
to distribution to participant laboratories.

The reference toxicant sample type for all freshwater test methods consisted of the blank sample matrix
(moderately hard synthetic freshwater for Ceriodaphnia and fathead test  methods and deionized water for the
Selenastrum  chronic method) spiked with KC1. Moderately hard synthetic freshwater was prepared by
adding the appropriate amounts of reagents (Section 7  of the WET method manual) to deionized water in
cleaned and rinsed 5-gallon or 3-gallon (depending on the volume needed for interlaboratory testing) plastic
carboys. The appropriate amount of reagent grade  KC1 was then added to the moderately hard synthetic
freshwater (or deionized water for the Selenastrum chronic method) to achieve the spiking concentrations
listed in Table 5.2 for the reference toxicant ampules.  When ampules were reconstituted according to
instructions provided in the participant laboratory SOPs (Appendix B), the resulting reconstituted sample
yielded the spiking concentrations listed in Table 5.1. The bulk reference toxicant sample for each test method
was mixed thoroughly prior to spiking, following spiking, and prior to removing aliquots for distribution to
participant laboratories. Bulk reference toxicant samples were stored in the dark at <4°C prior to distribution
to participant laboratories.
                                                 49

-------
Table 5.1.  Description of samples used for freshwater methods in the WET Study.
Test method
Ceriodaphnia
acute
Ceriodaphnia
chronic
Fathead acute
Fathead
chronic
Selenastrum
chronic
Sample type
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Sample matrix
MHSFb
MHSFb
Municipal effluent
River water
MHSFb
MHSF"
Municipal effluent
River water
MHSF"
MHSFb
Municipal effluent
River water
MHSFb
MHSFb
Municipal effluent
River water
Deionized water
Deionized water
Municipal effluent
River water
Spiking
agent
None
KC1
KC1
KC1
None
KC1
KC1
KC1
None
KC1
KC1
KC1
None
KC1
KC1
KC1
None
KC1
KC1
KC1
Spiked
concentration"
(mg/L as KC1)
-
1000
2680
1800
-
270
2100
1200
-
2200
5328
5000
-
1220
3600
2400
-
5655
11540
11713
Collection
date
-
-
11/08/99
11/05/99
-
-
10/22/99
10/08/99
-
-
11/01/99
10/18/99
-
-
10/01/99
09/24/99
-
-
03/06/00
03/20/00
Preparation
date
1 1/06/99
1 1/06/99
11/09/99
11/06/99
10/09/99
10/09/99
10/24/99
10/10/99
10/19/99
10/19/99
11/02/99
10/19/99
09/24/99
09/24/99
10/03/99
09/26/99
03/14/00
03/07/00
03/07/00
03/21/00
Shipment date
11/08/99, 11/10/99
11/08/99, 11/10/99
11/10/99
11/08/99
10/11/99, 10/25/99
10/11/99, 10/25/99
10/25/99
10/11/99
10/20/99, 11/03/99
10/20/99, 11/03/99
11/03/99
10/20/99
09/27/99, 10/04/99
09/27/99, 10/04/99
10/04/99
09/27/99
03/15/00,03/22/00,
03/29/00
03/08/00,03/15/00,
03/22/00, 03/29/00
03/08/00, 03/15/00
03/22/00
1 Spiked concentrations for reference toxicant samples represent concentrations after proper reconstitution of the ampule. See Table 5.2 for a description of ampule contents.
' Moderately hard synthetic freshwater prepared according to Section 7 of the WET method manuals.
                                                                           50

-------
Table 5.2.  Spiking concentrations in reference toxicant ampules.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Reference
toxicant
KC1
KG1
KC1
KC1
KC1
KC1
KC1
KC1
CuSO4
CuSO4
Spiked cone, in
prepared
ampule
(mg/L)
10,000
8,100
88,000
97,600
113,100
108,000
25,280
90,000
40.0
210
Volume of
ampule added
to
reconstituted
sample (inL)
100
100
100
100
200
100
500
500
100
100
Reconstituted
sample volume
(L)
1
3
4
8
4
9
4
15
4
21
Spiked cone, in
recon-stituted
sample
(mg/L)
1,000
270
2,200
1,220
5,655
1,200
3,160
3,000
1.0
1.0
The effluent sample type for all freshwater test methods consisted of a municipal effluent spiked with KC1.
The source and characteristics of the municipal effluent used as the matrix for this sample type is described in
Section 4.2.1.2. Effluent sample was collected from a small access pipe through which effluent was pumped
from the main discharge outfall. Sample was collected from this access pipe using a funnel and tubing to fill
5-gallon plastic carboys. Sample was then immediately transported to the referee  laboratory at ambient
temperature in 5-gallon plastic carboys. Upon arrival, the referee laboratory stored the sample in the dark at
<4°C.  Table 5.3 lists the volume of effluent collected for each test method. The effluent sample was then
homogenized in 50-gallon plastic containers. If the volume of sample collected for interlaboratory testing was
greater than 50 gallons for a given test method, sample was stored and homogenized in multiple 50-gallon
plastic containers connected with piping. Submersible pumps were used to circulate and homogenize sample
between individual containers. Following homogenization, the effluent sample for each test method was
spiked at the appropriate concentration with KC1 (Table 5.1). The appropriate amount of KC1 was initially
dissolved in a small volume of effluent and then added to the bulk effluent sample. When multiple containers
were used, equal amounts of KC1 were added to each container. The bulk effluent sample for each test method
was mixed thoroughly prior to spiking, following spiking, and prior to removing aliquots for distribution to
participant laboratories.  Bulk effluent samples were stored in the dark at <4°C prior to distribution to
participant laboratories.
                                                 51

-------
The receiving water sample type for all freshwater test methods consisted of a natural river water spiked with
KC1.  The source and characteristics of the river water used as the matrix for this sample type is described in
Section 4.2.1.3.  Receiving water sample was collected by filling 1-gallon plastic containers under the surface
of the water.  Receiving water sample was then transferred to 5-gallon plastic containers and immediately
transported to the referee laboratory at ambient temperature. Upon arrival at the referee laboratory, the
sample was stored at <4°C in the dark. Table 5.4 lists the volume of receiving water collected for each test
method. The receiving water sample was then homogenized and spiked as described in the preceding
paragraph for the effluent sample type.

Table 5.3.  Effluent sample volumes collected for interlaboratory testing.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Volume required
per participant lab
sample (L)
1
7
4
17.5
4
21
4
35
4
49
Number of
participant lab
samples required
28
28
28
28
16
16
8
8
13
14
Minimum
required volume
(L)'
28
196
112
490
64
336
32
280
52
686
Collected
volume
(L)
45
270
150
660
100
450
60
375
60
1000
' Minimum required volume = Volume per sample X number of samples.
Table 5.4.  Receiving water sample volumes collected for interlaboratory testing.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Volume required
per participant lab
sample (L)
1
7
4
17.5
4
21
4
35
4
49
Number of
participant lab
samples required
14
14
14
14
9
9
8
8
7
8
Minimum
required volume
(L)"
14
98
56
245
36
189
32
280
28
392
Collected
volume
(L)
19
135
90
345
60
278
75
389
36
540
1 Minimum required volume = Volume per sample X number of samples.
                                                 52

-------
5.2     Marine Methods

For the marine methods, Table 5.5 describes each of the sample types that were prepared and distributed for
interlaboratory testing.

5.2.1    Mysid Chronic and Sheepshead Acute and Chronic Test Methods

The blank sample type for the mysid chronic, sheepshead acute, and sheepshead chronic test methods
consisted of synthetic seawater prepared at a salinity of 25 ppt. The blank sample type was prepared by
adding the appropriate amounts of bioassay-grade Forty Fathoms artificial sea salts to deionized water in
cleaned and rinsed 5-gallon or 3-gallon (depending on the volume needed for interlaboratory testing) plastic
carboys. Following preparation, the bulk blank sample was properly mixed to dissolve the added reagents.
The bulk sample was again mixed prior to removing aliquots for packaging and distribution to participant
laboratories.  The bulk blank sample for each test method was stored in the dark at <4°C prior to distribution
to participant laboratories.

The reference toxicant sample type for the mysid chronic, sheepshead acute, and sheepshead chronic test
methods consisted of deionized water spiked with KC1.  For these test methods, KC1 was spiked in deionized
water rather than artificial seawater to improve the solubility of KC1 in the highly concentrated reference
toxicant ampule samples. The appropriate amount of reagent grade KC1 was added to deionized water in
cleaned and rinsed 5-gallon or 3-gallon (depending on the volume needed for interlaboratory testing) plastic
carboys to achieve the spiking concentrations listed in Table 5.2 for the reference toxicant ampules.  When
ampules were reconstituted according to instructions provided in the participant laboratory SOPs (Appendix
B), the resulting reconstituted sample yielded the spiking concentrations listed in Table 5.5. The bulk
reference toxicant sample for each test method was mixed thoroughly prior to spiking, following spiking, and
prior to removing aliquots for distribution to participant laboratories.  Bulk reference toxicant samples were
stored in the dark at <4°C prior to distribution to participant laboratories.
                                                 53

-------
Table 5.5. Description of samples used for marine methods in the WET Study.
Test
method
Mysid
chronic
Sheepshead
acute
Sheepshead
chronic
Silverside
acute
Silverside
chronic
Sample type
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Sample matrix
synthetic seawaterb
synthetic seawater11
municipal effluent
natural seawater
synthetic seawaterb
synthetic seawaterb
municipal effluent
natural seawater
synthetic seawaterb
synthetic seawaterb
municipal effluent
natural seawater
synthetic seawater11
synthetic seawaterb
industrial effluent
natural seawater
synthetic seawater6
synthetic seawaterb
industrial effluent
natural seawater
Spiking agent
None
KC1
KC1
KC1
None
KC1
KC1
KC1
None
KC1
KC1
KC1
None
CuSO4
CuSO4
CuSO4
None
CuSO4
CuSO4
CuSCX,
Spiked
concentration"
(mg/L as KC1
and as Cu)
-
1200
2000
2400
-
3160
5200
5600
-
3000
4880
4400
-
1.000
0.922
0.565
-
1.000
0.800
0.494
Collection
date
-
-
02/16/00
02/24/00
-
-
03/06/00
02/24/00
-
-
03/23/00
03/17/00
-
-
1 1/03/99
10/30/99
-
-
10/20/99
10/16/99
Preparation
date
02/18/00
02/18/00
02/18/00
02/25/00
03/03/00
03/10/00
03/10/00
03/03/00
03/17/00
03/25/00
03/24/00
03/17/00
10/31/99
10/31/99
11/07/99
10/31/99
10/17/99
10/17/99
10/23/99
10/17/99
Shipment date
02/21/00, 02/28/00
02/21/00, 02/28/00
02/21/00
02/28/00
03/06/00
03/13/00
03/13/00
03/06/00
03/20/00
03/27/00
03/27/00
03/20/00
11/01/99, 11/08/99
11/01/99, 11/08/99
11/08/99
11/01/99
10/18/99, 10/25/99
10/18/99
10/25/99
10/18/99
' Spiked concentrations for reference toxicant samples represent concentrations after proper reconstitution of the ampule. See Table 5.2 for a description of ampule contents.
b Synthetic seawater was prepared using bioassay grade Forty Fathoms added to deionized water.
                                                                           54

-------
The effluent sample type for the mysid chronic, sheepshead acute, and sheepshead chronic test methods
consisted of a municipal effluent spiked with KC1.  The municipal effluent described previously for freshwater
methods (Section 4.2.1.2) also was used for the mysid chronic, sheepshead acute, and sheepshead chronic test
methods.  The effluent was collected and transported to the referee laboratory as previously described for the
freshwater methods (Section 5.1). Table 5.3 lists the volume of effluent collected for each test method. The
effluent sample was then homogenized in 50-gallon plastic containers. If the volume collected and necessary
for interlaboratory testing was greater than 50 gallons for a given test method, sample was stored and
homogenized in multiple 50-gallon plastic containers connected with piping. Submersible pumps were used to
circulate and homogenize sample between individual containers. Following homogenization, the salinity of the
effluent sample was adjusted by the  addition of bioassay-grade Forty Fathoms artificial sea salts. Salinity of
the effluent sample was adjusted to 23 ppt, 20 ppt, and 21 ppt for the mysid chronic, sheepshead acute, and
sheepshead chronic test methods, respectively.  Since the addition of the KC1 spike increased salinity, the
initial salinity levels were selected to achieve a final salinity of 25 ppt in the  effluent sample following spiking
with KC1. Following salinity adjustment, the effluent sample was spiked at the appropriate concentration with
KC1 (Table 5.5).  The appropriate amount of KC1 was initially dissolved in a small volume of effluent and
then added to the bulk effluent sample. When multiple containers were used, equal amounts of KC1 were
added to each container.  The bulk effluent sample for each test method was mixed thoroughly prior to
spiking, following spiking, and prior to removing aliquots for distribution to participant laboratories.  Bulk
effluent samples were stored in the dark at <4°C prior to distribution to participant laboratories.

The receiving water sample type  for the mysid chronic, sheepshead acute, and sheepshead chronic test methods
consisted of a natural seawater spiked with KC1. The source and characteristics of the seawater used as the
matrix for this sample type is described in Section 4.2.4.3.  Seawater was collected within approximately 1
hour of high tide using a submersible pump, and transferred into 5-gallon carboys for transport.  The seawater
sample collected for each test method was filtered through a 5-um filter either at the time  of collection or upon
receipt at the referee laboratory.  Receiving water sample was transported to the referee laboratory at ambient
temperature.  Upon arrival, the referee laboratory stored the sample in the dark at <4°C.  Table  5.4 lists the
volume of receiving water collected for each test method. The receiving water sample was then homogenized
and spiked as described in the preceding paragraph for the effluent sample type.

5.2.2   Silverside Acute and Chronic Test Methods

The blank sample type for the silverside acute and chronic test methods consisted of synthetic seawater
prepared at a salinity of 25 ppt.  The blank sample type was prepared by adding the appropriate amounts of
bioassay-grade Forty Fathoms artificial  sea salts to deionized water in plastic carboys. Following preparation,
the bulk blank sample was properly mixed to dissolve the added reagents. The bulk sample was mixed again
prior to removing aliquots for packaging and distribution to participant laboratories.  The bulk blank sample
for both test methods was stored in the dark at <4°C prior to distribution to  participant laboratories.
                                                 55

-------
The reference toxicant sample type for the silverside acute and chronic test methods consisted of the blank
sample matrix (synthetic seawater at 25 ppt salinity) spiked with CuSO4.  The appropriate volume of a
concentrated CuSO4 stock solution was added to synthetic seawater in 10-L plastic containers to achieve the
spiking concentrations listed in Table 5.2 for the reference toxicant ampules.  When ampules were
reconstituted according to instructions provided in the participant laboratory SOPs, the resulting reconstituted
sample yielded the spiking concentrations listed in Table 5.5.  The bulk reference toxicant sample  was mixed
thoroughly prior to spiking, following spiking, and prior to removing aliquots for distribution to participant
laboratories.  The bulk reference toxicant sample was stored in the dark at <4°C prior to distribution to
participant laboratories.

The effluent sample type for the silverside acute and chronic test methods consisted of an industrial effluent
spiked with CuSO4. The source and characteristics of the  industrial effluent used as the matrix for this sample
type is described in Section 4.2.6.2. The effluent was collected using an automatic sampler set to  collect a
single grab sample.  Sample was collected and stored in 5-gallon buckets with plastic liners and plastic lids for
transport to the referee laboratory.  Upon arrival, the referee laboratory stored the sample in the dark at <4°C.
Table 5.3 lists the volume  of effluent collected for each test method.  The effluent sample was then
homogenized in 250-L plastic containers using a mechanical mixer. The volume required for the silverside
chronic test method necessitated the preparation of separate batches of effluent sample prior to shipment of
test samples for test initiation, the first renewal shipment, and the second renewal shipment.  For each batch, it
was necessary to mix and hold sample in two 250-L plastic containers. Sample was homogenized between the
two containers by repeatedly (six times) adding 40% of one container to the other and mixing each time. For
the silverside acute test method, all effluent sample was mixed and prepared in a single 250-L container.
Following homogenization, the salinity of the effluent sample was adjusted to 25 ppt by the addition of
bioassay-grade Forty Fathoms artificial sea salts.  The effluent sample was then spiked by adding the
appropriate volume of a concentrated CuSO4 stock solution to achieve the spiking concentrations listed in
Table 5.5.  The bulk effluent sample for each test method was mixed thoroughly prior to spiking, following
spiking, and prior to removing aliquots for distribution to participant laboratories. Bulk effluent samples were
stored in the  dark at <4°C prior to distribution to participant laboratories.

The receiving water sample type  for the silverside acute and chronic test methods consisted of a natural
seawater spiked with CuSO4. The source and characteristics of the seawater used as the matrix for this
sample type are described in Section 4.2.6.3.  Natural seawater was collected from the Scripps Institution of
Oceanography seawater system in La Jolla, CA.  Seawater was trucked to the referee  laboratory, where it was
stored and incorporated into the laboratory's flow-through seawater system that includes two 2,200-gallon
storage tanks, an in-line 20-um filter, and an in-line heater/chiller unit. The salinity of receiving water
(initially 34 ppt)  was adjusted to 25 ppt with the addition of deionized water.  The bulk receiving water sample
then was homogenized in 250-L plastic containers and spiked by adding the appropriate volume of a
concentrated CuSO4 stock solution to achieve the spiking concentrations listed in Table 5.5.  The bulk
receiving water sample for each test method was mixed thoroughly prior to spiking, following spiking, and
prior to removing aliquots for distribution to participant laboratories.  Bulk receiving water samples were
stored in the dark at <4°C prior to distribution to participant laboratories.

                                                  56

-------
5.3     Problems Encountered in Sample Preparation

The reference toxicant sample prepared for the Ceriodaphnia chronic test method produced toxicity in only
some of the participant laboratories. The referee laboratory also did not detect toxicity in this sample during
the interlaboratory testing phase. It was determined that the reference toxicant sample prepared for
interlaboratory testing was spiked at a level that was only slightly toxic and very near the minimum detection
level (100% sample). Depending on the sensitivity of test organisms at individual laboratories, some
laboratories identified the sample as toxic, while other laboratories did not.  The spiking level used for this
sample was based on preliminary testing results from three  tests that indicated an IC25 of 138, 132, and 134
mg/L KC1. The average of these tests (135 mg/L) was multiplied by two to obtain the target spiking level of
270 mg/L. It should be noted that one of the preliminary tests produced an IC25 of 320 mg/L KC1, indicating
that 135 mg/L may have been a low estimate of the IC25 for KC1.

The reference toxicant sample for the silverside acute test also did not produce a toxic response in
interlaboratory testing.  This was caused by precipitation of the spiked copper in the liquid ampule sample.
Precipitation of copper  in marine samples also was observed in preliminary testing for the mysid and
sheepshead test methods, so the spiking agent for these methods was changed (see Section 4.3).  For the
silverside methods, this potential problem was not identified during preliminary testing so no change in the
spiking agent was made prior to interlaboratory testing.  Part 4 preliminary testing of the reference toxicant
sample did not reveal copper precipitation and produced an  LC50 of 0.29 mg/L Cu. The same spiking levels
used in Part 4 testing were used for the interlaboratory sample, yet results for that sample showed
precipitation and no toxicity. The referee laboratory discovered that in Part 4 preliminary testing, the
reference toxicant ampule sample was prepared in deionized water, but the interlaboratory sample was
prepared in synthetic seawater.  To confirm that this difference in sample matrix resulted in the non-toxic
interlaboratory sample, the referee laboratory prepared two  additional reference toxicant ampule samples at
the same spiking level used for interlaboratory testing. One ampule was prepared in deionized water and one
was prepared in synthetic seawater. Testing results from the ampule prepared in deionized water were
consistent with Part 4 preliminary testing results (LC50 of 25.5% sample),  and results from the ampule
prepared in synthetic seawater were consistent with  interlaboratory testing results (LC50 >100% sample).
These additional tests confirmed that preparation of the reference toxicant ampule sample in synthetic
seawater caused the precipitation of the spiked copper and produced a nontoxic sample for interlaboratory
testing.
                                                 57

-------
6.0 PACKAGING AND DISTRIBUTION OF TEST SAMPLES

6.1     Sample Distribution Scheme

Laboratories participating in the base study design each received four blind test samples; laboratories
participating in the extended study design each received three blind test samples. The sample distribution
scheme in Table 6.1 shows the number and type of samples distributed to each laboratory according to the
laboratory's assigned position in the study. EPA sponsored laboratories were randomly assigned to positions
1-9, non-EPA sponsored laboratories in the base study design were randomly assigned to positions 10-20, and
laboratories participating in the extended study design were randomly ordered in positions 21- the total
number of laboratories. As noted in Table 6.1, an alternate sample distribution scheme was used for the
sheepshead acute and chronic test methods. Since fewer than nine laboratories participated in these methods,
the sample distribution scheme would not have met the data quality objective of obtaining six data sets for
each test method and sample type. For these test methods, one of each of the sample types was distributed to
each of the seven participant laboratories.

6.2     Packaging and Shipping of Samples

After bulk test samples were prepared according to Section 5, each bulk test sample was divided into
individual test sample aliquots for shipment to participant laboratories. Test sample aliquots were divided into
containers appropriate for the individual test sample volumes. Sample containers were pre-rinsed, filled with
the sample, and sealed with zero head-space. All blank and reference toxicant samples were prepared and
packaged as sample ampules. Ampules were small volume (generally 100 mL) liquid samples that were
reconstituted at participant laboratories to provide the necessary test sample volume. For the Selenastrum
chronic test and the sheepshead acute and chronic tests, larger volumes (200 mL, 500 mL, and 500 mL,
respectively) were used for ampules to reduce the concentration of KC1 in the reference toxicant ampule
sample and avoid potential solubility problems.  All ampule samples (blank and reference toxicant sample
types) for a given test method were shipped in the same container style and size. Effluent and receiving water
samples were prepared and packaged as whole volume samples. All whole volume samples for a given test
method were shipped in the same  container style and size.  Tables  6.2 and 6.3 show the volumes and numbers
of samples prepared for freshwater and marine test methods, respectively.

For chronic test methods that required daily renewal, samples were packaged and shipped in three separate
aliquots. The first aliquot (Initiation) was received by the participant laboratory on test Day 0 and was used
for test initiation on Day 0 and test renewal on Day 1. The second aliquot (Renewal 1) was received on test
Day 2 and used for test renewals on Day 2 and Day 3.  The final aliquot (Renewal 2) was received on test Day
4 and used for test renewal on Day 4, Day 5, Day 6,  and Day 7 if necessary (for the Ceriodaphnia chronic
test method).  For ampule samples, all three separate aliquots were received on test Day 0, and participant
laboratories were instructed to reconstitute the samples following the scheme described above (separate
                                                58

-------
aliquots reconstituted on test Days 0, 2, and 4).  For all acute test methods, a single aliquot of sample was
received and used for test initiation and any required renewals.
Table 6.1. Sample distribution scheme for the WET Study."
Participant laboratory
assigned position
Base study
design
EPA-
sponsored
Base study
design
non-EPA
sponsored
Extended
study design
non-EPA
sponsored
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21 -up
(odd #s)
21 -up
(even #s)
Number of samples of each type received
Blank
1
1
1
1
0
1
0
1
0
1
1
1
0
0
1
1
0
0
1
1
2
1
Effluent
1
2
0
1
2
1
2
1
2
1
2
0
2
2
2
0
2
2
2
0
0
0
Receiving water
1
1
1
0
1
0
1
0
1
1
1
1
0
0
1
1
0
0
1
1
0
0
Reference toxicant
1
0
2
2
1
2
1
2
1
1
0
2
2
2
0
2
2
2
0
2
1
2
" This sample distribution scheme was used for interlaboratory testing of all test methods except the sheepshead acute and chronic
test methods. For these test methods, one of each of the sample types was distributed to each of the seven participant laboratories.
Samples were cooled to <4°C prior to shipment and then packed in coolers (e.g., 28, 48, 54-qt) containing wet
ice.  Depending on the test method performed by an individual participant laboratory, multiple test samples
were shipped in one cooler if possible to reduce the number of coolers shipped. Test sample volumes that
exceeded the maximum weight limit for overnight shipping were divided into separate coolers for shipment.
Duplicate test  sample aliquots were shipped in the same cooler whenever possible; if test sample volume
prohibited shipping duplicates in the same cooler, they were shipped under the same airbill to ensure they were
shipped together.  All samples were shipped FedEx Priority Overnight for delivery on the day of scheduled
testing (see Section 2.2.4). Referee laboratories conducted testing simultaneously with participant laboratories
                                                  59

-------
on samples prepared identically to those for interlaboratory testing and shipped round-trip back to the referee
laboratory.

6.3     Sample Tracking

Each WET test method received an EPA episode number to designate samples prepared for that test method.
Each sample aliquot that was prepared and shipped was assigned a unique sample number and was
accompanied by an EPA traffic report form. Duplicate samples received different sample numbers to retain
the blind sample aspect of the study design. For chronic test methods that required additional shipments for
sample renewal, the sample number remained the same for each initiation and renewal shipment with the
addition of a letter (A, B, and C) after the sample number to designate the sample for use as Initiation (A),
Renewal 1 (B), or Renewal 2 (C). The  sample number was clearly and permanently marked on each container
and the  accompanying EPA traffic report form. Sample  numbers for each test method are given in Table  6.4.
Following completion of the study, each test sample was assigned a sample code (in addition to the sample
number previously assigned) as an alternate unique identifier.  For the results section of this report, samples
are identified by sample codes to aid in blinding the identity of individual participant laboratories.

Referee laboratories included an EPA traffic report form to document the chain-of-custody for each test
sample shipped. The traffic report form (see Appendix B) identified the episode number, sample number,
name and address of the referee laboratory, name and address of the participant laboratory, date shipped,
airbill number, tests requested, and pre-shipment sample information (sample preparation date and initial
water chemistry). A traffic report form specific to each sample was placed in a waterproof enclosure (i.e.,
Ziploc bag) and packed in the cooler with the respective  sample. Each cooler used in the study was
permanently numbered and labeled (with the referee laboratory name and address) to assist in locating lost
coolers and to assist in retrieving coolers from participant laboratories.

For each shipment event, the referee laboratory also completed a sample shipment documentation form. The
referee laboratory faxed this form to SCC immediately after sample pickup by FedEx.  The sample shipment
form documented the following information for each shipping event:
•   Sample number - the unique identifying number for each sample aliquot
•   Sample description - identified the sample as either blank, spiked effluent, spiked receiving water, or
    reference toxicant
    Participant laboratory name - name of the laboratory to which the sample was shipped
•   Airbill number - the overnight shipping service's number that identified each  individual shipment
•   Size of test containers - the volume of the test container in which the sample was shipped
    Cooler number - a unique identifying number for the cooler in which the sample was shipped.
•   Comments - any miscellaneous information related to sample shipment.
                                                60

-------
Table 6.2. Number and volume of samples required for freshwater methods in the WET Study.
Test method
Ceriodaphnia
acute
Ceriodaphnia
chronic
Fathead acute
Fathead
chronic
Selenastrum
chronic
Sample type
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Ampule or whole
volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Sample
volume3
100 mL
lOOmL
1 L
1 L
100 mL
lOOmL
2-3L
2-3L
lOOmL
lOOmL
4L
4L
100 mL
lOOmL
5-7.5L
5-7.5L
200 mL
200 mL
4L
4L
Number of samples required11
Initiation
34
32
28
14
35
49
28
14
28
39
30
14
25
38
28
14
9
14
16
9
Renewal 1
-
-
-
-
35
49
28
14
-
-
-
-
25
38
28
14
-
-
-
-
Renewal 2
-
-
-
-
35
49
28
14
-
-
-
-
25
38
28
14
-
-
-
-
Total
34
32
28
14
105
147
84
42
28
39
30
14
75
114
84
42
9
14
16
9
a For chronic test methods, sample volumes ranged depending upon the aliquot.  More volume was required for the Renewal 2 aliquot than the Initiation or Renewal 1 aliquots.
b Number of samples includes samples shipped round-trip back to referee laboratory.
                                                                       61

-------
Table 6.3. Number and volume of samples required for marine methods in the WET Study.
Test method
Mysid chronic
Sheepshead
acute
Sheepshead
chronic
Silverside
acute
Silverside
chronic
Sample type
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Blank
Reference toxicant
Effluent
Receiving water
Ampule or whole
volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Ampule
Ampule
Whole volume
Whole volume
Sample
volume"
100 mL
lOOmL
6-9L
6-9L
500 mL
500 mL
4L
4L
500 mL
500 mL
10- 15L
10- 15L
lOOmL
lOOmL
4L
4L
lOOmL
lOOmL
14-21 L
14-21L
Number of samples required11
Initiation
9
14
16
9
8
8
8
8
8
8
8
8
7
13
13
7
8
14
14
8
Renewal 1
9
14
16
9
-
-
-
-
8
8
8
8
-
-
-
-
8
14
14
8
Renewal 2
9
14
16
9
-
-
-
-
8
8
8
8
-
-
-
-
8
14
14
8
Total
27
42
48
27
8
8
8
8
24
24
24
24
7
13
13
7
24
42
42
24
a For chronic test methods, sample volumes ranged depending upon the aliquot. More volume was required for the Renewal 2 aliquot than the Initiation or Renewal 1 aliquots.
b Number of samples includes samples shipped round-trip back to referee laboratory.
                                                                       62

-------
SCC entered shipment information into a database and used this information on the day of expected sample
arrival to track the delivery of each sample through the FedEx automated shipment tracking system. If sample
shipment problems were encountered, SCC notified participant laboratories of the problem and instructed the
laboratory how to proceed. The instructions provided by SCC were specific to the individual case, but
laboratories were generally instructed to initiate testing on the day of sample arrival if the sample was
delivered prior to the close of business.  If the delivery of renewal samples was delayed, laboratories were
instructed to renew the test on time with remaining sample from the previous shipment. If sample shipments
could not be located or if tracking did not indicate progress of the sample, the referee laboratory was
instructed to resend the sample (prepared from the remaining bulk sample) for delivery the following day.

Upon receipt of each sample, participant laboratories were responsible for determining that the sample arrived
in satisfactory condition and for documenting receipt of the sample, post-shipment sample water quality
(temperature, pH, dissolved oxygen, and conductivity or salinity), and any problems on the EPA traffic report
form. Laboratories  faxed the completed traffic report form to  SCC immediately upon sample receipt and
retained a copy for inclusion in the data report. SCC's faxed  receipt of the completed traffic report form from
the participant  laboratory served as the notification that the sample had arrived in good condition at the
participant laboratory.

For ampule samples, participant laboratories were not required to measure post-shipment sample water
quality.  To avoid possible contamination between the highly  concentrated reference toxicant ampules and
blank samples, no direct measurements were made on the ampule samples. Temperature was measured in a
temperature check sample that was included with each cooler containing ampule samples. This sample, which
was clearly marked as a temperature check, contained tap water in the same volume and container as ampule
samples.
Table 6.4. Episode numbers and sample numbers used in the WET Study.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum growth
Mysidopsis chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
Episode number
6207
6208
6205
6206
6209
6216
6214
6215
6210
6211
Sample number range
03000-03107
04000-04128
01000-01110
02000-02104
05000 - 05047
12000- 12047
10000- 10031
11000- 11031
06000-06041
07000 - 07043
Sample code range
9217-9324
9325 - 9453
9001 -9111
9112-9216
9454 . 9501
9650 - 9697
9586-9617
9618-9649
9502 - 9541
9542 - 9585
                                                63

-------
6.4     Problems Encountered in Sample Distribution

On the first shipment day for the Ceriodaphnia acute test method (11/08/99), the referee laboratory
inadvertently switched the blank and reference toxicant ampules.  All participant laboratories that were
intended to receive blank samples received reference toxicant samples, and all laboratories that were intended
to receive reference toxicant samples received blank samples.  This error was not noticed at the time, and the
referee laboratory incorrectly reported sample types on the sample shipment documentation form faxed to
SCC. Since all samples were received by participant laboratories as  blind test samples, this error had no
effect on the participant laboratory testing of samples. This error did alter the intended sample distribution
scheme (Table 6.1), however, the number of blank and reference toxicant samples were approximately equal
so effects  on the study design were minimal. The error was first identified by SCC as a result of participant
laboratory data reports.  Test results and conductivity measurements  on the two samples indicated that the
samples were switched.  Since KC1 was used as the reference toxicant, conductivity measurements were used
to properly identify the blank and reference toxicant samples.  Conductivity of the reconstituted blank sample
was approximately 300 ^mhos/cm, and conductivity of the reconstituted reference toxicant sample was
approximately 2900 /^mhos/cm. The referee laboratory determined that the error was caused by inadvertently
filling ampules intended for blank samples with the reference toxicant bulk sample and filling ampules
intended for reference toxicant samples with the blank bulk sample.

Due to weather or other circumstances, sample shipments occasionally failed to arrive at the participant
laboratory on time.  Of the 1438 sample aliquots shipped in the WET Study, 1412 (or 98%) successfully
arrived on the intended delivery date. Tests that were initiated on samples greater than 36 hours old are
identified  with a data qualifier flag  in the results section of this report. No participant laboratory tested
samples that were greater than 72 hours old. In addition, sample types were tested for stability and persistence
of toxicity during the preliminary testing phase, so departure from recommended sample holding times did not
greatly affect test results in the WET Study.

Samples also occasionally arrived at participant laboratories at temperatures above 4°C. Only 7.4% of the
1438 sample aliquots arrived at above the recommended sample shipment temperature. This also had little
effect on test results in the WET Study since sample characteristics were known and selected toxicants were
not likely  to be altered as a result of slight temperature fluctuations.

On seven  occasions, participant  laboratories noted that sample labels were smeared by melting ice in the
coolers and were difficult to read.  This problem was addressed by placing the ice in double-lined plastic bags
within the coolers. On all occasions the sample numbers were identified using the accompanying traffic report
form. Participant laboratories then re-labeled the sample.
                                                 64

-------
7.0 INTERLABORATORY TESTING

Prior to interlaboratory testing of each test method, SCC provided participant laboratories with method-
specific SOPs documenting participant laboratory requirements (see Appendix B).  These SOPs described the
shipment and tracking of test samples, provided instructions for any necessary pre-test sample adjustments
(including preparation of reconstituted ampule samples), provided general and method-specific testing
requirements as described in Section 7.1 and 7.2, and described data reporting requirements.  Participant
laboratories were then provided with samples (prepared according to Section 5) for immediate testing.
Interlaboratory testing was conducted according to the schedule provided in Section 2.2.4 for the
Ceriodaphnia acute and chronic, fathead acute and chronic, Selenastrum chronic, mysid chronic, sheepshead
acute and chronic, and silverside acute and chronic test methods.  Interlaboratory testing was canceled for the
Champia chronic and Holmesimysis acute test methods (see Section 2.1).

7.1    General Testing Requirements

Except where indicated in the SOPs provided to participant laboratories, each test was conducted in
accordance with the general guidance and method specific requirements for effluent testing included in the
WET methods manuals. Additional general WET test requirements that were listed in participant laboratory
SOPs are provided below:

(1)    Tests were required to be conducted by the same laboratory personnel that routinely conduct WET
       tests at that laboratory facility and who were identified in the prequalification materials. The
       laboratory was required to contact SCC if these individuals could not be available during any part of
       the study.  Personnel conducting the tests were to be identified clearly and consistently in records.

(2)    To coordinate testing at participant laboratories, testing of each sample with each method was
       required to be initiated on the precise day specified in the finalized study schedule. The finalized
       study schedule was distributed to participating laboratories prior to commencement of each study
       round and in ample time to prepare for testing. Laboratories were required to test samples within 36
       hours from the time of sample preparation (determined in the WET Study as the time at which
       individual sample aliquots were divided from the bulk test sample for distribution to  participant
       laboratories). Laboratories were required to report deviations from the finalized study schedule to
       SCC immediately for approval.

(3)    Laboratories were required to conduct tests within the physical and chemical water quality ranges
       specified in the study plan, the SOW, specific instructions, and the methods manuals. Method specific
       instructions for any adjustments to the test samples prior to sample use (such as reconstitution of
       ampule samples or salinity adjustments) were  provided to the testing laboratories prior to test
       commencement. Laboratories were required to refrigerate (at 4°C ± 2°C) test samples immediately

                                                 65

-------
        upon receipt and throughout the period of testing. Routine or continuous monitoring of refrigeration
        temperature was recommended to ensure that these sample holding requirements were met.

(4)     Laboratories were required to measure test conditions (pH, conductivity or salinity, total alkalinity,
        total hardness, and dissolved oxygen) in each test according to guidance in the WET method manuals.

(5)     Laboratories were required to use the dilution and control waters specified in Tables 7.1  -7.12 for
        each test method.  Laboratories were required to prepare these dilution waters according to
        instructions in Section 7 of the method manuals.  For marine test methods, laboratories were required
        to prepare dilution waters that meet the salinity ranges specified in Tables 7.6 - 7.12.

(6)     Laboratories were required to conduct all tests as definitive tests with a control and a minimum of five
        test concentrations prepared using a dilution factor of 0.5.

(7)     Laboratories were required to conduct all tests using the number of replicates and number of test
        containers per concentration as specified in Tables 7.1 - 7.12.

(8)     For a given test method, laboratories were required to use the same type, size, shape, and material for
        all test chambers.  The test chamber material used had to be allowed by the WET method manuals.

(9)     Laboratories were required to randomize test vessels in accordance with the WET method manuals. In
        addition, block randomization and use of known parentage were required for the Ceriodaphnia
        chronic test method as described in method manual.  The Agency plans to amend Method 1002.0
        (Ceriodaphnia chronic test method) to require that test organisms be allocated among test replicates
        so that offspring of each female are evenly distributed among test replicates ("blocking-by-
        parentage").

(10)    While the method manual requires the termination of the Ceriodaphnia chronic test after the
        production of three broods in 60% of the controls, laboratories were required to conduct the
        Ceriodaphnia chronic test for eight days. Laboratories were required to record the survival, number
        of young per day, and number of broods at the end of Day 6, 7, and 8 (specifically, at 144 hours, at
        168 hours, and at  192 hours, respectively, from test initiation).  This was done to assess  the effect of
        the three brood test acceptance criterion on test results.  No test was allowed to be terminated prior to
        Day 8 for any reason, including a failure to meet test acceptance criteria. Laboratories were required
        to include the additional measurements on Days 6, 7, and 8 as raw data in the final data  report.
        However, laboratories were required to analyze data from the Ceriodaphnia chronic test using the
        three brood approach as specified in the method manual.
                                                 66

-------
(11)    Laboratories were required to conduct the Selenastrum chronic test simultaneously with and without
        EDTA for each sample.  For laboratories participating in the base study design, four samples were
        tested with and without EDTA for a total of eight analyses.

(12)    Laboratories were required to daily observe mortality and remove dead organisms in each test, except
        for the Selenastrum chronic and Champia chronic test methods. For the Ceriodaphnia chronic test
        method, laboratories were required to daily count young and determine the number of broods at each
        count.

(13)    Laboratories were required to contact SCC immediately if test results indicated extreme toxicity (i.e.,
        control mortalities, or complete mortality in all concentrations).  Laboratories were then required to
        investigate possible causes, first by checking for transcription and calculation mistakes, and then by
        investigating possible contamination in dilution waters, organism cultures, equipment, or other
        procedural steps.

(14)    If any initiated test failed to be completed for any reason, the laboratory was required to contact SCC
        immediately for problem resolution and scheduling of additional testing.  In this case, laboratories
        were required to report the incomplete test data and fully document the reason for not completing the
        test in the final report.

(15)    Laboratories were required to report all data obtained during the course of testing, including the
        response of control samples.

(16)    Laboratories were required to perform all QA/QC tests listed in Section 4 of the method manuals.
        Laboratories that purchased organisms were required to supply QA/QC from the test organism
        supplier and follow method manuals for the appropriate QA/QC for purchasing organisms.

(17)    Laboratories were required to perform a reference toxicant QC test  for each test method in the month
        that interlaboratory testing occurred. Results of this test had to be submitted with the final data
        package.

(18)    Laboratories were required to submit hard copies of all data and statistical  analyses. All bench sheets
        and raw data, including sample tracking and chemistry analysis data were required. Laboratories also
        were required to submit data electronically according to the electronic template (Microsoft Excel®
        spreadsheet) that was provided by SCC prior to test initiation.
                                                 67

-------
(19)    Laboratories were required to analyze data in accordance with the statistical programs specified in the
        method manuals. Statistical methods and programs used had to be reported along with sample
        calculations.

(20)    Laboratories were required to report an LC50 for each acute test. A NOEC and LC50 for survival,
        and a NOEC and IC25 for growth/reproduction was required as appropriate for each short-term
        chronic test as described in the method manuals and Table 2.3 of this report. Laboratories were
        required to report individual toxicity endpoints and were not allowed to average or perform other data
        manipulations unless required by a method's instructions.

7.2     Method-Specific Requirements

EPA acknowledges that the promulgated WET methods distinguish between requirements (indicated by the
compulsory terms "must" and "shall") and recommendations and guidance (indicated by discretionary terms
"should" and "may"). The latter terms indicate that the analyst has flexibility to optimize successful test
completion. Additionally, the method manuals allow variations of the methods that are typically fixed in the
permit; therefore, for the purposes of this study, a set of test condition variables were defined by EPA (for
example, dilution water, salinity, and acute test duration).

The summary of test conditions for the 12 WET methods evaluated in the WET Study are provided in Tables
7.1 - 7.12.  These tables were extracted from the summary test condition tables in the WET method manuals
and modified to fit the scope of this study. Items  shown in bold italic in these tables represent conditions
standardized for the purposes of this study where  method manuals provide a range.  These test conditions were
reiterated in participant laboratory SOPs (Appendix B).
                                                68

-------
Table 7.1. Summary of test conditions and test acceptability criteria for the Ceriodaphnia acute test method.
 1. Test type:
2. Test duration:
3. Temperature:
4. Light quality:
5. Light intensity:
6. Photoperiod:
7. Test chamber size:
8. Test solution volume:
9. Renewal of test solutions:
 10.  Age of test organisms:
 11.  No. organisms per test chamber:
 12.  No. replicate chambers per concentration:
 13.  No. organisms per concentration:
 14.  Feeding regime:

15.  Test chamber cleaning:
16.  Test chamber aeration:
17.  Dilution water:

18.  Test concentrations:
19.  Dilution factor
20.  Endpoint:
21.  Sample handling and holding requirements:

22.  Sample volume required:
23.  Test acceptability criterion:
Static non-renewal
48 h
25±1°C
Ambient laboratory illumination
10-20 uE/m2/s (50-100 ft-c) (ambient laboratory levels)
16 h light, 8 h darkness
30 ml
15 ml
None
Less than 24-h old
5
4
20
Feed YCT and Selenastrum while holding prior to the test; newly-released young should have food
available a minimum of 2 h prior to use in a test.
Cleaning not required
None
Moderately hard synthetic water prepared using MILLIPORE MILLI-Q® or equivalent deionized
water and reagent grade chemicals (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
>0.5
Mortality (LC50)
Samples treated as effluent samples for NPDES monitoring.  Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
1L
90% or greater survival in controls
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                        69

-------
Table 7.2.  Summary of test conditions and test acceptability criteria for the Ceriodaphnia chronic test method.
1. Test type:
2. Temperature (°C):
3. Light quality:
4. Light intensity:
5. Photoperiod:
6. Test chamber size:
7. Test solution volume:
8. Renewal of test solutions:
9. Age of test organisms:
10.  No. neonates per test chamber:1
11.  No. replicate test chambers per concentration:
12.  No. neonates per test concentration:
13.  Feeding regime:
14.  Cleaning:
15.  Aeration:
16.  Dilution water:

17.  Test concentrations:
18.  Dilution factor:
19.  Test duration:2
20.  Endpoints:
21.  Test acceptability criteria:

22.  Sample handling and holding requirements:

23.  Sample volume required:
Static renewal
25±1°C
Ambient laboratory illumination
10-20 uE/m2/s, or 50-100 ft-c (ambient laboratory levels)
16 h light, 8 h dark
30 ml
15 mL
Daily
Less than 24 h; and all released within a 8-h period
1
10
10
Feed 0.1 mL each of YCT and algal suspension per test chamber daily.
Use freshly cleaned glass beakers or new plastic cups daily
None
Moderately hard synthetic water prepared using MILLIPORE MILLI-Q® or equivalent deionized
water and reagent grade chemicals (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
>0.5
8 days
Survival and reproduction
80% or greater survival and an average of 15 or more young per surviving female in the control
solutions. 60% of surviving control organisms must produce three broods.
Samples treated as effluent samples for NPDES monitoring.  Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
1 L/day
1 Test vessels shall be randomized in accordance with the method manuals. In addition, block randomization and use of known parentage will be required for the Ceriodaphnia survival
and reproduction test as described in the manual and guidance will be reiterated in the specific instructions provided to the laboratories.
2The Ceriodaphnia dubia test which would otherwise be terminated after 3 broods according to methods manual Section 13.12.1 of that Method must be conducted for 8 days, with
endpoints (survival and number of young per day and number of broods at each recording interval) recorded at the end of the 6th, 7th and 8th day (specifically, at 144, 168, and 192
hours, respectively, from test initiation). No test shall be terminated prior to the 8th day for any reason, including a failure to meet test acceptance criteria.
                                                                           70

-------
Table 7.3. Summary of test conditions and test acceptability criteria for the fathead acute test method.
1. Test type:
2. Test duration:
3. Temperature:
4. Light quality:
5. Light intensity:
6. Photoperiod:
7. Test chamber size:
8. Test solution volume:
9. Renewal of test solutions:
10. Age of test organisms:
11. No. organisms per test chamber:
12. No. replicate chambers per concentration:
13. No. organisms per concentration:
14. Feeding regime:

15. Test chamber cleaning:
16. Test solution aeration:
17. Dilution water:

18. Test concentrations:
19. Dilution factor
20. Endpoint:
21. Sample handling and holding requirements:

22. Sample volume required:
23. Test acceptability criterion:
Static-renewal
96 h
25±1°C
Ambient laboratory illumination
10-20 uE/m2/s (50-100 ft-c) (ambient laboratory levels)
16 h light, 8 h darkness
250 mL
200 mL
At48h
1-14 days; 24-h range in age
10
2
20
Anemia nauplii are made available while holding prior to the test; add 0.2 mL Anemia nauplii
concentrate 2 h prior to test solution renewal at 48 h
Cleaning not required
None, unless DO concentration falls below 4.0 mg/L; rate should not exceed 100 bubbles/min
Moderately hard synthetic water prepared using MILLIPORE MILLI-Q® or equivalent deionized
water and reagent grade chemicals (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
>0.5
Mortality (LC50)
Samples treated as effluent samples for NPDES monitoring.  Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
2 Lfor effluents
90% or greater survival in controls
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         71

-------
Table 7.4. Summary of test conditions and test acceptability criteria for the fathead chronic test method.
1. Test type:
2. Temperature:
3. Light quality:
4. Light intensity:
5. Photoperiod:
6. Test chamber size:
7. Test solution volume:
8. Renewal of test solutions:
9. Age of test organisms:
10. No. larvae per test chamber:
11. No. replicate chambers per concentration:
12. No. larvae per concentration:
13. Source of food:
14. Feeding regime:
15. Cleaning:
16. Aeration:
17. Dilution water:

18. Test concentrations:
19. Dilution factor
20. Test duration:
21. Endpoints:
22. Test acceptability criteria:

23. Sample handling and holding requirements:

24. Sample volume required:
Static renewal
25 ± 1°C
Ambient laboratory illumination
10-20 uE/m2/s (50-100 ft-c) (ambient laboratory levels)
16 h light, 8 h darkness
500 ml
250 mL
Daily
Newly hatched larvae less than 24h old.  If shipped, not more than 48h old, 24h range in age
10
4
40
Newly hatched Artemia nauplii (less than 24 h old)
Feed 0.1 g newly hatched (less than 24-h old) brine shrimp nauplii three times daily at 4-h intervals
or, as a minimum, 0.15 g twice daily, 6 h between feedings (at the beginning of the work day prior to
renewal, and at the end of the work day following renewal). Sufficient nauplii are added to provide
an excess. Larvae fish are not fed during the final 12  h of the test
Siphon daily, immediately before test solution renewal
None, unless DO concentration falls below 4.0 mg/L. Rate should not exceed 100 bubbles/min
Moderately hard synthetic water prepared using MILLIPORE MILLI-Q® or equivalent deionized
water and reagent grade chemicals (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
>0.5
7 days
Survival and growth (weight as mean per original)
80% or greater survival in controls; average dry weight per surviving organism in control chambers
equals or exceeds 0.25 mg/surviving
Samples treated as effluent samples for NPDES monitoring. Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
2.5 L/day
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         72

-------
Table 7.5. Summary of test conditions and test acceptability criteria for the Selenastrum chronic test method.
1. Test type:
2. Temperature:
3. Light quality:
4. Light intensity:
5. Photoperiod:
6. Test chamber size:
7. Test solution volume:
8. Renewal of test solutions:
9. Age of test organisms:
10.  Initial cell density in test chambers:
11.  No. replicate chambers per concentration:
12.  Shaking rate:
13.  Aeration:
14.  Dilution water:


15.  Test concentrations:
16.  Test dilution  factor:
17.  Test duration:
18.  Endpoint:
19.  Test acceptability criteria:

20.  Sample handling and holding requirements:

21.  Sample volume required:
Static non-renewal
25 ± 1°C
"Cool white" fluorescent lighting
86 ± 8.6 uE/m2/s (400 ± 40 ft-c or 4306 lux)
Continuous illumination
250 ml
100 ml
None
4 to 7 days
10,000 cells/mL
4
100 cpm continuous
None
Algal stock culture medium, moderately hard synthetic water prepared using MILLIPORE MILLI-
Q® or equivalent deionized water and reagent grade chemicals(see Methods Manual Section 7,
Dilution Water)
Five concentrations and a control
>0.5
96 h
Growth (cell counts)
1 X 106 cells/mL with EDTA or 2 X 105 cells/mL without EDTA in the controls; Variability of controls
should not exceed 20%
Samples treated as effluent samples for NPDES monitoring. Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
2L
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                        73

-------
Table 7.6.  Summary of test conditions and test acceptability criteria for the mysid chronic test method.
1. Test type:
2. Salinity:
3. Temperature:
4. Light quality:
5. Light intensity:
6. Photoperiod:
7. Test chamber:
8. Test solution volume:
9. Renewal of test solutions:
10. Age of test organisms:
11. No. organisms per test chamber:
12. No. replicate chambers per concentration:
13. No. larvae per concentration:
14.  Source of food:
15.  Feeding regime:
16.  Cleaning:
17.  Aeration:
18.  Dilution water:

19.  Test concentrations:
20.  Dilution factor:
21.  Test duration:
22.  Endpoints:
23.  Test acceptability criteria:

24.   Sample handling and holding requirements:

25.  Sample volume required:
Static renewal
25%o  (±2%o)
26 ± 1°C
Ambient laboratory illumination
10-20 uE/m2/s (50-100 ft-c.) (ambient laboratory levels)
16 h light, 8 h darkness, with phase in/out period
8 oz plastic disposable cups, or 400 mL glass beakers
150 mL per replicate
Daily
7 days
5
8
40
Newly hatched Artemia nauplii (less than 24 h old)
Feed 150 24 h old nauplii per mysid daily, half after test solution renewal and half after 8-12 h.
Pipette excess food from cups daily immediately before test solution renewal and feeding.
None unless DO falls below 4.0 mg/L, then  gently aerate in all cups
25%o  (±2%o ) Bioassay Grade Forty Fathoms® artificial seawater prepared with MILL1PORE
MILLI-Q® or equivalent deionized water (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
0.5
7 days
Survival, growth, and egg development
80% or greater survival, average dry weight 0.20 mg or greater in controls; fecundity may be used if
50% or more of females in  controls produce eggs
Samples treated as effluent samples for NPDES monitoring. Samples are to be used on the day
specified by the finalized interlab oratory study testing schedule
3 L per day
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         74

-------
Table 7.7. Summary of test conditions and test acceptability criteria for the sheepshead acute test method.
1. Test type:
2. Test duration:
3. Temperature:
4. Light quality:
5. Light intensity:
6. Photoperiod:
7. Test chamber size:
8. Test solution volume:
9. Renewal of test solutions:
10. Age of test organisms:
11. No. organisms per test chamber:
12. No. replicate chambers per concentration:
13. No. organisms per concentration:
14. Feeding regime:

15. Test chamber cleaning:
16. Test solution aeration:
17. Dilution water:

18. Test concentrations:
19. Dilution factor
20. Endpoint:
21. Sample handling and holding requirements:

22. Sample volume required:
23. Test acceptability criterion:
24. Salinity:
Static renewal
96 h
25°C±1°C
Ambient laboratory illumination
10-20 uE/m2/s or (50-100 ft-c) (ambient laboratory levels)
16 h light, 8 h darkness
250 ml
200 ml
At48h
1-14 days; 24-h range in age
10
2
20
Artemia nauplii are made available while holding prior to the test; add 0.2 mL Anemia nauplii
concentrate 2 h prior to test solution renewal at 48 h
Cleaning not required
None, unless DO concentration falls below 4.0 mg/L; rate should not exceed 100 bubbles/min
25 %o ± 2%o Bioassay Grade Forty Fathoms® artificial seawater prepared with MILLIPORE
MILLI-Q® or equivalent deionized water (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
0.5
Mortality (LC50)
Samples treated as effluent samples for NPDES monitoring.  Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
1 Lfor effluents
90% or greater survival in controls
25%o (±2%o )
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         75

-------
Table 7.8. Summary of test conditions and test acceptability criteria for the sheepshead chronic test method.
1. Test type:
2. Salinity:
3. Temperature:
4. Light quality:
5. Light intensity:
6. Photoperiod:
7. Test chamber size:
8. Test solution volume:
9. Renewal  of test solutions:
10. Age of test organisms:
11. No. larvae per test chamber:
12. No. replicate chambers per concentration:
13. No. larvae per concentration:
14. Source of food:
15. Feeding regime:

16. Cleaning:
17. Aeration:

18. Dilution  water:

19. Test concentrations:
20. Dilution  factor:
21. Test duration:
22. Endpoints:
23. Test acceptability criteria:
24. Sample handling and holding requirements:

25. Sample volume required:
Static renewal
25%o (±2%o )
25 ± 1°C
Ambient laboratory illumination
10-20 uE/m2/s (50-100 ft-c) (ambient laboratory levels)
16 h light, 8 h darkness
600 mL beaker
500 mL/replicate (loading and DO restrictions must be met)
Daily
Newly hatched larvae (less than 24 h old; 24-h range in age)
10
4
40
Newly hatched Artemia nauplii, (less than 24-h old)
Feed once a day 0.10 g wet weight Anemia nauplii per replicate on Days 0-2; Feed 0.15 g wet weight
Artemia nauplii per replicate on Days 3-6
Siphon daily,  immediately before test solution renewal and feeding
None, unless DO falls below 4.0 mg/L, then aerate all chambers. Rate should be less than 100
bubbles/min
25%o (±2%o ) Bioassay Grade Forty Fathoms® artificial seawater prepared with MILLIPORE
MILLI-Q® or equivalent deionized water (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
0.5
1 days
Survival and growth (weight)
80% or greater survival in controls; average dry weight per surviving organism in control chambers
should be 0.60 mg or greater, if unpreserved, or 0.50 mg or greater after no more than 7 days in 4%
formalin or 70% ethanol
Samples treated as effluent samples for NPDES monitoring. Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
6 L per day
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         76

-------
Table 7.9. Summary of test conditions and test acceptability criteria for the silverside acute test method.
1. Test type:
2. Test duration:
3. Temperature:
4. Light quality:
5. Light intensity:
Static-ren ewal
96 h
25°C±1°C
Ambient laboratory illumination
10-20 ,uE/m2/s (50-100 ft-c) (ambient laboratory levels)
6. Photoperiod:
7. Test chamber size:
8. Test solution volume:
9. Renewal of test solutions:
10. Age of test organisms:
11. No. organisms per test chamber:
12. No. replicate chambers per concentration:
13. No. organisms per concentration:
14. Feeding regime:

15. Test chamber cleaning:
16. Test solution aeration:
17. Dilution water:

18. Test concentrations:
19. Dilution factor:
20. Endpoint:
21. Sample handling and holding requirements:

22. Sample volume required:
23. Test acceptability criterion:
24. Salinity:
16 h light, 8 h darkness
250 ml
200 ml
At48h
9-14 days; 24-h range in age
10
2
20
Artemia nauplii are made available while holding prior to the test; add 0.2 mL Anemia nauplii
concentrate 2 h prior to test solution renewal at 48 h
Cleaning not required
None, unless DO concentration falls below 4.0 mg/L; rate should not exceed 100 bubbles/min
25%o (±2%o ) Bioassay Grade Forty Fathoms® artificial seawater prepared with MILLIPORE
MILLI-Q® or equivalent deionized water (see Methods Manual Section 7, Dilution  Water)
Five concentrations and a control
>0.5
Mortality (LC50)
Samples treated as effluent samples for NPDES monitoring.  Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
1 L for effluents
90% or greater survival in controls
     (± 2%o )
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         77

-------
Table 7.10.  Summary of test conditions and test acceptability criteria for the silverside chronic test method.
1.   Test type:
2.   Salinity:
3.   Temperature:
4.   Light quality:
5.   Light intensity:
6.   Photoperiod:
7.   Test chamber size:
8.   Test solution volume:
9.   Renewal of test solutions:
10. Age of test organisms:
11. No. larvae per test chamber:
12. No. replicate chambers per concentration:
13. No. larvae per concentration:
14. Source of food:

15. Feeding regime:

16. Cleaning:
17. Aeration:

18. Dilution water:

19. Test concentrations:
20. Dilution factor:
21. Test duration:
22. Endpoints:
23. Test acceptability criteria:


24. Sample handling and holding requirements:

25. Sample volume required:
Static renewal
25%o  (±2%o)
25 ± 1°C
Ambient laboratory illumination
10-20  uE/m2/s (50-100 ft-c) (Ambient laboratory levels)
16 h light, 8 h darkness
/ L containers
750 mL/replicate (loading and DO restrictions must be met)
Daily
7-11 days post hatch; 24-h range in age
10
4
40
Newly hatched Artemia nauplii (survival of 7-9 days old Menidia beryllina larvae improved by feeding
24 h old Artemia)
Feed 0.10 g wet weight Artemia nauplii per replicate on days 0-2; Feed 0.15 g wet weight Artemia
nauplii per replicate on days 3-6
Siphon daily, immediately before test solution renewal and feeding
None, unless DO concentration falls below 4.0 mg/L, then aerate all chambers. Rate should be less than
lOObubbles/min.
25%o  (±2%o ) Bioassay Grade Forty Fathoms® artificial seawater prepared with MILLIPORE
MILLI-Q® or equivalent deionized water (see Methods Manual Section 7, Dilution Water)
Five concentrations and a control
>0.5
7 days
Survival and growth (weight)
80% or greater survival in controls, 0.50 mg average dry weight of control larvae when larvae dried
immediately after test termination, or 0.43 mg or greater average dry weight of control larvae, preserved
not more than 7 days  in 4% formalin or 70% ethanol
Samples treated as effluent samples for NPDES monitoring.  Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
6 L/day
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         78

-------
Table 7.11.  Summary of test conditions and test acceptability criteria for the Champia chronic test method.
1. Test type:                                             Static non-renewal
2. Salinity:                                              30%, (± 2%o)
3. Temperature:                                          23±1°C
4. Photoperiod:                                          16 h light, 8 h darkness
5. Light intensity:                                        75 uE/m2/s (500 ft-c)
6. Light source:                                          Cool-white fluorescent lights
7. Test chamber size:                                     200 mL polystyrene cups, or 250 mL Erlenmeyer flasks
8. Test solution volume:                                   100 mL
9. No. organisms per test chamber:                         5 female branch tips and 1  male plant
10. No. replicate chambers per concentration:                4
11. No. organisms per concentrations:                      24
12. Dilution water:                                       30%o salinity natural seawater
13. Test concentrations:                                   Five concentrations and a control
14. Test dilution factor:                                   >0.5
15. Test duration:                                         2 day exposure to effluent, followed by 5 to 7-day recovery period in control medium for cyctocarp
                                                        development
16. Endpoints:                                           Reduction in cystocarp production compared to controls
17. Test acceptability criteria:                              80% or greater survival, and an average of 10 cystocarps per plant in controls
18. Sample handling and holding requirements:              Samples treated as effluent samples for NPDES monitoring. Samples are to be used on the day
                                                        specified by the finalized interlaboratory study testing schedule
19. Sample volume required:                              2 L per test
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         79

-------
Table 7.12. Summary of test conditions and test acceptability criteria for the Holmesimysis acute test method. The acute test procedure described in
the Acute Methods Manual for Mysidopsis bahia will be used for this test with a salinity of 32%o  (± 2%o) and a temperature of 12 °C ± 1 °C.
1. Test type:
2. Test duration:
3. Temperature:
4. Light quality:
5. Light intensity:
6. Photoperiod:
7. Test chamber size:
8. Test solution volume:
9. Renewal of test solutions:
10. Age of test organisms:
11. No. organisms per test chamber:
12. No. replicate chambers per concentration:
13. No. organisms per concentration:
14. Feeding regime:

15. Test chamber cleaning:
16. Test solution aeration:
17. Dilution water:
18. Test concentrations:
19. Dilution factor:
20. Endpoint:
21. Sample handling and holding requirements:

22. Sample volume required:
23. Test acceptability criterion:
24. Salinity:
Static-renewal
96 h
12°C±1°C
Ambient laboratory illumination
10-20 |/E/m2/s (50-100 ft-c) (ambient laboratory levels)
16 h light, 8 h darkness
250 ml
200 ml
At48h
1 -5 days; 24-h range in age
10
4
40
Artemia nauplii are made available while holding prior to the test; feed 0.2 mL of concentrated
suspension of Artemia nauplii < 24-h old, daily (approximately 100 nauplii per mysid)
Cleaning not required
None, unless DO concentration falls below 4.0 mg/L; rate should not exceed 100 bubbles/min
32%o salinity natural seawater
Five concentrations and a control
>0.5
Mortality (LC50)
Samples treated as effluent samples for NPDES monitoring. Samples are to be used on the day
specified by the finalized interlaboratory study testing schedule
1 L for effluents
90% or greater survival in controls
32%o (±2%o)
NOTE: Test vessels shall be randomized in accordance with the method manuals.
                                                                         80

-------
8.0 DATA REPORTING AND EVALUATION

8.1     Report Submission

Within 30 days following the completion of interlaboratory testing for a given method, each laboratory was
required to submit a data report detailing the conduct and results of WET testing completed on each sample.
Table 8.1 lists the report due dates and number of reports received for each test method.  Reports were received
for all samples from all participating laboratories with the exception of one participant laboratory (laboratory ID
#3) for the Ceriodaphnia chronic method.  Participant laboratory #3 received samples for the Ceriodaphnia
chronic method and initiated testing. During testing of the final two of three samples received, an overnight
power failure caused the malfunction of water baths containing test chambers.  The malfunctioning water baths
over-heated test chambers, killing all test organisms (including controls) and terminating the test prematurely. The
laboratory was unable to retest within reasonable sample holding times, so all further testing of this method was
canceled at this laboratory.  The laboratory was non-EPA-sponsored, and the sponsor declined to reimburse the
laboratory for testing costs since all tests were not completed. In turn, the laboratory declined to submit a data
report for this test method.  As a result, interlaboratory data for the Ceriodaphnia chronic method include results
from one referee laboratory and 34 of the 35 participant laboratories.

According to the Participant Laboratory Statement of Work and SOP (Appendix B), each data report was
required to consist of:
        Narrative summary of testing - The narrative summary was intended to quickly and clearly identify the
        laboratory, test method, samples tested, summarized test results, any problems associated with the
        samples or conduct of the tests, any modifications from approved procedures, and any laboratory
        comment on the performance of the method.
•       Hardcopy results - This deliverable consisted of the items outlined in Table 8.2, all raw data (biological
        and chemical), all laboratory benchsheets, all pertinent sample information including copies of all
        completed EPA traffic report forms, and all pertinent quality assurance information including results of
        the monthly QA/QC reference toxicant tests.
•       Electronic results - Laboratories also were required to submit selected raw and summarized data
        electronically using method-specific Microsoft Excel® spreadsheet templates that were provided to
        participant laboratories by SCC.  Electronic data included general information, sample collection/receipt
        information, test condition information, raw biological data, raw water quality data, and summarized test
        results. Electronic data submission facilitated automated review and statistical analysis of study results.
                                                  81

-------
Table 8.1. Report due dates.
Due date
11/12/99
12/02/99
12/03/99
12/08/99
12/13/99
12/13/99
04/06/00
04/17/00
05/03/00
05/04/00
Test method
Fathead chronic
Silverside chronic
Ceriodaphnia chronic
Fathead acute
Ceriodaphnia acute
Silverside acute
Mysid chronic
Sheepshead acute
Selenastrum chronic
Sheepshead chronic
Number of reports received3
28
11
35
30
29
10
12
8
12
8
"The number of laboratory reports received includes participant and referee laboratories.
Table 8.2. Data reporting elements.
 Section 1 - Summary Page
      1.1   Laboratory name
      1.2   Laboratory address and phone number
      1.3   Name and signature of laboratory QA Officer, certifying that data have been internally reviewed and that
           personnel meticulously followed the methods, and the procedures are deemed to be compliant with the
           methods and acceptable for reporting purposes
      1.4   Laboratory contact responsible for study
      1.5   Analyst(s) who performed WET tests (full name)
      1.6   Toxicity tests performed
      1.7   Detailed explanations of any difficulties encountered and any approved modifications to the techniques
           specified in the SOW, specific instructions, or the methods manuals
      1.8   Number of successful tests completed
 Section 2 - Sample Information
     2.1   Number of samples received and EPA sample number assigned to each sample
     2.2   Dates of sample receipt
     2.3   Sample temperature when received at laboratory
     2.4   Physical and chemical data of sample contents (as required in appropriate method)
     2.5   Dilution water
           2.5.1    Source and time frame water is used or how maintained
           2.5.2    Collection or preparation date(s), where applicable
           2.5.3    Pretreatment information
           2.5.4    Physical and chemical characteristics (pH, hardness, conductivity, salinity, etc.)
     2.6   Sample storage information
     2.7   Sample preparation for testing information
                                                      82

-------
Table 8.2. Data reporting elements, (continued)
Section 3 - Test Conditions
    3.1   Toxicity test method used (title, number, source)
    3.2   Endpoint(s) of test(s)
    3.3   Deviations from reference method(s), if any, and reason(s)
    3.4   Date and time test(s) started, date and time samples were prepared and solutions transferred for renewals
    3.5   Date and time test(s) terminated
    3.6   Type and volume of test chambers
    3.7   Volume of solution used per chamber
    3.8   Number of organisms per test chamber
    3.9   Number of replicate test chambers per treatment
    3.10  Feeding frequency and amount and type of food (be specific with sources, concentrations of foods (i.e, algae
          concentration, YCT solids level, preparation dates)
    3.11  Acclimation of test organisms (temperature mean and range and, where applicable, salinity mean and
          range)
    3.12  Test temperature (mean and range)
    3.13  Test salinity, where applicable (mean and range)
    3.14  Specify if aeration was needed
    3.15  Specify if organisms were dried immediately for weighing or preserved prior to drying
    3.16  Specify how food was prepared and sources of food. Include test results that validate the quality of batch
          food preparations (i.e., Ceriodaphnia dubia tests on YCT preparation)
    3.17  Describe how routine chemistries on new solutions were made (in actual test chamber or in beakers after
          dispensing)
    3.18  Describe how randomization was conducted, especially blocking and known parentage; report how brood
          distinctions were made  and  male (if any) identification was made
Section 4 - Test Organisms
    4.1   Scientific name of test species, verification of species documented
    4.2   Age (life stage) of test species (be specific for all species); age at time of test initiation (for example, for C.
          dubia be sure to clarify the window of age of the neonates as well as the overall age of the animals)
    4.3   Mean length and weight (where applicable)
    4.4   Source and QA/QC test conditions
    4.5   Holding Conditions
    4.6   Diseases and treatment (where applicable)
    4.7   Taxonomic key used for species identification
Section 5 - Quality Assurance
    5.1    Reference toxicant used routinely; source; date received; lot number
    5.2   Date and time of most recent reference toxicant test; test results and current control (cusum) chart including
          20 most recent data points
    5.3   Dilution water used in reference toxicant tests (with characteristics provided)
    5.4   Physical and chemical methods used
    5.5   Reference toxicant results (NOEC, IC25, or LC50 where applicable, LOEC or EC50)
                                                     83

-------
 Table 8.2. Data reporting elements, (continued)
 Section 6 - Results
     6.1    Copies of all bench sheets. Be sure to count and notate broods for reproduction test with Ceriodaphnia
     6.2   Raw toxicity data in tabular form, including daily records of affected organisms in each replicate at each
           concentration (including controls) and plots of toxicity data
     6.3   Table of endpoints (LC50, IC25, NOEC for each endpoint) and confidence limits (where applicable)
     6.4   Statistical methods and software used to calculate endpoints
     6.5   Summary table of physical and chemical data	
8.2       Data Review

8.2.1      Data Package Receipt and Gross Completeness Check

Data reports from all laboratories were submitted to SCC for review and verification of test results. Upon
receipt, SCC personnel date stamped data packages and performed an initial review to ensure that all required
information was provided.  Most laboratories reported results using their standard reporting formats rather than
the suggested format in Table 8.2.  This was acceptable provided that all pertinent information was included.  If
necessary information was not provided in the data report, SCC personnel contacted the laboratory and asked
them to supply the additional information.

8.2.2      Data Accuracy and Quality Check

Following initial review of data packages for completeness, SCC personnel performed a detailed review of data
reports to ensure that data were  accurate and generated in accordance with the required procedures. The
following review steps were completed for each data report:
       Cross reference of raw data  - Raw data on submitted electronic benchsheets were compared to
       hardcopy laboratory benchsheets to ensure that no transcription or data entry errors occurred. Every
       entry of biological raw data (including individual replicate values of daily survival, daily reproduction,
       and weight) was cross referenced against hardcopy benchsheets to ensure accuracy.  Sample collection,
       test condition, and water quality raw data entered on electronic  benchsheets were spot checked against
       hardcopy benchsheets.  If errors were found during the spot check, more intensive review was initiated.
       All data fields that triggered automated data qualifier flags in the electronic benchsheet were also
       individually compared to hardcopy benchsheets to ensure accuracy.  When errors or inconsistencies were
       identified, the electronic benchsheet was corrected to match the hardcopy laboratory benchsheets.
•       Narrative summary and hardcopy report review - SCC reviewed the narrative summary and hardcopy
       report to ensure that tests were conducted in accordance with the WET method manuals, the WET Study
       plan, and guidance provided in method-specific SOPs. If any deviations from the required or
       recommended procedures were identified, SCC data reviewers verified that those deviations were
       captured by automated or manual data qualifier flags in the electronic benchsheet.
                                                   84

-------
       Data qualifier flags - The electronic benchsheets for each method were programmed to automatically
       identify and flag excursions in sample holding times or sample temperatures, failure of test acceptability
       criteria, excursions in test or dilution water quality, and test condition deviations.  For all flags that were
       automatically identified, SCC evaluated the electronic benchsheet and hardcopy report to ensure that the
       flag was warranted. Additional parameters, test conditions, and comments that were unable to be
       programmed for automated review were checked and flagged manually if necessary. Table 8.3 lists and
       describes the categories of data qualifier flags used in the WET Study.

       Tests that were flagged for a failure to meet test acceptability criteria (flags a,, a2, a3, a4, a6, and a7) were
       considered invalid and were not included in the analysis of method performance (see Section 9.1.1). The
       large number of other flags (b, - g,2) described in Table 8.3 and identified for particular tests in Section
       9, emphasizes the extensive nature of test review rather than deficiencies in test conduct or quality
       control. While the presence of these flags may indicate a deviation from optimal test conditions, their
       presence alone was not used to invalidate and exclude test data from the analysis of method performance.
       Rather, these flags were used to identify potential causes for aberrant test results and to support inclusion
       or exclusion of data in outlier analyses (see Section 9.1.4).

       Reference toxicant test review - Laboratories were required to conduct a reference toxicant test within
       the month of testing for the WET Study and provide the test results and current control chart with the
       data report. SCC data review staff verified that a reference toxicant test was conducted during the
       required time frame and reviewed the laboratory's control chart to ensure that the  current test result fell
       within the control chart limits (2 standard deviations for point estimates and 1 concentration interval for
       hypothesis testing endpoints). If the reference toxicant test fell outside of the control chart limits or was
       not conducted during the required time frame, a data qualifier flag for reference toxicant testing was
       associated (in the results database) with all test results from this laboratory for the given method.
Table 8.3. Test data qualifier flags.
Flag code
Flag Description
Test acceptability
a.
a2
a.i
a4
as
a<,
a7
Survival of control organisms failed to meet the minimum test acceptability criteria for the method.
Growth of control organisms (measured as the mean weight of surviving control organisms) failed to
test acceptability criteria for the method.
Reproduction of control organisms in the Ceriodaphnia chronic test failed to meet the minimum test
criteria requiring that 60% of surviving control organisms have 3 broods prior to test termination at 8
meet the minimum
acceptability
days.
Mean reproduction of surviving control organisms in the Ceriodaphnia chronic test failed to meet the minimum test
acceptability criteria of 15 neonates.
Fecundity endpoints were not generated since less than 50% of control females in the mysid chronic
test produced eggs.
Mean algal growth of control organisms in the Selenastrum chronic test did not meet the minimum test acceptability
criteria requiring a mean cell density of IxlO6 cells/mL with EDTA or 2xl05 cells/mL without EDTA.
Algal growth variability between control replicates in the Selenastrum chronic test did not meet the
acceptability criteria requiring less than 20% variability (measured as % CV).
minimum test
                                                  85

-------
Table 8.3. Test data qualifier flags (continued)
Flag code
Flag Description
Sample Receipt
b,
b2
b,
b4
b5
Sample temperature was >4°C upon arrival at the participant lab.
Sample was >36 hr old at the time of test initiation.
Sample was >72 hr old at the time of test initiation.
Sample was aerated upon receipt due to over saturation of dissolved oxygen.
Sample was inadvertently delivered to and opened by a laboratory not involved in the study. The sample was quickly
rerouted to the correct laboratory and tests were initiated on time.
Dilution Water
C|
C2
C3
C4
C5
Dilution water used for the test was different from that required in the study.
Alkalinity of dilution water was >10% outside of recommended ranges for moderately hard synthetic water.
Hardness of dilution water was >10% outside of recommended ranges for moderately hard synthetic water.
pH of dilution water was >10% outside of recommended ranges for moderately hard synthetic water.
Salinity of dilution water was outside of the range required in the study.
Water Quality
d,
d2
d,
d4
d5
d()
d7
Temperature of one or more test concentrations was outside of range required in the study.
Dissolved oxygen (DO) was less than 4 mg/L in one or more test concentrations.
Aeration was not provided in test when DO was <4 mg/L.
pH was <6 or >9 in one or more test concentrations.
Salinity of one or more test concentrations was outside of the range required in the study.
Salinity was adjusted during the test to compensate for evaporation due to test aeration.
Total ammonia was >5 mg/L in one or more test concentrations.
Test Conditions
c.
e2
e3
64
e5
e6
e7
e8
e,
SID
c.
Number of organisms per test container differed from the required number of organisms due to accidental loss of one
or more test organisms.
Test chamber size was different from that required in the study.
Test solution volume in test containers was outside of the range required in the study.
Test renewals were conducted more than 2 hours outside of the required time for test renewal.
Feeding schedule used during the test differed from feeding schedule recommended in the method manuals.
Test termination was conducted more than 2 hours prior to proper test termination time.
Dilution series used for testing was 12.5, 25, 50, 75, and 100% instead of the standard dilution series required in the
study (6.25, 12.5, 25, 50, and 100%).
Lighting cycle was interrupted for 2 hours during power outage.
Continuous shaking rate of 100 cpm was not used.
Initial cell density average was more than 10% outside of required 10,000 ccll/mL inoculation level.
Initial cell density variability among replicates was greater than a CV of 10%.
                                                86

-------
 Table 8.3. Test data qualifier flags (continued)
Flag code
Flag Description
Organisms
f,
f2
f3
f4
f,
Age of organism was outside of range required in the study.
Organism culture contaminated with rotifers.
Organism culture crashed just prior to testing.
Males were identified in the test.
Organisms were cultured at 20°C and directly transferred to test temperature of 25°C.
Quality control
gi
g2
g3
g»
g5
g«
Si
&
e»
gio
gll
gl2
Reference toxicant test conducted during the month of study testing was outside of the 2 standard deviation control
chart limits or the test was not conducted.
Minimum significant difference (MSD) was greater than the recommended criteria for this method (USEPA, 2000b).
ASTM h statistic for this test was greater than the recommended criteria,
ASTM k statistic for this test was greater than the recommended criteria,
indicating that this test may be an outlier.
indicating that this test may be an outlier.
Interrupted concentration - response relationship was observed (i.e., a test concentration was determined to be
significantly different from the control, while one or more higher test concentrations were not significantly different from
the control).
One or multiple test replicates were lost due to laboratory error.
Two cell count methods were used. Test failed test acceptability criteria
passed test acceptability criteria for growth using Hemacytometer method.
for growth using coulter counter method, but
Test was repeated due to laboratory error. Initial test was incorrectly terminated at 48 hours.
Test was repeated due to laboratory error. Initial test was incorrectly renewed with the wrong sample.
Two sets of controls were conducted for this test, and one did not meet test acceptability criteria for reproduction.
Cell density was measured using turbidity technique instead of cell counts.
Referee laboratory test repeated. Initial test failed test acceptability criteria for survival.
8.2.3   Effect Concentration Re-calculation and Verification

To confirm that test results were calculated correctly and according to WET method manual requirements for
statistical data analysis, all test results were re-calculated by SCC using reviewed raw data from the electronic
benchsheets.  SCC conducted statistical analysis of WET test data using ToxCalc version 5.0 (Tidepool
Scientific, 1996). Biological test data was electronically copied from electronic benchsheets directly into the
ToxCalc software to avoid additional data transcription or data entry errors. Statistical methods for analysis
were selected according to the WET method manuals using the EPA flowchart option in the ToxCalc software.
Test results for the endpoints listed in Table 2.3 were calculated for the respective test methods. Any error
messages that were produced by the software were noted and evaluated to ensure that the software defaulted to
the correct alternate statistical method.
                                                   87

-------
As part of the effect concentration re-calculation and verification process, SCC reviewed the concentration-
response curve generated for each sample and endpoint.  SCC reported test results for each sample in accordance
with EPA's guidance on the evaluation of concentration-response relationships (USEPA, 2000a).  When SCC
observed unexpected concentration-response patterns, EPA's guidance (USEPA, 2000a) was followed for
determining whether the derived effect concentration was reliable and should be reported, the effect concentration
was anomalous and should be explained, or the test was inconclusive and the sample should be retested.  When
EPA's concentration-response relationship guidance recommended retesting of a sample, SCC reported the result
for this sample as inconclusive.  Table 8.4 lists the samples that produced unexpected concentration-response
curves and were affected by the concentration-response guidance.

After test results were re-calculated by SCC, these results were compared to results as reported by the participant
laboratory.  If re-calculated and laboratory-reported results differed, the  summarized test data and statistical
analyses used for calculation were reviewed to isolate the source of the deviation. Following re-calculation and
review, test results were incorporated into a results database for the WET Study.
Table 8.4.  Sample results affected by EPA guidance on concentration-response relationships (USEPA,
2000a).
Method
Ceriodaphnia
chronic
Sample
code
9328
9332
9333
9341
9343
9379
9380
9392
Concentration
-response
pattern
observed"
4
4
5
5
6
5
6
6
Effect on reported test result
Reproduction NOEC and IC25 were reported as inconclusive since test
MSD was greater than recommended criterion
Calculated IC25 was determined to be anomalous due to ICp smoothing;
IC25 was adjusted to >100% since mean response in the 100% treatment
was within 25% of control mean
6.25% treatment was determined to be anomalous and reproduction
NOEC was reported as highest concentration not significantly different
from control
6.25% and 12.5% treatments were determined to be anomalous and
reproduction NOEC was reported as highest concentration not
significantly different from control
Reproduction NOEC was reported as concentration below the LOEC
6.25% treatment was determined to be anomalous and survival NOEC
was reported as highest concentration not significantly different from
control
Survival NOEC was reported as inconclusive since test MSD was greater
than recommended criterion
Reproduction NOEC was reported as concentration below the LOEC

-------
Table 8.4. Sample results affected by EPA guidance on concentration-response relationships (USEPA,
2000a). (continued)
Method
Ceriodaphnia
chronic
(continued)
Fathead
chronic
Selenastrum
chronic
Sample
code
9408
9415
9122
9129
9145
9161
9168
9193
9194
9209
9212
9454
(w/o EDTA)
9455
(w/ EDTA)
9455
(w/o EDTA
9468
(w/ EDTA)
9468
(w/o EDTA)
9473
(w/o EDTA)
Concentration
-response
pattern
observed"
4
5
5
6
5
6
5
6
6
5
6
4
4
4
5
5
5
Effect on reported test result
Calculated IC25 was determined to be anomalous due to ICp smoothing;
1C25 was adjusted to >100% since mean response in the 100% treatment
was within 25% of control mean
Survival NOEC was reported as inconclusive since test MSD was greater
than recommended criterion
6.25% treatment was determined to be anomalous and growth NOEC was
reported as highest concentration not significantly different from control
Growth NOEC was reported as concentration below the LOEC
25% treatment was determined to be anomalous and survival NOEC was
reported as highest concentration not significantly different from control
Growth NOEC was reported as concentration below the LOEC
6.25% treatment was determined to be anomalous and growth NOEC was
reported as highest concentration not significantly different from control
Growth NOEC was reported as concentration below the LOEC
Growth NOEC was reported as concentration below the LOEC
Growth NOEC was reported as inconclusive since test MSD was greater
than recommended criterion
Survival and growth NOECs were reported as concentration below the
LOEC
Growth NOEC, IC25, and IC50 were reported as inconclusive since test
MSD was greater than recommended criterion
Growth NOEC, IC25, and IC50 were reported as inconclusive since
control response was marginal and below laboratory's normal range of
control performance
Calculated IC25 was determined to be anomalous due to ICp smoothing;
1C25 was adjusted to >100% since mean response in the 100% treatment
was within 25% of control mean
50% treatment was determined to be anomalous and growth NOEC was
reported as highest concentration not significantly different from control
Growth NOEC was reported as inconclusive since test MSD was greater
than recommended criterion
Growth NOEC was reported as inconclusive since test MSD was greater
than recommended criterion
                                               89

-------
Table 8.4. Sample results affected by EPA guidance on concentration-response relationships (USEPA,
2000a). (continued)
Method
Mysid chronic
Silverside
chronic
Sample
code
9682
9694
9696
9545
9556
Concentration
-response
pattern
observed"
5
5
5
4
5
Effect on reported test result
12.5% treatment was determined to be anomalous and fecundity NOEC
was reported as highest concentration not significantly different from
control
12.5% treatment was determined to be anomalous and fecundity NOEC
was reported as highest concentration not significantly different from
control
12.5% treatment was determined to be anomalous and growth NOEC was
reported as highest concentration not significantly different from control
Growth NOEC and IC25 were reported as inconclusive since test MSD
was greater than recommended criterion
12.5% treatment was determined to be anomalous and growth NOEC was
reported as highest concentration not sienificantlv different from control
1 Concentration-response patterns are numbered as identified in Chapter 4 of USEPA, 2000a.
                                                  90

-------
9.0 RESULTS

9.1     Analysis of Results

SCC personnel entered re-calculated and verified test results for each sample tested into a results database along
with associated data qualifier flags, sample information, and summary test information (e.g., control mean,
control CV, test minimum significant difference, etc.). Information in the results database was used to evaluate
the test completion rate, false positive rate, and precision for each test method.  All calculated test results
presented in this section were rounded to three significant figures for consistency. A formal analysis of
measurement error for each data type was not conducted, however, three significant figures is believed to be
consistent with most WET test measurements (e.g., test concentrations,  weights, counts). Summary statistics
presented in this section (e.g., mean, standard deviation, CV) were calculated from test results prior to rounding
(so summary statistics of rounded results may differ slightly).

9.1.1    Valid Tests

Only valid tests were  used in the determination of false positive rates and precision. A valid test was defined as a
test that met the required test acceptability criteria for the method as stated in the WET method manuals. When
EPA guidance on the  evaluation of concentration-response relationships (USEPA, 2000a) recommended retesting
of a sample, the test result in question was identified as inconclusive and was not included in the analysis of false
positive rates and precision for the methods.  Also, test results from referee laboratories were excluded from
determinations of test completion rates, false positive rates, and precision for the methods.  While referee
laboratory testing was conducted similarly to and simultaneously with participant laboratory testing, the identity
of samples was not blinded to the referee laboratory.

9.1.2   Successful Test Completion Rate

The test completion rate was calculated independently for each test method as the percentage of initiated and
properly terminated tests that met the test acceptability criteria as specified in the WET method manuals.
Participant laboratories that failed to complete tests due to reasons unrelated to  the test methods themselves (i.e.,
laboratory error) were not included in the test completion rate calculations or statistical analyses.  This occurred
for only four samples (9586, 9587, 9589, and 9618).  In each case, the referee laboratory sent a new sample
aliquot from the original bulk sample preparation for retesting at the participant laboratory. Results from the
repeated tests are presented in this report and were used in the determination of test completion rates, false
positive rates, and precision.

9.1.3    False Positive Rate

The false positive rate was calculated independently for each test method and for each endpoint reported (LC50,
survival NOEC, IC25 for growth, IC25 for reproduction, NOEC for growth, and NOEC for reproduction). The

                                                   91

-------
false positive rate was determined as the rate at which test results indicated toxicity in blank samples, and was
calculated as:
                Number of valid tests indicating toxicity in blank samples
                 Total number of valid tests conducted on blank samples
9.1.4   Precision

Precision estimates were generated independently for each test method, point estimate, and sample type tested.
For sample types that were tested using within-laboratory replication, estimates were provided for within-
laboratory precision (based on the within-laboratory variance component), between-laboratory precision (based
on the between-laboratory variance component), and total precision (based on the total variance).  For sample
types that were not tested using within-laboratory replication, a single precision estimate was generated based on
the total variance.

When test results were calculated as outside of the test concentration range (i.e.,  >100% or <6.25%), these data
points were censored for the purposes of estimating precision. Results that were >100% were censored as 100%,
and results that were <6.25% were censored as 6.25%. IC25 results of >12.5%, >25%, and >50% were  also
possible for the mysid chronic fecundity endpoint. These values were censored as 12.5%, 25%, and 50%,
respectively.  Since this censoring of data could result in a bias of the precision estimates, precision estimates
were not calculated for data sets that required greater than 20% of data to be censored in this manner. This
occurred in testing of the blank sample type for all test methods, the reference toxicant sample type for the
Ceriodaphnia chronic and silverside acute test methods (see Section 5.3), and all sample types for the mysid
chronic fecundity endpoint.

Only participant laboratory results from valid tests (see Section 9.1.1) were included in the calculation of
precision estimates; invalid tests were excluded. SCC conducted an outlier analysis of valid test results to
determine if any additional test results should be excluded from the analysis  of precision.  SCC used the
calculation of ASTM's h and k statistics (ASTM, 1997) to evaluate data consistency and identify potential
outliers.  ASTM h statistics were used to examine the consistency of test results  from laboratory to laboratory.
ASTM k statistics were used to examine the consistency of within-laboratory precision from laboratory to
laboratory. For each test method,  ASTM h statistics were calculated for each laboratory and each sample type
(except the blank sample type) using the equation below.
                       ,
                       h =
                                                   92

-------
where,          JC = a laboratory's average test result for a given sample type (if the laboratory only tested one
                sample of a given sample type, that individual result was used)

                X = the average of individual laboratory averages (2_,x/p)
                                                                i
               p = number of laboratories testing a given sample type

For each test method, ASTM k statistics were calculated for each laboratory and each sample type (except the
blank sample type) using the equation below. The k statistic was not calculable for laboratories that did not test
replicate samples of the same sample type.
/t - 1 	
p
M
V i /
/
IP

J                                     I
                      2^(x- x)   (n- l)
               p = number of laboratories testing a given sample type
               x = an individual test result

                X = a laboratory's average test result for a given sample type
               n = number of test results for a given sample type from a single laboratory

For each test method and sample type (excluding the blank sample type), SCC compared the h and k statistics
calculated for each laboratory to critical values of h and k statistics at the recommended 0.5% significance level
(see ASTM, 1997 for table of critical values).  Test results from laboratories with a calculated h statistic above
the critical value were significantly different (at the 0.5% significance level) from results reported by other
laboratories for the same sample type. These inconsistent test results were flagged (see Table 8.3) and identified
as potential outliers.  Laboratories with a calculated k statistic above the critical value experienced greater within-
laboratory variability  than other laboratories testing the same sample type.  These inconsistent test results also
were flagged (see Table 8.3) for further investigation.

Since estimates of coefficients of variation can be biased by extreme values and by small data sets, it was
important to closely investigate individual data points before discarding them as outliers. An individual test result
was only discarded as an outlier if the laboratory was identified by ASTM h statistics as an outlier and a
reasonable cause for producing the aberrant result could be determined. The data qualifier flags associated with
each test result were useful in this determination. In general, a very conservative approach to excluding outliers
was taken. Only 15 tests in the entire study of 698 tests were identified by ASTM h statistics as potential
outliers.  Two of these tests were  also flagged for extreme ASTM k statistics.  Of the 15 tests identified as
                                                   93

-------
potential outliers, only 8 were excluded from precision estimates based on the determination of a cause for
inconsistent results. Table 9.1 shows the test results that were identified as potential outliers and provides a
rationale for the inclusion or exclusion of these data points from precision estimates.

Precision was estimated by the coefficient of variation (CV) for point estimates.  For NOEC values, precision
was simply described by the range and distribution of NOEC values and the percentage of values falling within
one concentration of the median (as described in the WET method manuals for evaluating routine reference
toxicant test results using NOECs).  The CV for point estimates was calculated as:
                                X.

where,          S2 = variance (S = the standard deviation)

                x = mean of valid test results for a given method, endpoint, and sample type

For test methods and sample types that included within-laboratory replication (i.e., multiple tests on the same
sample type from a given laboratory), the variance identified in the above equation was obtained by maximum
likelihood estimation using the PROC MIXED procedure in SAS version 8 (SAS Institute, 2000).  This
procedure estimated the within-laboratory, between-laboratory, and total variance components.  Each of these
variance components were individually used to calculate within-laboratory, between-laboratory, and total CVs.

For test methods and sample types that did not include within-laboratory replication, the variance identified in the
above equation was obtained by the following equation using the PROC MEANS procedure in SAS version 8
(SAS Institute, 2000).
where,         n = number of valid test results for a given method, endpoint, and sample type
               Xj = individual result / (z ranging from 1 to n)
               S2 = variance of the n test results (S = the standard deviation)

                x = mean of the n test results
                                                   94

-------
Table 9.1.  Test results identified as potential outliers by ASTM h statistics.
Test method
Ceriodaphnia
chronic
Fathead acute
Fathead
chronic
Selenastrum
chronic
Sheepshead
acute
Silverside
chronic
Lab
ID
42
42
62
62
205
205
205
125
125
39
125
425
101
421
421
Sample
code
9347
9348
9033
9034
9065
9066
9177
9162
9163
9468
(w/
EDTA)
9476
(w/o
EDTA)
9617
9600
9582
9583
Test endpoint
survival and
reproduction
survival and
reproduction
survival
survival
survival
survival
growth
growth
growth
growth
growth
survival
survival
survival and
growth
survival and
growth
Included/e
xcluded
excluded
excluded
included
included
included
included
included
excluded
excluded
excluded
excluded
included
included
excluded
excluded
Rationale
These duplicate samples were flagged for both h
and k statistics, indicating that the laboratory's
mean test result was significantly different from
other laboratories and that within-lab variability
was also significantly higher for this laboratory;
this high within-lab variability could explain the
inconsistent mean test result from this laboratory
No data qualifier flags
Only data qualifier flag was for test chamber size
No data qualifier flags
Dilution water quality was highly variable for this
laboratory. For these two tests, dilution water
hardness at test initiation was 111 mg/L,
compared to the expected hardness range of 80-
100 mg/L for moderately hard reconstituted
water. For the remaining two fathead chronic
tests conducted in this laboratory, dilution water
alkalinity at test initiation was 45 mg/L,
compared to the expected alkalinity range of 60-
70 mg/L.
Cell growth was marginal and failed test
acceptability criteria when measured using
coulter counter; other flags included sample
temperature, test temperature, test pH, continuous
shaking rate, and interrupted concentration-
response relationship
Reference toxicant test was not conducted
concurrently with tests or within the month of
testing
Data qualifier flags were observed for salinity of
30 ppt on test day 4 and for test termination at
93.7 hours (rather than 96), however, the data
point was included due to the small size of the
data set for this method (this result represented
14% of the data set)
No data qualifier flags
Reference toxicant test conducted concurrently
with samples was outside of control chart limits
                                                   95

-------
9.2     Ceriodaphnia Acute Test Method Results

A total of 28 participant laboratories conducted the Ceriodaphnia acute test method in the WET Study. These
laboratories tested a total of 34 blank samples, 30 reference toxicant samples, 27 effluent samples, and 13
receiving water samples. For each sample tested, a 48-hour LC50 was generated as a test result.  Results of
Ceriodaphnia acute testing are shown in Tables 9.2 - 9.5 for each sample type.

The sample distribution scheme used for the Ceriodaphnia acute test method was inadvertently altered from the
original study design (see Section 6.4) due to an error in sample distribution.  On the first week of testing, all
laboratories that were intended to receive blank samples were shipped reference toxicant samples, and all
laboratories that were intended to receive reference toxicant samples were shipped blank samples.  This error was
not identified prior to the second week of testing, so shipments during the second week were conducted as
planned. This caused some laboratories to receive three blank samples and no reference toxicant samples and
other laboratories to receive three reference toxicant samples and no blank samples.  This sample distribution
error should not affect the evaluation of the Ceriodaphnia acute test method.

9.2.1    Successful Test Completion Rate

A total of 104 Ceriodaphnia acute tests were initiated by 28 participant laboratories. All 104 tests were
completed, however, tests conducted on five samples (9232, 9222, 9233, 9234, and 9231) were invalid due to
failure to meet test acceptability criteria for survival. The resulting successful test completion rate calculated in
the WET Study for the Ceriodaphnia acute test  method was 95.2%. Four of the five invalid tests were conducted
in a single laboratory (Lab 29).  This laboratory failed to properly culture test organisms at the test temperature
of 25°C. The laboratory transferred organisms cultured at 20°C to the test temperature of 25°C, causing
significant mortality in all test treatments and failure of test acceptability criteria.

9.2.2    False Positive Rate

A total of 33 valid tests were completed on blank samples by 14 participant laboratories (Table 9.2).  The LC50
calculated for all 33 blank samples was >100%,  indicating no toxicity and no false positives. The resulting false
positive rate calculated in the WET Study for the Ceriodaphnia acute test method was 0.00%.

9.2.3    Precision

Precision of the Ceriodaphnia acute test method was estimated by calculating the CV of LC50 values obtained
for the reference toxicant, effluent, and receiving water samples. Within-laboratory, between-laboratory, and
total CVs were calculated for the reference toxicant and effluent samples. Only a total CV was calculated for the
receiving water samples since no within-laboratory replication was provided for this sample type. All valid
participant laboratory test data for the reference  toxicant, effluent, and receiving water samples were used in
estimating precision.  No test results were identified by ASTM h statistics as possible outliers.

                                                   96

-------
Table 9.6 summarizes the precision of point estimates from the Ceriodaphnia acute test method. Within-
laboratory CVs ranged from 9.68% to 14.6%, between-laboratory CVs ranged from 15.2% to 32.8%, and total
CVs ranged from 21.1% to 34.2%.  Total CVs were lower for reference toxicant samples (21.1%) than for
effluent (34.2%) or receiving water (31.8%) samples. As expected, the majority of variability was due to the
between-laboratory component, with within-laboratory CVs averaging 12.1% and between-laboratory CVs
averaging 24.0%.  Averaging the CVs based on total variance for the three sample types, a total CV of 29.0%
was obtained for the  Ceriodaphnia  acute test method in the WET Study.
                                                97

-------
Table 9.2. Results for Ceriodaphnia acute test method performed on blank samples.

LabID
29
6
18
18
18
33
33
33
46
46
46
62
62
69
69
70
70
70
73
73
73
105
157
157
157
251
251
311
311
417
417
452
452
452
Summary
Statistics

Sample code
9232
9221
9224
9225
9226
9235
9236
9237
9243
9244
9245
9250
9251
9254
9255
9258
9259
9260
9262
9263
9264
9270
9283
9284
9285
9295
9296
9299
9300
9315
9316
9322
9323
9324
N
Min
Max
Median
Mean

Test date
11/09/99
11/09/99
1 1/09/99
11/11/99
11/11/99
11/09/99
1 1/09/99
11/11/99
1 1/09/99
1 1/09/99
11/11/99
1 1/09/99
11/09/99
1 1/09/99
11/09/99
11/09/99
11/09/99
11/11/99
1 1/09/99
11/09/99
11/11/99
1 1/09/99
11/09/99
11/09/99
11/11/99
11/09/99
1 1/09/99
1 1/09/99
11/09/99
11/09/99
11/11/99
11/09/99
11/11/99
11/11/99





False positives
False Dositive rate
Survival Information
LC50
(% sample)
Invalid11
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
33
>100
>100
>100
>100
0
0.00%
Control mean
(%)
5.00
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
90.0
90.0
100
100
100
100
100
100
100
100
100
100
100
100
100
100
95.0
95.0







Control CV (%)
41.8
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
11.2
11.2
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.26
9.26








Flags"
a,, e2, f5

d,,e2
d,,e2
d,,e2



c4, d4
c4, d4
C4




C2
C2
C2



Ss


b,
















a Data qualifier flags are described in Table 8.3.
 Results from invalid tests were excluded from summary statistics.
                                                     98

-------
Table 9.3. Results for Ceriodaphnia acute test method performed on reference toxicant samples.

LabID
Referee
Referee
3
3
3
25
25
25
42
42
42
60
101
101
101
113
113
113
125
125
141
141
141
205
238
406
416
416
416
425
425
425
Summary
Statistics

Sample code
9307
9310
9217
9218
9219
9227
9229
9230
9239
9241
9242
9246
9265
9267
9268
9273
9274
9275
9276
9279
9280
9281
9282
9287
9291
9303
9311
9312
9313
9318
9320
9321
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
1/09/99
1/11/99
1/09/99
1/09/99
1/11/99
1/09/99
1/11/99
1/11/99
1/09/99
1/11/99
1/11/99
1 1/09/99
1 1/09/99
11/11/99
11/11/99
1 1/09/99
11/09/99
11/11/99
1 1/09/99
11/11/99
1 1/09/99
11/11/99
11/11/99
11/09/99
11/09/99
11/09/99
11/09/99
11/11/99
11/11/99
11/09/99
11/11/99
11/11/99





STD
CV%
STD
cv%
STD
CV%
Survival Information
LC50
(% sample)
40.6"
34.4b
31.9
27.6
18.3
34.4
33.0
33.0
37.9
35.4
50.0
17.1
21.7
25.9
27.7
35.4
32.0
35.4
26.2
24.8
31.2
32.7
35.4
39.2
30.8
31.8
34.2
24.1
26.8
35.4
30.0
28.9
30
17.1
50.0
31.8
30.9
4.52
14.6%
4.70
15.2%
6.52
21.1%
Control mean
(%)
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
90.0
100
100
100
100
100
100
100
100
100
100
100
100
100
100











Control CV (%)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
18.7
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00












Flags"


b4

6,
e,
63
63

b,
b,
C|, 63
e2, e3
e2, e,
e2, e,





C2
C2
C2
chc2, d.
e2, e6
c.
C2, Cj
C2, C,
C2, C,
C|
C|
Ci,gs











1 Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                     99

-------
Table 9.4.  Results for Ceriodaphnia acute test method performed on effluent samples.

LabID
Referee
6
29
29
6
33
60
60
62
62
69
69
70
105
105
125
157
205
205
238
238
251
251
311
311
406
406
417
Summary
Statistics

Sample code
9309
9222
9233
9234
9223
9238
9248
9249
9252
9253
9256
9257
9261
9271
9272
9278
9286
9289
9290
9293
9294
9297
9298
9301
9302
9305
9306
9317
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99
11/11/99





STD
cv%
STD
cv%
STD
cv%
Survival Information
LC50
(% sample)
24.6"
Invalid0
Invalid0
Invalid0
9.99
13.0
13.2
10.3
28.2
26.8
27.7
33.0
21.2
23.6
23.1
20.3
35.4
36.6
30.8
16.9
16.5
17.1
19.5
24.1
21.8
27.7
27.0
15.0
24
9.99
36.6
22.4
22.5
2.17
9.68%
7.36
32.8%
7.68
34.2%
Control mean
(%)
100
40.0
0.00
0.00
100
100
95.0
95.0
90.0
100
100
100
90.0
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100











Control CV (%)
0.00
59.5
0.00
0.00
0.00
0.00
9.26
9.26
18.7
0.00
0.00
0.00
11.2
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00












Flags"

a.
a,, dh e2, fs
ab d,, e2, fs


C|,e,
Ci,e3




C2



b,
b,,b4, c,,c2)d|
b|,b4)c,,c2) d,
e2
e2




bi, c,,c2
C|,C2












1 Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
0 Results from invalid tests were excluded from summary statistics.
                                                       100

-------
Table 9.5.  Results for Ceriodaphnia acute test method performed on receiving water samples.

LabID
Referee
29
6
25
42
60
101
105
125
205
238
406
417
425
Summary
Statistics

Sample code
9308
9231
9220
9228
9240
9247
9266
9269
9277
9288
9292
9304
9314
9319
N
Min
Max
Median
Mean
STD
CV%

Test date
1 1/9/99
11/9/99
1 1/9/99
11/9/99
1 1/9/99
1 1/9/99
1 1/9/99
1 1/9/99
1 1/9/99
11/9/99
1 1/9/99
11/9/99
1 1/9/99
1 1/9/99







Survival Information
LC50
(% sample)
30.8b
Invalid0
20.8
34.2
26.8
16.5
21.5
11.1
33.2
34.2
23.3
19.9
20.4
17.7
12
11.1
34.2
21.2
23.3
7.40
31.8%
Control mean
(%)
100
5.00
100
100
100
100
100
100
90.0
100
100
100
100
100







Control CV
(%)
0.00
41.8
0.00
0.00
0.00
0.00
0.00
0.00
11.2
0.00
0.00
0.00
0.00
0.00








Flags"
d,
a,, e2, f5

£3

Ci, 63
e2, e3


bb b4, ct, c2, d,
e2, e6
Cl

c.







* Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
Table 9.6.  Precision of point estimates from the Ceriodaphnia acute test method.
Sample type

Reference toxicant
Effluent
Receiving water
Average
CV (%)
Within-lab"
14.6
9.68
-
12.1
Between-lab"
15.2
32.8
-
24.0
Total
21.1
34.2
31.8
29.0
a Within and between-laboratory components of variability were not calculated for the receiving water sample type since no within-
replication was provided for this sample type.
                                                       101

-------
9.3     Ceriodaphnia Chronic Test Method Results

A total of 35 participant laboratories conducted the Ceriodaphnia chronic test method in the WET Study.  One of
the participant laboratories did not submit a data report (see Section 8.1), so summarized results are based on 34
participant laboratories.  These laboratories tested a total of 34 blank samples, 48 reference toxicant samples, 27
effluent samples, and 13 receiving water samples. For each sample tested, a survival NOEC, a reproduction
NOEC, a survival LC50, and a reproduction IC25 were generated as test results. While all Ceriodaphnia
chronic tests performed in the WET Study were conducted for eight days, results presented in this section were
determined using the current WET method manual criterion for test termination. This criterion states that tests
should be terminated when 60% of the surviving control females have produced their third brood, or at the end of
8 days, whichever occurs first. As described in the WET method manual, reproduction was measured as the total
number of young produced per original female at the time of appropriate test termination.  Organisms positively
identified as males were excluded from the reproduction analysis. Also, test concentrations above the survival
NOEC were excluded from hypothesis testing conducted on the reproduction endpoint.  Results of Ceriodaphnia
chronic testing are shown in  Tables 9.7 - 9.10 for each sample type.

Precision estimates were not obtained from the reference toxicant sample type since greater than 20% of test
results were outside of the test concentration range. For the reference toxicant sample, 97.3% of LC50s were
>100% sample; 72.2% of IC25s were >100% sample. This was caused by a reference toxicant sample that was
only moderately toxic. The spiking concentration of KC1 for this sample was selected to achieve an IC25 of
approximately 50% sample based on preliminary testing. Despite preliminary testing efforts, the spiking level
selected was insufficient to produce this targeted level of effect (see Section 5.3).

9.3.1    Successful Test Completion Rate

A total of 122 Ceriodaphnia chronic tests were initiated by 34 participant laboratories. All 122 tests were
completed, however, tests conducted on 22 samples were invalid due to failure to meet test acceptability criteria
for survival or reproduction.  The resulting successful test completion rate calculated in the WET Study for the
Ceriodaphnia chronic test method was 82.0%.  In addition, the reproduction results for sample 9328 and the
survival NOEC results for samples 9415 and 9380 were reported as inconclusive based on an evaluation of the
concentration-response relationship (Table 8.4). If these tests are considered unsuccessful in addition to invalid
tests (since the test would be repeated in a regulatory context), the successful test completion rate becomes
79.5%.

Of the 34 participant laboratories, 24 produced valid results for all samples tested.  The 22 invalid tests  were
concentrated in the remaining 10 laboratories.  Of these 10 laboratories, 8 laboratories performed invalid tests on
50% or more of the samples tested. Two laboratories performed invalid tests on all samples tested. This
attributed to the relatively low successful test completion rate achieved for the Ceriodaphnia chronic test method
in the WET Study.
                                                  102

-------
9.3.2    False Positive Rate

A total of 27 valid tests were conducted on blank samples by 22 participant laboratories (Table 9.7). No false
positives were observed for the survival endpoints.  The survival NOEC was  100% for all 27 blank samples, and
the LC50 was >100% for all 27 blank samples. One false positive was observed for sublethal endpoints. The
reproduction NOEC for sample 9450 was 25%, and the reproduction IC25 for this sample was 15.883%.  The
resulting false positive rate calculated in the WET Study for the Ceriodaphnia chronic test method was 3.70%
for the reproduction endpoint. The one false positive that was observed for this method originated from a
laboratory that failed all other Ceriodaphnia chronic tests conducted.

In addition to the false positive reported above, a participant laboratory reported a reproduction IC25 of less than
100% for sample 9332, indicating a false positive result.  Based on EPA guidance for evaluating concentration-
response relationships (USEPA, 2000a), this value was determined to be an anomalous result of the ICp
smoothing procedure and the IC25 was corrected to >100% (Table 8.4). A participant laboratory also reported a
survival NOEC and reproduction NOEC of less than 100% for sample 9379,  indicating a false positive result.
This sample exhibited an interrupted concentration-response curve, and based on EPA guidance for evaluating
concentration-response relationships, the survival NOEC and growth NOEC were re-calculated and reported as
100% (Table 8.4).  Sample 9341 also produced an interrupted concentration-response curve, but the reproduction
NOEC was similarly re-calculated and reported as 100% (Table 8.4).

9.3.3    Precision

Precision of the Ceriodaphnia chronic test method was estimated by calculating the CV of LC50 and IC25 values
obtained for the effluent and receiving water samples. CVs were not calculated for the reference toxicant sample
type because greater than 20% of the data set contained values outside of the  concentration range. For the
effluent sample type, within-laboratory, between-laboratory, and total CVs were calculated. Only a total CV was
calculated for the receiving water samples since no within-laboratory replication was provided for this sample
type.  Survival and reproduction results from samples 9347 and 9348 were excluded from the analysis of
precision. These test samples were identified by ASTM h statistics as possible outliers, and review of data
qualifier flags revealed a possible cause (Table 9.1).

Table 9.11 summarizes the precision of point estimates from the Ceriodaphnia chronic test method.  Within-
laboratory and between-laboratory CVs for LC50 values were 7.09% and 21.8%, respectively. Total CVs ranged
from 20.0% for the receiving water sample type to 23.0% for the effluent sample type. Averaging the CVs based
on total variance for these two sample types, a total CV of 21.5% was obtained for the Ceriodaphnia chronic
survival endpoint.

CVs for IC25 values were higher than those for the survival endpoint (LC50). Within-laboratory CVs were
18.5%, between-laboratory CVs were 27.3%, and total CVs ranged from 32.9% to 37.3%. Averaging the CVs
for the IC25 based on total variance for the two sample types, a total CV of 35.1% was obtained for the
Ceriodaphnia chronic reproduction endpoint.

                                                 103

-------
The precision of NOEC values was determined by evaluating the range and distribution of NOEC values and the
percentage of values falling within and beyond one concentration from the median.  Table 9.12 describes the
precision of NOEC values for the Ceriodaphnia chronic method. For the survival endpoint, NOEC values
spanned two concentrations for the receiving water sample types and three concentrations for the reference
toxicant and effluent sample types. The percentage of values within one concentration of the median was 97.2%,
91.3%, and 100% for the reference toxicant, effluent, and receiving water sample types, respectively.

Reproduction endpoints were less precise for the reference toxicant sample type, spanning five concentrations, but
were comparable for the effluent and receiving water sample types, spanning three and two concentrations,
respectively. The percentage of values within one concentration of the median was 83.3%, 100%, and 100% for
the reference toxicant, effluent, and receiving water sample types, respectively. Of the eight results (for samples
9342, 9343, 9360, 9392, 9397, 9415, 9361, and 9362) that were beyond one concentration from the median,
three (samples 9343, 9392, and 9415) were the result of interrupted concentration-response curves. Three results
that were beyond one concentration from the median were from very statistically sensitive tests, with %MSDs
below EPA's recommended lower bound of 11% (USEPA, 2000b). Percent MSDs for these three samples (9360,
9361, and 9362) were 8.7%, 8.1%, and 9.2% respectively.
                                                 104

-------
Table 9.7. Results for Ceriodaphnia chronic test method performed on blank samples.

LabID
Referee
46
69
311
311
333
406
406
6
25
27
27
30
30
33
44
44
49
71
73
101
105
113

Sample
code
9425
9352
9363
9411
9412
9420
9422
9423
9330
9332
9337
9338
9340
9341
9344
9349
9350
9356
9367
9371
9376
9379
9381

Test date
10/12/99
10/26/99
10/12/99
10/26/99
10/26/99
10/26/99
10/12/99
10/12/99
10/26/99
10/12/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/12/99
10/12/99
10/12/99
10/13/99
10/12/99
10/26/99
10/26/99
10/12/99
Survival Information
NOEC
(% sample)
100b
Invalid0
Invalid'
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
LC50
(% sample)
>100b
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid'
Invalid0
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
Control
mean (%)
87.5
70.0
100
90.0
100
70.0
70.0
77.8
100
80.0
100
100
80.0
100
100
100
100
90.0
100
100
100
100
100
Reproduction Information
NOEC
(% sample)
100b
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
IC25
(% sample)
>100b
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
Control
mean
(neonates)
17.1
2.40
9.10
9.50
8.40
26.1
21.1
21.6
25.4
18.8
20.1
24.2
16.0
23.5
15.8
30.8
29.5
24.1
23.1
21.5
20.4
24.9
26.5
Control CV
(%)
48.6
79.1
53.7
43.9
56.7
69.7
50.4
54.9
26.0
63.8
27.8
12.3
58.6
22.1
51.7
6.46
4.86
18.7
15.1
20.5
17.2
27.4
12.4
Day test
termination
7
8
8
8
8
7
6
6
6
7
6
7
7
6
6
7
7
6
7
6
7
6
7

Flags"
di, g2, g6
ai> a3, a4, C4:
db d4, e5, f,,
g2
a4> C2> 66> I3>
Rt
a,, a4, c2, g2
a,, a4, c2, g2
a,, e2, e3, e6,
g2, g5
a,, c,, c2, g2
a,, c,, c2, g2
f4
e6

b,
g2
g5
§2
e2
e2
C2



C2. e6: §5
c,, efi
                                                               105

-------
Table 9.7. Results for Ceriodaphnia chronic test method performed on blank samples, (continued)

LabID
113
125
231
299
299
416
417
421
425
448
450
452
Summary
Statistics

Sample
code
9382
9384
9402
9409
9410
9429
9432
9436
9439
9445
9446
9450
N
Min
Max
Median
Mean

Test date
10/12/99
10/12/99
10/27/99
10/27/99
10/27/99
10/12/99
10/12/99
10/12/99
10/12/99
10/26/99
10/12/99
10/12/99





False positives
False positive rate
Survival Information
NOEC
(% sample)
100
100
100
100
100
100
100
100
100
100
100
100
27
100
100
100

0
0 00%
LC50
(% sample)
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
27
>100
>100
>100
>100
0
0 00%
Control
mean (%)
100
90.0
100
90.0
100
100
90.0
100
100
100
100
90.0







Reproduction Information
NOEC
(% sample)
100
100
100
100
100
100
100
100
100
100
100
25
27
25
100
100

1
3.70%
IC25
(% sample)
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
15.9
27
15.9
>100
>100
96.9
1
3.70%
Control
mean
(neonates)
27.5
17.3
22.7
22.2
24.8
31.6
18.8
30.0
18.9
23.6
22.2
19.4







Control CV
(%)
9.11
53.1
16.6
12.2
15.7
15.6
28.7
32.2
33.8
4.09
47.9
73.9







Day test
termination
7
7
8
6
6
7
8
7
8
6
6
7








Flags"
c2, e6
C2, g2
b,
b,,d,
b,,d,
c2, c,
C2
d,,e6
c,

<=2. §2
f>, gl,g?







'" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
                                                                          106

-------
Table 9.8. Results for Ceriodaphnia chronic test method performed on reference toxicant samples.

LabID
Referee
6
46
46
105
105
238
311
333
421
452
452
6
27
30
33
33
42
42
44
49
49
62
62
73
73

Sample code
9428
9329
9354
9355
9377
9378
9405
9413
9419
9438
9452
9453
9328
9336
9339
9342
9343
9345
9346
9351
9357
9358
9359
9360
9372
9373

Test date
10/26/99
10/12/99
10/26/99
10/26/99
10/12/99
10/12/99
10/12/99
10/26/99
10/26/99
10/26/99
10/27/99
10/27/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/26/99
10/26/99
10/26/99
10/12/99
10/12/99
10/26/99
10/26/99
Survival Information
NOEC
(% sample)
100b
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid0
100
100
100
100
100
100
100
100
100
100
100
100
100
100
LC50
(% sample)
>100"
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
Control
mean (%)
100
100
100
100
100
60.0
30.0
100
50.0
77.8
90.0
80.0
100
100
100
100
100
100
100
100
90.0
100
100
100
100
90.0
Reproduction Information
NOEC
(% sample)
100b
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Inconclusive*
100
50
25
12.5
100
100
100
100
100
100
6.25
100
100
IC25
(% sample)
>100b
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Inconclusive11
>100
<6.25
28.1
24.9
>100
>100
>100
>100
>100
>100
>100
>100
>100
Control
mean
(neonates)
25.3
10.8
5.60
5.00
14.8
7.30
9.50
6.30
13.1
28.0
7.80
9.00
15.9
24.7
25.8
19.2
21.1
18.0
17.2
32.8
19.8
21.5
25.6
27.4
22.7
23.0
Control CV
(%)
11.9
59.0
68.5
21.1
46.7
95.1
16.6
65.3
115
53.7
58.2
35.1
48.6
9.36
38.7
39.0
18.1
13.4
12.2
13.2
40.7
14.7
8.48
7.14
12.8
20.1
Day test
termination
6
8
8
8
6
7
8
8
6
7
8
8
6
7
6
6
6
7
7
7
6
6
6
6
6
6

Flags"
d,
84, g2
a,, a4, c4, d,,
d4, e5, f,, g2
a,, a4, c4, d,,
d4, e,, f,, g,
a4, c2, e6, g2
a,, a4, c2, e6,
82, g<
a,, a,, a4, e7
a,, a4, C2, g2
a,, a4, e2, e3,
eh, R,
a,,c,, d],e6
a,, a4> b2, f2,
gb g2
a3, a4, b2, f2,
Ri,R2
§2
b,
g2

&


e.
CT
c,, f4




                                                                107

-------
 Table 9.8.  Results for Ceriodaphnia chronic test method performed on reference toxicant samples,  (continued)

LabID
101
101
113
125
125
134
141
141
205
231
231
299
313
313
406
416
416
417
417
421
448
448
450
Summary
Statistics

Sample code
9374
9375
9383
9386
9387
9389
9392
9393
9397
9400
9401
9408
9414
9415
9424
9430
9431
9434
9435
9437
9443
9444
9449
N
Min
Max
Median
Mean

Test date
10/12/99
10/12/99
10/26/99
10/26/99
10/26/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/20/99
10/20/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/12/99
10/12/99
10/26/99





Survival Information
NOEC
(% sample)
100
100
100
100
100
100
100
100
25
100
100
100
100
Inconclusive11
100
100
100
100
100
100
100
100
100
36
25
100
100

LC50
(% sample)
>100
>100
>100
>100
>100
>100
>100
>100
55.8
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
37
55.8
>100
>100
98.8
Control
mean (%)
100
100
100
100
100
100
100
100
100
100
90.0
100
100
100
100
100
100
90.0
100
100
100
100
100





Reproduction Information
NOEC
(% sample)
100
100
100
100
100
100
12.5
50
6.25
50
50
100
100
12.5
100
100
100
100
100
100
100
100
100
36
6.25
100
100

IC25
(% sample)
>100
>100
>100
>100
>100
>100
93.6
>100
<6.25
81.2
92.0
>100
85.6
7.67
>100
>100
>100
>100
82.7
>100
>100
>100
>100
36
<6.25
>100
>100
86.0
Control
mean
(neonates)
22.8
24.2
32.0
22.6
21.5
25.2
20.7
19.1
16.5
24.6
24.9
24.0
23.3
19.4
26.3
24.7
25.8
17.5
19.6
30.1
21.0
21.4
24.3





Control CV
(%)
7.40
8.22
13.4
16.7
19.0
29.1
13.3
10.6
21.6
8.40
12.6
33.0
58.0
40.9
15.5
26.0
12.9
43.8
22.8
33.1
21.2
8.87
15.2





Day test
termination
6
6
7
6
6
6
7
7
6
6
6
6
7
6
6
6
6
8
8
7
6
6
6






Flags"


e*
C2
C2
f-t, 82
c,, c,, d4, gs
c2, c,
bi,c,,c2


di, e6
g2
g2,g^
Cbf.
C2, C,
C2, C-,
C2, g2
c,,f4
c3, d,, e6


C2





a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
d Results were excluded from summary statistics.  Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
                                                                              108

-------
Table 9.9. Results for Ceriodaphnia chronic test method performed on effluent samples.

LabID
Referee
6
25
25
42
42
62
62
69
69
71
71
105
134
134
141
141
205
205
231
238
238
313
313
333
425

Sample
code
9427
9331
9334
9335
9347
9348
9361
9362
9365
9366
9369
9370
9380
9390
9391
9394
9395
9398
9399
9403
9406
9407
9416
9417
9421
9441

Test date
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/27/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
Survival Information
NOEC
(% sample)
25b
Invalid0
25
25
50e
25=
6.25
6.25
25
25
25
25
Inconclusive*1
12.5
25
25
12.5
25
12.5
25
12.5
12.5
25
25
25
25
LC50
(% sample)
33.0b
Invalid0
33.4
33.7
1Q.T
50.0e
11.8
12.5
35.4
35.4
35.4
35.4
27.9
26.5
26.0
30.9
26.8
28.7
26.8
28.7
16.5
20.2
33.0
28.7
27.3
30.8
Control
mean (%)
100
100
100
90.0
100
100
90.0
100
100
100
100
100
90.0
100
100
100
100
100
100
100
100
100
100
100
80.0
100
Reproduction Information
NOEC
(% sample)
12.5b
Invalid0
12.5
25
50=
25C
6.25
6.25
25
12.5
25
25
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
12.5
IC25
(% sample)
18.5b
Invalid0
21.1
21.9
60.4e
37.8C
8.96
9.38
31.3
21.8
31.1
29.5
14.1
18.6
17.0
16.3
16.2
13.5
14.3
18.2
15.3
17.1
20.9
16.7
17.0
16.9
Control
mean
(neonates)
24.0
13.1
22.7
20.2
15.7
18.1
18.8
20.2
19.9
29.3
20.5
22.2
32.4
27.0
38.4
15.9
15.8
18.7
17.8
22.3
15.7
17.1
24.9
19.1
20.5
15.7
Control CV
(%)
8.10
17.1
13.0
36.7
18.0
19.0
11.2
10.6
14.9
17.8
26.8
24.3
40.3
45.8
9.94
9.58
7.78
35.3
25.2
21.8
25.0
18.2
18.5
33.3
57.0
12.4
Day test
termination
6
6
8
8
7
7
6
6
7
6
6
6
8
6
6
6
6
6
6
8
7
8
6
6
6
8

Flags3

a4


83,84
83,84
b,
b,,d,
C2
C2


c2, e6, g2, 85
482
f4
C2, C3
c2, c3, d4
b,,c,,c2
t>i, C,, C2
b,,b2
bb c2, e2
t>i, c2, e.


e2, e3, e6, g2
c,, e^
                                                                109

-------
 Table 9.9.  Results for Ceriodaphnia chronic test method performed on effluent samples, (continued)

LabID
425
450
Summary
Statistics

Sample
code
9442
9448
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
10/26/99
10/26/99





STD
CV%
STD
cv%
STD
cv%
Survival Information
NOEC
(% sample)
25
12.5
23
6.25
25
25







LC50
(% sample)
26.0
21.9
24
11.8
35.4
28.3
27.5
1.95
7.09%
6.00
21.8%
6.31
23.0%
Control
mean (%)
100
88.9











Reproduction Information
NOEC
(% sample)
6.25
12.5
24
6.25
25
12.5







IC25
(% sample)
<6.25
16.4
24
<6.25
31.3
17.0
17.9
3.31
18.5%
4.88
27.3%
5.90
32.9%
Control
mean
(neonates)
16.4
21.8











Control CV
(%)
11.9
40.0











Day test
termination
8
6












Flags"
c,, e6, g,0
C2, g6











' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
A Results were excluded from summary statistics. Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
' Results were identified as outliers, a probable cause was identified, and therefore results were excluded from summary statistics.
                                                                               110

-------
Table 9.10.  Results for Ceriodaphnia chronic test method performed on receiving water samples.

LablD
Referee
46
238
452
25
69
71
125
134
205
333
417
425
450
Summary
Statistics

Sample code
9426
9353
9404
9451
9333
9364
9368
9385
9388
9396
9418
9433
9440
9447
N
Min
Max
Median
Mean
STD
CV%

Test date
10/12/99
10/26/99
10/12/99
10/12/99
10/12/99
10/12/99
10/13/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99
10/12/99







Survival Information
NOEC
(% sample)
25"
Invalid0
Invalid0
Invalid0
50
25
25
25
25
25
25
25
25
25
10
25
50
25



LC50
(% sample)
35.4b
Invalid0
Invalid0
Invalid0
59.0
40.5
40.6
35.4
37.9
34.6
53.6
40.8
33.0
44.3
10
33.0
59.0
40.6
42.0
8.38
20.0%
Control
mean (%)
100
80.0
40.0
90.0
100
100
100
90.0
100
100
100
100
100
80.0







Reproduction Information
NOEC
(% sample)
25b
Invalid0
Invalid'
Invalid0
25
25
12.5
25
25
12.5
25
25
12.5
25
10
12.5
25
25



IC25
(% sample)
31.3"
Invalid0
Invalid0
Invalid0
53.8
27.5
23.0
31.3
31.6
10.5
28.1
32.6
22.3
33.4
10
10.5
53.8
29.7
29.4
11.0
37.3%
Control
mean
(neonates)
20.5
5.50
7.60
10.2
26.6
15.0
20.7
17.3
31.6
17.2
36.2
18.4
19.5
21.0







Control CV
(%)
30.3
67.6
29.2
45.0
10.7
37.7
20.1
53.1
39.9
34.1
7.79
40.5
13.7
77.3







Day test
termination
6
8
7
8
6
8
6
7
7
6
6
8
8
7








Flags3

a3, a4, bb b2,
b3, c4, d1; d4,
65, f3, g2
a,, a4, bb e2,
g2
a4, f2, g,, g?
e6, gs
C2, f3, g2

C2, g2
f4
b,,c,,c2
b,, e2, e3, e4
c2, f4, g2
Ci
c2, d,, g2







1 Data qualifier flags are described in Table 8.3.
' Results from the referee laboratory were excluded from summary statistics.
: Results from invalid tests were excluded from summary statistics.
                                                                          Ill

-------
Table 9.11.  Precision of point estimates from the Ceriodaphnia chronic test method.
Sample type


Reference toxicant b
Effluent
Receiving water
Average
CV (%)
LC50
Within-lab •
-
7.09
-
7.09
Between-lab '
-
21.8
-
21.8
Total
-
23.0
20.0
21.5
IC25
Within-lab "
-
18.5
-
18.5
Between-lab *
-
27.3
-
27.3
Total
-
32.9
37.3
35.1
" Within- and between-laboratory components of variability were not calculated for the receiving water sample type since no within-replication was provided for this sample type.
b Precision estimates were not calculated for the reference toxicant sample type since the majority of results for this sample type were outside of the test concentration range (ie.,
Table 9.12.  Precision of NOEC values from the Ceriodaphnia chronic test method.
Sample type
Reference
toxicant
Effluent
Receiving water
Endpoint
Survival
Reproduction
Survival
Reproduction
Survival
Reproduction
NOEC Frequency
<6.25%
0
0
0
0
0
0
6.25%
0
2
2
3
0
0
12.5%
0
3
6
17
0
3
25%
1
1
15
4
9
7
50%
0
4
0
0
1
0
100%
35
26
0
0
0
0
Median
(% sample)
100
100
25
12.5
25
25
% within 1
cone, of
median
97.2
83.3
91.3
100
100
100
% beyond 1
cone, of
median
2.78
16.7
8.70
0.00
0.00
0.00
                                                                           112

-------
9.4     Fathead Acute Test Method Results

A total of 29 participant laboratories conducted the fathead acute test method in the WET Study. These
laboratories tested a total of 27 blank samples, 38 reference toxicant samples, 29 effluent samples, and 13
receiving water samples. For each sample tested, a 96-hour LC50 was generated as a test result. Results of
fathead acute testing are shown in Tables 9.13 - 9.16 for each sample type.

9.4.1    Successful Test Completion Rate

A total of 107 fathead acute tests were initiated by 29 participant laboratories. All 107 tests were completed and
met test acceptability criteria. The resulting successful test completion rate calculated in the WET Study for the
fathead acute test method was 100%. Two tests conducted by the referee laboratory were invalid due to  control
survival of 65%. These two tests were initiated on the same day, so poor health of organisms used for testing on
that day is a likely cause.

9.4.2    False Positive Rate

A total of 27 blank samples were analyzed by 22 participant laboratories (Table 9.13).  The LC50 calculated for
all 27 blank samples was >100%, indicating no toxicity and no false positives. The resulting false positive rate
calculated in the WET Study for the fathead acute test method was 0.00%.

9.4.3    Precision

Precision of the fathead acute test method was estimated by calculating the CV of LC50 values obtained for the
reference toxicant, effluent, and receiving water samples. Within-laboratory, between-laboratory, and total CVs
were calculated for the reference toxicant and effluent samples.  Only a total CV was calculated for the receiving
water samples since no within-laboratory replication was provided for this sample type. All participant
laboratory test data for the reference toxicant, effluent, and receiving water samples were used in estimating
precision.  While results on four test samples were identified by ASTM h  statistics as possible outliers, these
results were not excluded from the analysis of precision since a reasonable cause for the outliers could not be
identified (Table 9.1).

Table 9.17  summarizes the precision of point estimates from the fathead acute test method.  CVs were consistent
among sample types; within-laboratory CVs ranged from 7.62% to 10.3%, between-laboratory CVs ranged from
19.2% to 19.7%, and total CVs ranged from 17.2% to 21.8%. As expected, the majority of variability was due
to the between-laboratory component, with within-laboratory CVs averaging 8.96% and between-laboratory CVs
averaging 19.5%. Averaging the CVs based on total variance for the three sample types, a  total CV of 20.0%
was obtained for the fathead acute test method in the WET Study.
                                                  113

-------
Table 9.13.  Results for fathead acute test method performed on blank samples.

LabID
Referee
3
3
6
27
33
42
42
46
60
60
70
73
101
146
157
157
205
231
238
244
251
311
313
406
416
452
452
Summary
Statistics

Sample code
9093
9002
9003
9004
9012
9018
9024
9025
9029
9030
9031
9037
9041
9044
9056
9061
9062
9063
9069
9073
9077
9081
9082
9085
9089
9097
9109
9110
N
Min
Max
Median
Mean

Test date
10/21/99
1 1/04/99
11/04/99
10/21/99
10/21/99
1 1/04/99
10/21/99
10/21/99
11/04/99
10/22/99
10/22/99
10/21/99
10/21/99
10/21/99
10/21/99
1 1/04/99
1 1/04/99
10/21/99
1 1/04/99
1 1/04/99
1 1/04/99
11/04/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99





False positives
False nositive rate
Survival Information
LC50
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
27
>100
>100
>100
>100
0
0.00%
Control mean
(%)
95.0
90.0
100
100
95.0
100
100
95.0
100
90.0
100
100
100
100
100
100
100
100
100
100
100
100
100
100
95.0
100
100
100







Control CV
(%)
8.66
17.1
0.00
0.00
8.66
0.00
0.00
8.66
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
8.66
0.00
0.00
0.00








Flags"




e4

bs
b5
c4, d4
b2, c,, d7, e2, e3
b2, c,, d7, e2, e3



d,,e4


b,,e2
e2, e3
d,, e2


d,,e4

C|










' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                     114

-------
Table 9.14. Results for fathead acute test method performed on reference toxicant samples.

LabID
Referee
3
6
6
25
25
33
41
41
42
46
46
60
62
62
73
73
101
101
105
141
141
146
157
231
231
238
238
244
244
251
251
311
311
313
313
417

Sample code
9096
9001
9006
9007
9008
9009
9017
9020
9021
9026
9027
9028
9032
9033
9034
9042
9043
9046
9047
9049
9052
9053
9059
9060
9067
9068
9071
9072
9075
9076
9079
9080
9083
9084
9087
9088
9102

Test date
11/04/99
10/21/99
11/04/99
1 1/04/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
11/04/99
10/21/99
10/21/99
1 1/04/99
10/21/99
10/21/99
11/04/99
1 1/04/99
11/04/99
11/04/99
10/21/99
10/21/99
10/21/99
11/04/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
1 1/04/99
11/04/99
11/04/99
11/04/99
10/21/99
Survival Information
LC50
(% sample)
Invalid1"
41.8
43.6
46.7
35.4
40.1
32.3
26.8
24.1
31.9
23.4
29.6
33.9
63.9
55.5
35.4
34.2
55.4
53.6
36.7
34.0
37.9
32.7
33.0
34.2
38.2
39.2
46.7
33.7
34.3
34.2
34.6
39.2
42.0
45.6
46.8
42.1
Control mean
(%)
65.0
95.0
100
100
100
100
100
100
100
100
95.0
95.0
100
100
100
100
100
100
100
95.0
95.0
95.0
100
100
100
100
100
100
100
100
100
100
100
100
95.0
95.0
100
Control CV
(%)
7.92
8.66
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
8.66
8.66
0.00
0.00
0.00
0.00
0.00
0.00
0.00
8.66
8.66
8.66
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
8.66
8.66
0.00

Flags"
a.



gi
gi

C2
e2

c4, d4
c4, d4
Ci, d2, d3, e2, e,
g3
g3




e2


d,, e4

e2,e3
e2,e-,
d,, e2
d,,e2




d,
d,


e,, e4
                                               115

-------
 Table 9.14. Results for fathead acute test method performed on reference toxicant samples, (continued)

LabID
425
452
Summary
Statistics

Sample code
9106
9111
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
10/21/99
11/04/99





STD
cv%
STD
cv%
STD
cv%
Survival Information
LC50
(% sample)
43.5
30.9
38
23.4
63.9
36.0
38.6
2.94
7.62%
7.60
19.7%
8.15
21.1%
Control mean
(%)
100
100











Control CV
(%)
0.00
0.00












Flags"

6|











1 Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                     116

-------
Table 9.15.  Results for fathead acute test method performed on effluent samples.

LabID
Referee
25
25
27
27
33
41
41
62
62
70
70
105
105
141
141
146
205
205
231
238
244
406
406
416
416
417
417
425
425
Summary
Statistics

Sample code
9095
9010
9011
9014
9015
9019
9022
9023
9035
9036
9039
9040
9050
9051
9054
9055
9058
9065
9066
9070
9074
9078
9091
9092
9099
9100
9103
9104
9107
9108
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
1 1/4/99
1 1/4/99
1 1/4/99
1 1/4/99
1 1/4/99
11/4/99
11/4/99
11/4/99
11/4/99
11/4/99
11/4/99
1 1/4/99
1 1/4/99
1 1/4/99
1 1/4/99
11/4/99
11/4/99
11/4/99
11/4/99
11/4/99
11/4/99
11/4/99
11/4/99
1 1/4/99
1 1/4/99
1 1/4/99
1 1/4/99
1 1/4/99
11/4/99
11/4/99





STD
CV%
STD
CV%
STD
cv%
Survival Information
LC50
(% sample)
Invalid6
21.8
22.5
18.9
19.8
19.6
16.2
18.3
27.7
23.3
21.8
21.0
23.9
15.8
17.5
15.7
18.2
37.7
32.7
18.9
19.6
19.6
27.7
26.2
22.3
23.1
18.2
18.3
19.7
19.1
29
15.7
37.7
19.7
21.6
2.22
10.3%
4.14
19.2%
4.70
21.8%
Control mean
(%)
65.0
100
100
100
100
100
95.0
100
100
100
100
100
100
100
95.0
95.0
100
100
100
100
100
100
100
100
95.0
100
100
100
95.0
90.0











Control CV (%)
24.0
0.00
0.00
0.00
0.00
0.00
8.66
0.00
0.00
0.00
0.00
0.00
0.00
0.00
8.66
8.66
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
8.66
0.00
0.00
0.00
8.66
17.1












Flags"
a,, d.
gi
gi



e2
b,,e2
64
e4
e.

d2, e2
d2, e2


di, e4
e2, g3
e2, g3
62, e3
di, e2

C,, 64
Ci, e4


e2,e4
e2, e4
d,
d,











1 Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                     117

-------
Table 9.16.  Results for fathead acute test method performed on receiving water samples.

LabID
Referee
6
27
33
70
101
105
146
205
313
406
416
417
425
Summary
Statistics

Sample code
9094
9005
9013
9016
9038
9045
9048
9057
9064
9086
9090
9098
9101
9105
N
Min
Max
Median
Mean
STD
CV%

Test date
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99
10/21/99







Survival Information
LC50
(% sample)
19.6"
17.7
17.7
16.6
18.9
26.2
16.9
20.3
16.2
23.9
19.1
16.9
21.6
25.0
13
16.2
26.2
18.9
19.8
3.41
17.2%
Control mean
(%)
95.0
100
100
100
100
95.0
95.0
100
100
100
100
100
100
100







Control CV
(%)
8.66
0.00
0.00
0.00
0.00
8.66
8.66
0.00
0.00
0.00
0.00
0.00
0.00
0.00








Flags"


e4



e2
di, e4
bb e2

c.

ea, e4








' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
Table 9.17.  Precision of point estimates from the fathead acute test method.
Sample type

Reference toxicant
Effluent
Receiving water
Average
CV (%)
Within-lab "
7.62
10.3
-
8.96
Between-lab •
19.7
19.2
-
19.5
Total
21.1
21.8
17.2
20.0
" Within- and between-laboratory components of variability were not calculated for the receiving water sample type since no within-
replication was provided for this sample type.
                                                      118

-------
9.5     Fathead Chronic Test Method Results

A total of 27 participant laboratories conducted the fathead chronic test method in the WET Study. These
laboratories tested a total of 24 blank samples, 37 reference toxicant samples, 27 effluent samples, and 13
receiving water samples.  For each sample tested, a 7-day survival NOEC, a 7-day growth NOEC, a 7-day
survival LC50, and a 7-day growth IC25 were generated as test results. As described in the WET method
manual, growth was measured as the total weight per replicate divided by the number of original organisms in
that replicate. This definition provides a combined growth and survival endpoint that is more accurately termed
biomass. Also, test concentrations above the survival NOEC were excluded from hypothesis testing conducted
on the growth endpoint. Results of fathead chronic testing are shown in Tables 9.18-9.21 for each sample type.

9.5.1    Successful Test Completion Rate

A total of 101 fathead chronic tests were initiated by 27 participant laboratories.  All  101 tests were completed,
however, tests conducted on samples 9199 and 9118  were invalid due to failure to meet test acceptability criteria
for growth and survival, respectively. The resulting successful test completion rate calculated in the WET Study
for the fathead chronic test method was 98.0%. In addition, the growth NOEC result for sample 9209 was
reported as inconclusive based on an evaluation of the concentration-response relationship (Table 8.4). If this
test is considered unsuccessful (since the test would be repeated in a regulatory context), the successful test
completion rate becomes 97.0%.

9.5.2    False Positive Rate

A total of 24 blank samples were analyzed by 20 participant laboratories  (Table 9.18). No false positives were
observed for the survival endpoints. The survival NOEC was 100% for all 24 blank samples, and the LC50 was
>100% for all 24 blank samples. One false positive was observed for sublethal endpoints. The growth NOEC for
sample 9158 was 50%, and the growth IC25  for this  sample was 93.6%.  The resulting false positive rate
calculated in the WET Study for the fathead chronic test method was 4.35% for the growth NOEC and 4.17% for
the growth IC25. The one false positive that was observed was due to poor survival in a single replicate of the
100% test concentration treatment. For this sample, the survival in the 100% test concentration was 90%, 100%,
90%, and 50% for the 4 replicates, respectively.  Disregarding replicate 4, the survival for this treatment would
be identical to the control survival (95%).

In addition to the false positive reported above, participant laboratories reported a NOEC of less than 100% for
two additional samples  (9145 and 9209), indicating false positive results.  These samples exhibited an interrupted
concentration-response curve. Based on EPA guidance for evaluating concentration-response relationships, the
growth and survival NOEC for sample 9145  was re-calculated and reported as 100%, and the growth NOEC for
sample 9209 was reported as inconclusive (Table 8.4).
                                                 119

-------
9.5.3    Precision

Precision of the fathead chronic test method was estimated by calculating the CV of LC50 and IC25 values
obtained for the reference toxicant, effluent, and receiving water samples. Within-laboratory, between-
laboratory, and total CVs were calculated for the reference toxicant and effluent samples.  Only a total CV was
calculated for the receiving water samples since no within-laboratory replication was provided  for this sample
type.  Results from two samples (9162 and 9163) were excluded from the analysis of precision. These test
samples were identified by ASTM h statistics as possible outliers, and review of data qualifier  flags revealed a
possible cause (Table 9.1). One additional sample (9177) was identified as a possible outlier but was not
excluded from the analysis of precision since a reasonable cause could not be identified.

Table 9.22 summarizes the precision of point estimates from the fathead chronic test method.  CVs for LC50
values were consistent among sample types; within-laboratory CVs ranged from 6.66% to 9.16%, between-
laboratory CVs ranged from  10.5% to 12.0%, and total CVs ranged from 12.4% to 15.1%. As expected, the
majority of variability was due to the between-laboratory component, with within-laboratory CVs averaging
7.91% and between-laboratory CVs averaging 11.3%. Averaging the CVs for the LC50 based on total variance
for the three sample types, a total CV of 13.4% was obtained for the  fathead chronic survival endpoint.

CVs for IC25 values were higher than those for the survival endpoint (LC50). Within-laboratory CVs ranged
from 10.0% to 19.1%, between-laboratory CVs ranged from 12.9% to 17.2%, and total CVs ranged from 19.8%
to 23.1%. Within-laboratory CVs (averaging 14.6%) were only slightly lower than between-laboratory CVs
(averaging 15.1%).  Averaging the CVs for the IC25 based on total variance for the three sample types, a total
CV of 20.9% was obtained for the fathead chronic growth endpoint.

The precision of NOEC values was determined simply by evaluating the range distribution of NOEC values and
the percentage of values falling within and beyond one concentration from the median.  Table 9.23 describes the
precision of NOEC values for the fathead chronic method.  For the survival endpoint, NOEC values spanned four
concentrations for the reference toxicant  sample type and two concentrations for the effluent and receiving water
sample types.  The percentage of values within one concentration of the median was 97.2%, 100%, and 100% for
the reference toxicant, effluent, and receiving water sample types, respectively.

Growth endpoints were less precise, spanning five concentrations for the reference toxicant sample  type and four
concentrations for the effluent and receiving water sample types. The percentage of values within one
concentration of the median was 86.1%,  91.7%, and 76.9% for the reference toxicant, effluent, and receiving
water sample types, respectively. Of the 10 results (for samples 9150, 9177, 9193, 9194, 9212, 9129, 9166,
9161, 9176, and 9180) that were beyond 1 concentration from the median, 5 results (samples 9193, 9194, 9212,
9129, and 9161) were the consequence of interrupted concentration-response curves. For each of these samples,
test concentrations higher than the reported NOEC were not significantly different from the control, but the
NOEC was reported as the concentration below the LOEC based on EPA's concentration-response guidance
(Table 8.4). Three results that were beyond one concentration from the median were from very statistically
sensitive tests, with %MSDs  near or below EPA's recommended lower bound of 9.4% (USEPA, 2000b). Percent

                                                 120

-------
MSDs for these three samples (9150, 9166, and 9180) were 9.2%, 9.7%, and 10.3% respectively. The remaining
two samples (9176 and 9177) that were beyond one concentration from the median were from the same
laboratory (205) and were conducted on the same day, indicating that test organisms or specific test procedures in
this laboratory may have produced conditions that caused greater sensitivity.
                                                121

-------
Table 9.18. Results for fathead chronic test method performed on blank samples.

LabID
Referee
3
3
6
25
27
33
42
42
46
49
49
62
73
101
105
125
231
238
299
311
333
333
406
425

Sample
code
9201
9113
9114
9117
9119
9123
9131
9135
9136
9138
9142
9143
9145
9151
9152
9158
9160
9182
9186
9188
9192
9196
9197
9198
9209

Test date
09/28/99
10/05/99
10/05/99
10/05/99
09/28/99
09/28/99
09/29/99
09/28/99
09/28/99
09/28/99
09/28/99
09/28/99
09/28/99
10/05/99
09/28/99
10/05/99
09/28/99
10/05/99
10/05/99
09/28/99
09/28/99
10/05/99
10/05/99
09/28/99
09/28/99
Survival Information
NOEC
(% sample)
100b
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
LC50
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
Control
mean (%)
100
100
95.0
97.5
100
100
100
90.0
92.5
96.7
95.2
97.5
95.0
100
100
95.0
92.5
97.5
95.0
100
100
97.5
95.0
90.0
100
Control CV
(%)
0.00
0.00
11.4
5.94
0.00
0.00
0.00
9.94
6.32
6.93
6.73
5.94
7.07
0.00
0.00
11.4
6.32
5.94
7.07
0.00
0.00
5.94
11.4
9.94
0.00
Growth Information
NOEC
(% sample)
100b
100
100
100
100
100
100
100
100
100
100
100
100
100
100
50
100
100
100
100
100
100
100
100
Inconclusive0
IC25
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
93.6
>100
>100
>100
>100
>100
>100
>100
>100
>100
Control
mean (mg)
0.477
0.381
0.365
0.371
0.511
0.788
0.383
0.471
0.506
0.300
0.360
0.379
0.293
0.807
0.375
0.392
0.382
0.663
0.373
0.770
0.503
0.602
0.693
0.262
0.643
Control CV
(%)
5.23
7.59
11.7
8.46
4.90
4.50
16.1
10.6
5.02
12.0
9.88
3.77
18.2
8.70
22.9
12.5
16.0
6.38
13.4
2.93
4.58
5.75
17.2
9.51
6.89

Flags3
di,e,



e.
b,
b2
b,
b,,d7
c4, d4, f,
e.
£1
d,,g5


c4, e2
C2, g2
e2, e3
c2, e2, e6, g2
d,, e,, e2, e6


e,,e6
b,, c,, f,
c,, g,, g,
                                                                122

-------
 Table 9.18. Results for fathead chronic test method performed on a blank samples, (continued)

LabID
Summary
Statistics

Sample
code
N
Min
Max
Median
Mean

Test date





False positives
False oositive rate
Survival Information
NOEC
(% sample)
24
100
100
100

0
0.00%
LC50
(% sample)
24
>100
>100
>100
>100
0
0.00%
Control
mean (%)







Control CV
(%)







Growth Information
NOEC
(% sample)
23
50
100
100

1
4.35%
IC25
(% sample)
24
93.6
>100
>100
99.7
1
4.17%
Control
mean (mg)







Control CV
(%)








Flags"







" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results were excluded from summary statistics.  Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
                                                                               123

-------
Table 9.19. Results for fathead chronic test method performed on reference toxicant samples.

LabID
Referee
406
3
6
6
25
27
27
30
30
42
49
62
62
73
73
105
105
134
141
141
157
205
231
238

Sample
code
9204
9199
9112
9115
9116
9122
9125
9126
9127
9128
9137
9144
9147
9148
9149
9150
9156
9157
9165
9168
9169
9173
9177
9181
9184

Test date
10/05/99
10/05/99
09/29/99
09/29/99
09/29/99
10/05/99
10/05/99
10/05/99
09/28/99
09/28/99
10/05/99
10/05/99
10/05/99
10/05/99
09/28/99
09/28/99
09/29/99
09/29/99
09/28/99
09/28/99
09/28/99
09/28/99
09/28/99
09/28/99
09/28/99
Survival Information
NOEC
(% sample)
50"
Invalid0
50
50
50
50
25
25
25
25
50
50
50
50
50
25
50
50
50
50
50
50
50
50
50
LC50
(% sample)
70.7b
Invalid0
83.7
70.7
66.2
51.4
40.6
54.9
53.2
58.4
66.4
68.5
67.9
62.1
64.0
65.0
63.8
65.5
65.0
69.5
73.7
66.8
69.0
68.3
73.7
Control
mean (%)
100
95.0
100
97.5
97.5
100
100
100
95.0
100
95.0
85.0
92.5
90.0
97.5
100
92.5
100
100
100
100
95.0
97.5
100
100
Control CV
(%)
0.00
7.07
0.00
5.94
5.94
0.00
0.00
0.00
7.07
0.00
7.07
10.9
6.32
14.0
5.94
0.00
6.32
0.00
0.00
0.00
0.00
7.07
5.94
0.00
0.00
Growth Information
NOEC
(% sample)
50"
Invalid0
50
50
50
50
25
25
25
25
50
50
50
50
25
12.5
50
50
25
50
50
25
6.25
50
50
IC25
(% sample)
63 .3"
Invalid0
64.2
62.5
56.7
51.4
33.5
47.5
26.8
45.0
57.1
58.4
53.6
54.2
51.5
56.8
58.2
58.9
49.4
57.4
63.2
53.4
20.2
59.1
59.4
Control
mean (mg)
0.435
0.230
0.534
0.250
0.274
0.467
0.671
0.692
0.687
0.574
0.506
0.386
0.275
0.303
0.754
0.697
0.246
0.258
0.617
0.433
0.433
0.505
0.478
0.490
0.442
Control CV
(%)
14.3
9.51
4.26
15.3
7.97
9.59
7.57
7.40
32.5
14.2
7.33
7.55
4.70
16.9
7.83
6.87
14.1
17.0
14.1
15.6
15.6
11.0
7.13
9.53
18.1

Flags"
di,e,
a2> Cb 66> *l>
g?
b2, c2, f,
b,
b,
e,,g,
b,
b,


b,
d,




b2, d2, e2
b2, d2, e2

C2, C,, g.
c2, c,
c.
g?
e2, e3
e?
                                                                 124

-------
 Table 9.19. Results for fathead chronic test method performed on reference toxicant samples, (continued)

LabID
238
299
299
311
311
333
406
416
416
425
425
452
452
Summary
Statistics

Sample
code
9185
9190
9191
9193
9194
9195
9200
9205
9206
9211
9212
9213
9214
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
09/28/99
10/05/99
10/05/99
10/05/99
10/05/99
09/28/99
10/05/99
09/28/99
09/28/99
10/05/99
10/05/99
09/29/99
09/29/99





STD
cv%
STD
cv%
STD
cv%
Survival Information
NOEC
(% sample)
50
25
25
50
50
50
50
50
50
25
6.25
50
50
36
6.25
50
50







LC50
(% sample)
74.5
59.5
49.6
68.8
69.5
70.7
60.9
71.2
70.7
63.3
54.3
69.5
71.2
36
40.6
83.7
66.6
65.1
4.33
6.66%
6.84
10.5%
8.10
1 2 4%
Control
mean (%)
100
100
100
100
100
92.5
90.0
100
100
100
100
97.5
100











Control CV
(%)
0.00
0.00
0.00
0.00
0.00
11.3
9.94
0.00
0.00
0.00
0.00
5.94
0.00











Growth Information
NOEC
(% sample)
50
25
25
<6.25
12.5
50
50
50
50
25
6.25
50
50
36
<6.25
50
50







IC25
(% sample)
56.3
35.7
34.0
56.6
58.0
60.2
58.8
58.9
63.1
57.0
50.4
60.9
57.3
36
20.2
64.2
56.9
52.9
5.30
10.0%
9.11
17.2%
10.5
1 9 9%
Control
mean (mg)
0.532
0.743
0.686
0.413
0.392
0.632
0.235
0.454
0.414
0.563
0.493
0.588
0.710











Control CV
(%)
16.1
2.97
9.18
9.46
3.76
10.3
12.2
2.22
9.56
4.44
9.14
12.8
6.65












Flags"
c3, e2, e6
d,, e2, e6
d,, e2, e6
g5
gs

c,, efi, f,

-------
Table 9.20. Results for fathead chronic test method performed on effluent samples.

LabID
Referee
6
25
30
30
33
33
46
46
101
101
105
125
125
134
134
141
141
157
157
205
205
231
238
416
416

Sample
code
9203
9118
9121
9129
9130
9133
9134
9140
9141
9154
9155
9159
9162
9163
9166
9167
9170
9171
9174
9175
9178
9179
9183
9187
9207
9208

Test date
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
10/05/99
Survival Information
NOEC
(% sample)
25b
Invalid0
12.5
12.5
12.5
12.5
12.5
12.5
12.5
25
25
12.5
12.5
25
12.5
12.5
12.5
12.5
12.5
25
12.5
12.5
12.5
12.5
12.5
25
LC50
(% sample)
31. lb
Invalid0
27.8
21.6
19.9
24.5
24.5
19.3
20.0
32.1
30.5
26.3
21.1
27.6
30.3
31.3
29.7
27.6
29.6
28.4
27.7
25.4
24.1
24.2
21.0
28.8
Control
mean (%)
97.5
70.0
92.5
92.5
85.0
95.0
87.5
95.0
97.5
97.5
97.5
95.0
92.5
92.5
100
100
100
100
97.5
95.0
100
100
100
100
100
97.5
Control CV
(%)
5.94
23.8
6.32
11.3
20.2
11.4
20.0
11.4
5.94
5.94
5.94
11.4
11.3
11.3
0.00
0.00
0.00
0.00
5.94
11.4
0.00
0.00
0.00
0.00
0.00
5.94
Growth Information
NOEC
(%sample)
12.5b
Invalid0
12.5
<6.25
12.5
12.5
12.5
12.5
12.5
25
12.5
12.5
<6.25d
<6.25d
<6.25
6.25
12.5
6.25
12.5
25
<12.5
12.5
12.5
12.5
12.5
12.5
IC25
(%sample)
26.9b
Invalid0
20.6
16.1
16.1
14.9
21.2
16.1
16.3
29.2
20.5
18.4
7.04d
<6.25d
18.4
26.1
23.8
15.6
29.0
26.8
20.5
17.3
19.0
16.9
14.8
23.7
Control
mean (mg)
0.430
0.348
0.448
0.490
0.375
0.686
0.644
0.345
0.395
0.490
0.528
0.366
0.390
0.390
0.607
0.593
0.435
0.435
0.545
0.494
0.403
0.420
0.545
0.451
0.703
0.698
Control CV
(%)
10.7
27.4
11.6
10.9
29.9
20.7
16.6
16.8
14.7
14.3
5.66
5.22
8.36
8.36
7.78
3.49
4.79
4.79
12.4
13.7
6.53
15.4
6.81
6.61
9.21
15.2

Flags3
6]
a.
e3
d2, g5
d2
bi, g2
b,
c4, d4, f,
c4, d4, f,


d2, e2
C3, g3
C3, §3

d,
c2, c3, d2
c2, c3, d2


b,, b4, e7
b,, b4, e7
b,, e2, e3
e2
C2, C3, f,
c,, c,, f,
                                                                 126

-------
 Table 9.20.  Results for fathead chronic test method performed on effluent samples, (continued)

LabID
452
452
Summary
Statistics

Sample
code
9215
9216
N
Min
Max
Median
Mean
Within-lab
Between-
lab
Total

Test date
10/05/99
10/05/99





STD
cv%
STD
cv%
STD
cv%
Survival Information
NOEC
(% sample)
12.5
25
26
12.5
25
12.5







LC50
(% sample)
30.4
34.1
26
19.3
34.1
27.6
26.5
2.42
9.16%
3.16
12.0%
3.99
15 1%
Control
mean (%)
100
95.0











Control CV
(%)
0.00
11.4











Growth Information
NOEC
(%sample)
12.5
25
24
<6.25
25
12.5







IC25
(%sample)
26.2
29.8
24
14.8
29.8
19.8
20.7
3.96
19.1%
2.68
12.9%
4.78
23.1%
Control
mean (mg)
0.461
0.420











Control CV
(%)
10.0
6.90












Flags'
e?
e2











" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
d Results were identified as outliers, a probable cause was identified, and therefore results were excluded from summary statistics.
                                                                              127

-------
Table 9.21. Results for fathead chronic test method performed on receiving water samples.

LabID
Referee
25
27
33
46
62
101
125
134
157
205
231
299
425
Summary
Statistics

Sample
code
9202
9120
9124
9132
9139
9146
9153
9161
9164
9172
9176
9180
9189
9210
N
Min
Max
Median
Mean
STD
CV%

Test date
9/28/99
9/28/99
9/28/99
9/29/99
9/28/99
9/28/99
9/28/99
9/28/99
9/28/99
9/28/99
9/28/99
9/28/99
9/28/99
9/28/99







Survival Information
NOEC
(% sample)
25b
25
12.5
25
12.5
25
12.5
25
25
25
25
25
12.5
25
13
12.5
25
25



LC50
(% sample)
34.9b
31.2
28.7
34.2
28.0
32.5
31.2
37.4
35.4
34.1
41.8
37.2
27.7
37.1
13
27.7
41.8
34.1
33.6
4.22
1 2 6%
Control
mean (%)
97.5
95.0
100
100
100
95.0
97.5
92.5
97.5
90.0
97.5
100
100
100







Control CV
(%)
5.94
7.07
0.00
0.00
0.00
7.07
5.94
6.32
5.94
14.0
5.94
0.00
0.00
0.00







Growth Information
NOEC
(% sample)
25b
25
12.5
25
6.25
25
12.5
<6.25
12.5
25
<6.25
<6.25
12.5
25
13
<6.25
25
12.5



IC25
(%sample)
29.5b
29.3
28.0
29.3
19.0
29.2
25.9
28.6
25.4
29.8
16.1
22.6
18.8
32.4
13
16.1
32.4
28.0
25.7
5.08
19.8%
Control
mean (mg)
0.473
0.510
0.655
0.381
0.563
0.285
0.353
0.382
0.575
0.545
0.433
0.506
0.719
0.675







Control CV
(%)
4.34
5.48
8.70
13.9
11.8
8.35
23.2
16.0
5.88
14.4
17.7
8.43
10.2
12.7








Flags"
6|
63

b2
c4, d,, d4, e6,
f,
d,

C2, g5


b,, b4, d2
b,, e2, e1
e2, efi
C2







1 Data qualifier flags are described in Table 8.3.
' Results from the referee laboratory were excluded from summary statistics.
                                                                            128

-------
Table 9.22. Precision of point estimates from the fathead chronic test method.
Sample type


Reference toxicant
Effluent
Receiving water
Average
CV (%)
LC50
Within-lab •
6.66
9.16
-
7.91
Between-lab '
10.5
12.0
-
11.3
Total
12.4
15.1
12.6
13.4
IC25
Within-lab "
10.0
19.1
-
14.6
Between-lab a
17.2
12.9
-
15.1
Total
19.9
23.1
19.8
20.9
' Within- and between-laboratory components of variability were not calculated for the receiving water sample type since no within-replication was provided for this sample type.
Table 9.23. Precision of NOEC values from the fathead chronic test method.
Sample type
Reference
toxicant
Effluent
Receiving water
Endpoint
Survival
Growth
Survival
Growth
Survival
Growth
NOEC Frequency
<6.25%
0
1
0
2
0
3
6.25%
1
2
0
3
0
1
12.5%
0
2
20
16
4
4
25%
8
10
6
3
9
5
50%
27
21
0
0
0
0
100%
0
0
0
0
0
0
Median
(% sample)
50
50
12.5
12.5
25
12.5
% within 1
cone, of
median
97.2
86.1
100
91.7
100
76.9
% beyond 1
cone, of
median
2.78
13.9
0.00
8.33
0.00
23.1
                                                                         129

-------
9.6     Selenastrum Chronic Test Method Results

A total of 11 participant laboratories conducted the Selenastrum chronic test method in the WET Study.  These
laboratories tested a total of 8 blank samples, 13 reference toxicant samples, 15 effluent samples, and 8 receiving
water samples. Each of these samples was tested with and without the addition of EDTA to the sample and
dilution water. For each test conducted, a 96-hour growth NOEC, a 96-hour growth IC25, and a 96-hour growth
IC50 were generated as test results. The growth endpoint was measured as cell density (cells per mL). Test
acceptability criteria were assessed independently for tests conducted with EDTA (minimum control cell density
of 1x106 cells/mL) and without EDTA (minimum control cell density of 2x105 cells/mL). The  test acceptability
criteria for control variability was determined by calculating the CV of growth in control replicates (control CV
must be less than 20% to meet test acceptability criteria). Results of Selenastrum chronic testing are shown in
Tables 9.24 - 9.31 for each sample type tested with and without the addition of EDTA.

9.6.1    Successful Test Completion Rate

A total of 44 Selenastrum chronic tests were initiated with the addition of EDTA, and 44 tests were initiated
without the addition of EDTA. All 88 tests were completed, however, 16 tests conducted with  EDTA and 15
tests conducted without EDTA were invalid due to failure to meet test acceptability criteria for control growth or
control variability.  The resulting successful test completion rate calculated in the WET Study for the
Selenastrum chronic test method was 63.6% with EDTA and 65.9% without EDTA. In addition, growth
endpoints  for samples  9473 without EDTA, 9455 with EDTA, 9454 without EDTA, and 9468 without EDTA
were reported as inconclusive based on an evaluation of the concentration-response relationship (Table 8.4). If
these tests are considered unsuccessful (since the tests would be repeated in a regulatory context), the successful
test completion rate becomes 61.4% with EDTA and 59.1% without EDTA.

Two distinct patterns of test failures were observed for laboratories conducting the Selenastrum chronic test
method. The first pattern involved the failure of both tests (with and without EDTA) conducted on a given
sample and a given day. This pattern was observed in laboratories 39, 299, 33, and 459. All but 2 of the 18 test
failures from these laboratories were observed to occur for tests both with and without EDTA on a given sample.
For instance, laboratory 39 failed tests with and without EDTA on samples 9466 and 9467, but passed tests with
and without EDTA on samples 9468 and 9469.  The second pattern of test failures involved the failure of all tests
conducted with a given nutrient type (with or without EDTA). Laboratories 33, 406, and 209 exhibited this
pattern. Laboratories 33 and 406 failed all tests (8 of 8) with EDTA and passed all but 1 (7  of 8) test without
EDTA.  Laboratory 209 failed all tests (4 of 4)  without EDTA and passed all tests with EDTA.  This pattern was
further explained by the culturing methods used in each laboratory. Laboratories 33 and 406, which failed all
tests with EDTA, normally cultured organisms without the addition of EDTA.  Laboratory 209, which failed all
tests without EDTA, normally cultured organisms with EDTA.
                                                 130

-------
9.6.2    False Positive Rate

A total of five valid tests were conducted on blank samples with the addition of EDTA (Table 9.24). No false
positives were observed for the growth NOEC, IC25, or IC50.  The growth NOEC was 100% for all 5 blank
samples tested with EDTA, and the IC25 and IC50 values were >100% for all 5 blank samples tested with
EDTA.  The resulting false positive rate calculated in the WET Study for the Selenastrum chronic test method
conducted with the addition of EDTA was 0.00%.

A total of six valid tests were conducted on blank samples without the addition of EDTA (Table 9.25). For the
growth NOEC, one false positive was observed and one test result was determined as inconclusive due to
evaluation of the concentration-response relationship (Table 8.4).  The growth NOEC for sample 9457 tested
without EDTA was reported as 6.25%. Two false positives were observed for the IC25.  The IC25 for samples
9457 and 9473 were reported as 11.7% and 24.6%, respectively. No false positives were observed for the IC50.
The resulting false positive rate calculated in the WET Study for the Selenastrum chronic test method conducted
without the addition of EDTA was 20.0%, 33.3%, and 0.00% for the growth NOEC, IC25, and IC50
respectively.

9.6.3    Precision

Precision of the Selenastrum chronic test method was estimated by calculating the CV of IC25 and IC50 values
obtained for the reference toxicant, effluent,  and receiving water samples.  Within-laboratory, between-
laboratory, and total CVs were calculated for the reference toxicant and effluent samples.  Only a total CV was
calculated for the receiving water samples since no within-laboratory replication was provided for this sample
type.  Results from two samples (9468 w/ EDTA and 9476 w/o EDTA) were excluded from the analysis of
precision. These test samples were identified by ASTM h statistics as possible outliers, and review of data
qualifier flags revealed a possible cause (Table 9.1).

Table 9.32 summarizes the precision of point estimates from the Selenastrum chronic test method conducted with
EDTA.  CVs for IC25 values ranged from 10.9% to 39.5% for the within-laboratory variability component,
8.48% to 20.8% for the between-laboratory variability component, and 23.5% to  40.4% for the total variability
component.  Uncharacteristically, within-laboratory CVs (averaging 25.2%) were higher than between-laboratory
CVs (averaging 14.6%).  This is likely due to the unusually large within-laboratory CV for the effluent sample
type.  Averaging the CVs based on total variance for the three sample types, a total CV of 34.3% was obtained
for the IC25 in the Selenastrum chronic test  method conducted with EDTA. IC50 values were slightly more
precise than IC25 values. For the IC50, CVs averaged 5.83%,  13.3%, and 32.2% for the within-laboratory,
between-laboratory, and total variability components, respectively.

Table 9.33 summarizes the precision of point estimates from the Selenastrum chronic test method conducted
without EDTA. The Selenastrum test method was less precise when conducted without the addition of EDTA.
CVs for IC25 values ranged from 21.0% to 25.6% for the within-laboratory variability component, 60.3% to
83.6% for the between-laboratory variability component, and 24.1% to 87.5% for the total variability component.

                                                 131

-------
Within-laboratory CVs averaged 23.3%, and between-laboratory CVs averaged 72.0%.  Averaging the CVs
based on total variance for the three sample types, a total CV of 58.5% was obtained for the IC25 in the
Selenastrum chronic test method conducted without EDTA. IC50 values were again slightly more precise than
IC25 values. For the IC50, CVs averaged 14.5%, 43.9%, and 58.5% for the within-laboratory, between-
laboratory, and total variability components, respectively.

The precision of NOEC values was determined by evaluating the range and distribution  of NOEC values and the
percentage of values falling within and beyond one concentration from the median.  Table 9.34 describes the
precision of NOEC values for the Selenastrum chronic method. For tests conducted with EDTA, NOEC values
spanned three concentrations for the effluent sample type and four concentrations for the reference toxicant and
receiving water sample types.  The percentage of values within one concentration of the  median was 85.7%,
100%, and 85.7% for the reference toxicant, effluent, and receiving water sample types,  respectively.

For tests conducted without EDTA, NOEC values spanned six concentrations for the reference toxicant sample
type, four concentrations for the effluent sample type, and two concentrations for the receiving water sample type.
The percentage of values within one concentration of the median was 40%, 50%, and 100% for the reference
toxicant, effluent, and receiving water sample types, respectively.
                                                 132

-------
Table 9.24.  Results for Selenastrum chronic test  method performed on blank samples with EDTA.

LabID
Referee
33
333
406
3
36
62
209
459
Summary
Statistics

Sample Code
9497
9459
9489
9491
9457
9463
9473
9481
9499
N
Min
Max
Median
Mean

Nutrient
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA





False positives
False nositive rate

Test date
03/30/00
03/16/00
03/30/00
03/23/00
03/30/00
03/16/00
03/30/00
03/30/00
03/16/00







Growth Information
NOEC
(% sample)
Invalid6
Invalid0
Invalid0
Invalid0
100
100
100
100
100
5
100
100
100

0
0 00%
IC25
(% sample)
Invalid15
Invalid0
Invalid0
Invalid0
>100
>100
>100
>100
>100
5
>100
>100
>100
>100
0
0.00%
IC50
(% sample)
Invalid6
Invalid0
Invalid0
Invalid0
>100
>100
>100
>100
>100
5
>100
>100
>100
>100
0
0.00%
Control Mean
(Cells/mL)
1.71e+06
2.28e+05
1.66e+06
6.03e+05
4.79e+06
1.32e+06
2.98e+06
3.51e+06
1.94e+06







Control CV
(%)
28.0
3.25
29.6
26.7
9.77
9.53
15.2
12.6
5.50








Flags'
a7, d4, g2
a6
a7, d4, g2
a6, a7, d,, d4,
69, gl, J?7

d4
d4, g2
69,811
d4, g2







a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
                                                                         133

-------
Table 9.25. Results for Selenastrum chronic test method performed on blank samples without EDTA.

LabID
Referee
209
333
3
33
36
62
406
459
Summary
Statistics

Sample Code
9497
9481
9489
9457
9459
9463
9473
9491
9499
N
Min
Max
Median
Mean

Nutrient
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA





False positives
False oositive rate

Test date
03/30/00
03/30/00
03/30/00
03/30/00
03/16/00
03/16/00
03/30/00
03/23/00
03/16/00







Growth Information
NOEC
(% sample)
Invalidb
Invalid0
Invalid0
6.25
100
100
Inconclusived
100
100
5
6.25
100
100

1
20.0%
IC25
(% sample)
Invalid6
Invalid0
Invalid0
11.7
>100
>100
24.6
>100
>100
6
11.7
>100
>100
72.7
2
33.3%
IC50
(% sample)
Invalid11
Invalid0
Invalid0
>100
>100
>100
>100
>100
>100
6
>100
>100
>100
>100
0
0.00%
Control Mean
(Cells/mL)
1.64e+06
4.38e+05
9.59e+05
2.65e+05
4.89e+05
1.44e+06
3.68e+06
5.65e+05
7.75e+05







Control CV
(%)
27.8
39.2
28.8
7.29
4.35
17.9
8.72
16.2
8.70








Flags3
a7, d,, d4
a7, e9, g2, g5,
gn
a7, g2


d4, g6
d4, g2, g5
d,, d4, e,, g,, g2
d4, g2







a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
d Results were excluded from summary statistics. Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
                                                                             134

-------
Table 9.26. Results for Selenastrum chronic test method performed on reference toxicant samples with EDTA.

LabID
Referee
33
33
299
406
406
3
36
39
125
209
209
459
459
Summary
Statistics

Sample Code
9495
9460
9461
9485
9492
9493
9455
9465
9469
9477
9478
9479
9500
9501
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Nutrient
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA





STD
CV%
STD
cv%
STD
cv%

Test date
03/16/00
03/23/00
03/30/00
03/30/00
03/23/00
03/30/00
03/16/00
03/30/00
03/30/00
03/30/00
03/09/00
03/17/00
03/23/00
03/30/00











Growth Information
NOEC
(% sample)
<6.25b
Invalid0
Invalid'
Invalid0
Invalid"
Invalid0
Inconclusive11
25
25
6.25
50
25
25
50
7
6.25
50
25







IC25
(% sample)
8.40b
Invalid'
Invalid0
Invalid0
Invalid0
Invalid0
Inconclusive11
39.4
40.7
19.9
42.8
39.6
43.0
50.8
7
19.9
50.8
40.7
39.5
4.30
10.9%
8.20
20.8%
9.26
23.5%
IC50
(% sample)
35.5"
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Inconclusive11
68.4
73.2
54.8
68.1
65.5
75.6
71.7
7
54.8
75.6
68.4
68.2
2.39
3.51%
6.36
9.33%
6.80
9.97%
Control Mean
(Cells/mL)
1.98e+06
2.70e+05
4.37e+05
1.50e+05
5.68e+05
6.18e+05
1.53e+06
1.29e+06
1.13e+06
2.19e+06
2.87e+06
2.27e+06
2.24e+06
1.43e+06











Control CV
(%)
4.43
5.60
3.46
14.4
17.9
13.8
11.1
17.6
7.19
5.07
19.6
2.65
2.70
11.8












Flags"
d,,d4
a&
36
a6, d,, d4, e,, e8,
e9, e,0, e,,, g2
a6, db d4, e,, g,,
§2
a6, bb d,, e,, g,,
g2
d.
d,, e,
d4, e2, e-,, e,, g,
81
69, §2, §11
di, e,, g,,

d4, g2, §6











" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
d Results were excluded from summary statistics. Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
                                                                             135

-------
Table 9.27.  Results for Selenastrum chronic test method performed on reference toxicant samples without EDTA.

LabID
Referee
209
209
299
3
33
33
36
39
125
406
406
459
459
Summary
Statistics

Sample Code
9495
9478
9479
9485
9455
9460
9461
9465
9469
9477
9492
9493
9500
9501
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Nutrient
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA





STD
CV%
STD
cv%
STD
cv%

Test date
03/16/00
03/09/00
03/17/00
03/30/00
03/16/00
03/23/00
03/30/00
03/30/00
03/30/00
03/30/00
03/23/00
03/30/00
03/23/00
03/30/00











Growth Information
NOEC
(% sample)
6.25"
Invalid0
Invalid'
Invalid'
100
100
100
12.5
12.5
6.25
25
6.25
<6.25
25
10
<6.25
100
18.75







IC25
(% sample)
10.2b
Invalid"
Invalid"
Invalid"
>100
>100
>100
15.7
22.7
13.2
19.4
12.9
9.69
35.9
10
9.69
>100
21.0
42.9
11.0
25.6%
35.9
83.6%
37.6
87 5%
IC50
(% sample)
37.6b
Invalid"
Invalid"
Invalid"
>100
>100
>100
21.5
57.3
23.7
33.4
22.0
24.7
53.9
10
21.5
>100
43.7
53.7
12.8
23.8%
29.9
55.6%
32.5
60 5%
Control Mean
(Cells/mL)
1.85e+06
1.48e+06
l.OOe+04
1.71e+05
2.71e+05
2.25e+05
4.77e+05
2.02e+06
1.93e+06
1.18e+06
3.33e+05
4.48e+05
2.27e+06
1.27e+06











Control CV
(%)
9.08
102
0.00
22.3
3.13
0.915
7.23
6.99
11.4
5.49
6.67
9.19
4.18
14.7












Flags"
AI, d4, g6
a?, e9, g2, g,i
a6,d,, e,, g2, g,,
36, a7, d,, e3, e8,
e9, e,,,g2



At, «9
A4, e2, e3, e,, g,
gi
d,, d4, e,, g,, g2
t>i, A,, e,, g.

gi











' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
" Results from invalid tests were excluded from summary statistics.
                                                                         136

-------
Table 9.28.  Results for Selenastrum chronic test method performed on effluent samples with EDTA.

LabID
Referee
33
39
39
299
299
333
406
459
3
36
62
62
125
125
333
Summary
Statistics

Sample Code
9494
9458
9466
9467
9482
9483
9487
9490
9498
9454
9462
9470
9471
9474
9475
9486
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Nutrient
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA





STD
CV%
STD
cv%
STD
cv%

Test date
03/09/00
03/09/00
03/09/00
03/16/00
03/09/00
03/16/00
03/16/00
03/09/00
03/10/00
03/09/00
03/09/00
03/09/00
03/16/00
03/09/00
03/16/00
03/09/00











Growth Information
NOEC
(% sample)
<6.25"
Invalid"
Invalid0
Invalid'
Invalid0
Invalid'
Invalid'
Invalid'
Invalid'
<6.25
6.25
12.5
12.5
12.5
6.25
6.25
7
<6.25
12.5
6.25







IC25
(% sample)
10.7b
Invalid0
Invalid0
Invalid0
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
17.7
13.3
27.2
16.7
36.5
22.6
10.3
7
10.3
36.5
17.7
20.6
8.15
39.5%
1.75
8.48%
8.33
40 4%
IC50
(% sample)
29.9"
Invalid0
Invalid'
Invalid'
Invalid'
Invalid'
Invalid'
Invalid0
Invalid0
45.6
34.2
54.9
46.7
62.6
60.9
58.3
7
34.2
62.6
54.9
51.9
4.22
8.14%
8.91
17.2%
9.86
19.0%
Control Mean
(Cells/mL)
2.18e+06
2.52e+05
1.27e+06
5.02e+05
2.75e+06
4.35e+05
2.64e+06
7.58e+05
5.90e+05
4.66e+06
1.94e+06
3.37e+06
3.90e+06
2.30e+06
2.33e+06
3.54e+06











Control CV
(%)
4.45
6.19
28.9
21.2
45.4
30.4
27.4
21.2
17.0
3.87
18.9
4.29
8.29
10.0
7.95
8.28












Flags"
d,
36, gs
a7, b,, d4, e,, g2
a6, a7, d,, d4, e9,
g2
a7, d4, 63, e,, e,0,
e,i,g?
a6, a7, d,, d4, e3,
e,, e,,,, e,,, g.
a7, b,, d,, d4, g2
a6, a7, bi, d], d4,
69, gb g2
36, b,, b2, g2
d4
di, d4, e6
d4
d4
gi
gi
d,,d4











a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
' Results from invalid tests were excluded from summary statistics.
                                                                         137

-------
Table 9.29. Results for Selenastrum chronic test method performed on effluent samples without EDTA.

LabID
Referee
39
39
299
299
333
333
406
459
3
33
36
62
62
125
125
Summary
Statistics

Sample Code
9494
9466
9467
9482
9483
9486
9487
9490
9498
9454
9458
9462
9470
9471
9474
9475
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Nutrient
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA





STD
CV%
STD
cv%
STD
cv%

Test date
03/09/00
03/09/00
03/16/00
03/09/00
03/16/00
03/09/00
03/16/00
03/09/00
03/10/00
03/09/00
03/09/00
03/09/00
03/09/00
03/16/00
03/09/00
03/16/00











Growth Information
NOEC
(% sample)
6.25"
Invalid*1
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Inconclusive11
50
12.5
6.25
12.5
50
25
6
6.25
50
18.75







IC25
(% sample)
8.85"
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Inconclusive*1
65.1
20.1
11.9
10.3
57.6
43.2
6
10.3
65.1
31.7
34.7
7.27
21.0%
20.9
60.3%
22.2
63.9%
IC50
(% sample)
19.7"
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Invalid0
Inconclusive11
80.2
43.6
35.1
37.5
74.0
68.6
6
35.1
80.2
56.1
56.5
2.97
5.25%
18.1
32.1%
18.4
32.5%
Control Mean
(Cells/mL)
1.44e+06
1.07e+05
3.10e+05
6.28e+04
4.00e+04
2.61e+06
1.44e+06
5.43e+05
6.23e+05
5.11e+05
2.53e+05
2.23e+06
4.11e+06
3.93e+06
8.38e+05
1.03e+06











Control CV
(%)
18.2
75.0
29.0
62.9
41.7
20.9
53.8
26.9
26.2
8.03
6.51
10.2
14.1
16.4
18.5
12.3












Flags"
dl
a6, a7, b,, d4, e9, g2
a7, d,, d4, e,, g2, g5
a6, a7, d,, d4, e3, e,, g2
a6, a7, d,, d4, e3, e9, en,
§2
a7, d,, d4
a7, b,, d,, d4, g2
a7, b,, db d4, e9, g,, g2
a7, b,, b2, g2
d4, g2

d,,d4
d4
d4
gl, §2
gi











a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
11 Results were excluded from summary statistics. Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
                                                                             138

-------
Table 9.30. Results for Selenastrum chronic test method performed on receiving water samples with EDTA.

LabID
Referee
3
36
39
62
125
209
299
333
Summary
Statistics

Sample Code
9496
9456
9464
9468
9472
9476
9480
9484
9488
N
Min
Max
Median
Mean
STD
CV%

Nutrient
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA
EDTA








Test date
03/23/00
03/23/00
03/23/00
03/23/00
03/24/00
03/23/00
03/23/00
03/23/00
03/23/00







Growth Information
NOEC
(% sample)
<6.25b
6.25
25
100C
6.25
12.5
12.5
<6.25
25
7
<6.25
25
12.5



IC25
(% sample)
<6.25b
17.2
24.7
>100C
20.2
24.7
11.1
<6.25
19.7
7
<6.25
24.7
19.7
17.7
6.88
38.9%
IC50
(% sample)
26.8"
35.8
38.0
>100
36.8
64.5
18.1
<6.25
41.2
8
<6.25
>100
37.4
42.6
28.8
67.6%
Control Mean
(Cells/mL)
1.94e+06
4.00e+06
2.33e+06
1.35e+06
3.39e+06
2.46e+06
2.28e+06
2.53e+06
2.79e+06







Control CV
(%)
8.48
2.65
17.1
9.47
3.66
5.59
15.6
19.8
18.7








Flags3
d,,d4
d4
d4, e,
b,, d,, d4, e,,
gi, g-;, g?
b,, d4, f,
gi
69, 82, gn
db d4, e3, e,,
610, g6
b,, d,, d4







' Data qualifier flags are described in Table 8.3.
3 Results from the referee laboratory were excluded from summary statistics.
: Results were identified as outliers, a probable cause was identified, and therefore results were excluded from summary statistics.
                                                                            139

-------
Table 9.31.  Results for Selenastrum chronic test method performed on receiving water samples without EDTA.

LabID
Referee
209
299
3
36
39
62
125
333
Summary
Statistics

Sample Code
9496
9480
9484
9456
9464
9468
9472
9476
9488
N
Min
Max
Median
Mean
STD
CV%

Nutrient
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA
w/o EDTA








Test date
03/23/00
03/23/00
03/23/00
03/23/00
03/23/00
03/23/00
03/24/00
03/23/00
03/23/00







Growth Information
NOEC
(% sample)
12.5"
Invalid0
Invalid0
<6.25
6.25
Inconclusive"1
6.25
<6.25e
6.25
4
<6.25
6.25
6.25



IC25
(% sample)
15.8"
Invalid0
Invalid0
<6.25
10.1
10.3
9.76
32.1e
6.34
5
<6.25
10.3
9.76
8.54
2.05
24.1%
IC50
(% sample)
39.2"
Invalid0
Invalid0
39.1
18.4
>100
15.0
55.9
14.6
6
14.6
>100
28.8
40.5
33.4
82.5%
Control Mean
(Cells/mL)
1.36e+06
2.07e+05
1.12e+05
2.88e+05
1.50e+06
1.06e+06
3.49e+06
1.38e+06
1.14e+06







Control CV
(%)
14.3
23.0
39.0
12.2
13.8
2.89
11.1
6.96
7.77








Flags8
d,,d4
a7, e,, g2, g,,
a6, a7, d,, e3, e9,
^10

d4, e9
b,, d,, d4, e9,
82, §5, g6
b2, d,, f,
gl, S3
b,, d,, g2







" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
d Results were excluded from summary statistics.  Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
e Results were identified as outliers, a probable cause was identified, and therefore results were excluded from summary statistics.
                                                                              140

-------
Table 9.32.  Precision of point estimates from the Selenastrum chronic test method conducted with EDTA.
Sample type


Reference toxicant
Effluent
Receiving water
Average
CV (%)
IC25
Within-Iab a
10.9
39.5
-
25.2
Between-lab a
20.8
8.48
-
14.6
Total
23.5
40.4
38.9
34.3
IC50
Within-lab '
3.51
8.14
-
5.83
Between-lab "
9.33
17.2
-
13.3
Total
9.97
19.0
67.6
32.2
' Within and between-laboratory components of variability were not calculated for the receiving water sample type since no within-replication was provided for this sample type.
Table 9.33. Precision of point estimates from the Selenastrum chronic test method conducted without EDTA.
Sample type


Reference toxicant
Effluent
Receiving water
Average
CV (%)
IC25
Within-lab •
25.6
21.0
-
23.3
Between-lab "
83.6
60.3
-
72.0
Total
87.5
63.9
24.1
58.5
IC50
Within-lab "
23.8
5.25
-
14.5
Between-lab "
55.6
32.1
-
43.9
Total
60.5
32.5
82.5
58.5
a Within- and between-laboratory components of variability were not calculated for the receiving water sample type since no within-replication was provided for this sample type.
                                                                         141

-------
Table 9.34. Precision of NOEC values from the Selenastrum chronic test method.
Nutrient
With EDTA
Without
EDTA
Sample type
Reference
toxicant
Effluent
Receiving water
Reference
toxicant
Effluent
Receiving water
NOEC Frequency
<6.25%
0
1
1
1
0
1
6.25%
1
3
2
2
1
3
12.5%
0
3
2
2
2
0
25%
4
0
2
2
1
0
50%
2
0
0
0
2
0
100%
0
0
0
3
0
0
Median
25
6.25
12.5
18.75
18.75
6.25
% within 1
cone, of
median
85.7
100
85.7
40.0
50.0
100
% beyond 1
cone, of
median
14.3
0.00
14.3
60.0
50.0
0.00
                                                               142

-------
9.7     Mysid Chronic Test Method Results

A total of 11 participant laboratories conducted the mysid chronic test method in the WET Study. These
laboratories tested a total of 8 blank samples, 13 reference toxicant samples, 15 effluent samples, and 8 receiving
water samples. For each sample tested, a 7-day survival NOEC, a 7-day growth NOEC, a 7-day fecundity
NOEC, a 7-day LC50, a 7-day growth IC25, and a 7-day fecundity IC25 were generated as test results. As
described in the WET method manual, growth was measured as the total weight per replicate divided by the
number of original organisms in that replicate.  This definition provides a combined growth and survival endpoint
that is more accurately termed biomass. The fecundity endpoint was measured as the percentage of females with
eggs.  Replicates without identified females were excluded from the analysis of fecundity. Results for the
fecundity endpoint were not calculated if less than 50% of control females produced eggs. Also, test
concentrations above the survival NOEC were excluded from hypothesis testing conducted on growth and
fecundity endpoints.  Results of mysid chronic testing are shown in Tables 9.35 - 9.38 for each sample type.

9.7.1    Successful Test Completion Rate

A total of 44 mysid chronic tests were initiated by  11 participant laboratories. All 44 tests were completed,
however, the test conducted on  sample 9690 was invalid due to failure to meet test acceptability criteria for
survival. The resulting successful test completion rate calculated in the WET Study for the mysid chronic test
method was 97.7%. One test in the referee laboratory was also invalid due to failure to meet test acceptability
criteria for survival. Of the 44 mysid  chronic tests initiated, laboratories  were able to report results for the
fecundity endpoint in only 22 (or 50%) tests.  The remaining 22 tests did not meet  the minimum control fecundity
necessary to report fecundity results.

9.7.2    False Positive Rate

A total of seven valid tests were conducted by seven participant laboratories on blank samples (Table 9.35).  No
false positives were observed for survival, growth, or fecundity endpoints.  The survival NOEC was 100% for all
7 blank samples, and the LC50 was >100% for all 7 blank samples. The NOEC for growth was 100% for all 7
blank samples, and the IC25 was >100% for all 7 blank samples. The fecundity endpoint was only calculable for
four samples. The NOEC for fecundity was  100% for all 4 samples, and the IC25 for fecundity was >100% for
all 4 samples.  The resulting false positive rate calculated in the WET Study for the mysid chronic test method
was 0.00%.

A false positive growth NOEC  result was reported by the participant laboratory for sample 9658. Upon SCC re-
calculation  and verification, the growth NOEC was properly reported as 100%.  SCC test review discovered that
the participant laboratory had calculated growth based on the weight per surviving mysid rather than per original
mysid. Also,  sample 9696 produced an interrupted concentration-response curve,  however, based on EPA
guidance for evaluating concentration-response relationships (USEPA, 2000a), the NOEC for this sample should
be reported as 100%.
                                                 143

-------
9.7.3    Precision

Precision of the mysid chronic test method was estimated by calculating the CV of LC50 and growth IC25 values
obtained for the reference toxicant, effluent, and receiving water samples.  Precision estimates were not made for
fecundity IC25 values, since greater than 20% of this data set consisted of censored values (i.e., > values).
Within-laboratory, between-laboratory, and total CVs were calculated for  the reference toxicant and effluent
samples. Only a total CV was calculated for the receiving water samples since no within-laboratory replication
was provided for this sample type. All participant laboratory test data for the reference toxicant, effluent, and
receiving water samples were used in estimating precision. No test results were identified by ASTM h statistics
as possible outliers.

Table 9.39 summarizes the precision of point estimates from the mysid chronic test method. CVs for LC50
values were consistent among sample types; within-laboratory CVs ranged from 6.09% to 7.06%, between-
laboratory CVs ranged from 24.6% to 30.0%, and total CVs ranged from  25.6% to 37.4%.  As expected, the
majority of variability was due to the between-laboratory component, with within-laboratory CVs averaging
6.58% and between-laboratory CVs averaging 27.3%.  Averaging the CVs for the LC50 based on total variance
for the three sample types, a total CV of 31.2% was obtained for the mysid chronic survival endpoint.

CVs for growth IC25 values were higher than those for the survival endpoint (LC50). Within-laboratory CVs
ranged from 5.26% to 8.69%, between-laboratory  CVs ranged from 36.6% to 40.0%, and total CVs ranged from
37.0% to 45.9%.  Within-laboratory CVs (averaging 6.98%) were much lower than between-laboratory CVs
(averaging 38.3%).  Averaging the CVs for the IC25 based on total variance for the three sample types, a total
CV of 41.3%  was obtained for the mysid chronic growth endpoint.

The precision of NOEC values was determined by evaluating the range and distribution of NOEC values and the
percentage of values falling within and beyond one concentration from the median. Table 9.40 describes the
precision of NOEC values for the mysid chronic test method. For the survival endpoint, NOEC values spanned
three concentrations for all three sample types, and 100% of NOEC values were within one concentration of the
median. Growth endpoints performed similarly, with the exception of the  reference toxicant sample.  Growth
NOEC values spanned four concentrations for the reference toxicant sample type, and 7.69% were beyond one
concentration from the median. The data set for the fecundity endpoint was much smaller than that for the
survival and growth endpoints, since only 50% of tests achieved the necessary fecundity in controls (egg
production in  50% of control females). In this reduced data set, fecundity NOEC values spanned three
concentrations for each sample type and 75%, 87.5%, and 66.7% of values were within one concentration of the
median for the reference toxicant, effluent, and receiving water sample types, respectively.
                                                 144

-------
Table 9.35.  Results for mysid chronic test method performed on blank samples.

LabID
Referee
417
22
39
73
77
101
125
420
Summary
Statistics

Sample
Code
9687
9690
9650
9658
9664
9666
9672
9676
9696
N
Min
Max
Median
Mean

Test date
02/29/00
02/22/00
02/22/00
02/22/00
02/29/00
02/22/00
02/29/00
02/29/00
02/29/00





False positives
False positive rate
Survival Information
NOEC
(%
sample)
Invalid6
Invalid0
100
100
100
100
100
100
100
7
100
100
100

0
0.00%
LC50
(%
sample)
Invalid1"
Invalid0
>100
>100
>100
>100
>100
>100
>100
7
>100
>100
>100
>100
0
0.00%
Control
mean
(%)
76.3
78.1
95.0
90.0
100
100
100
92.5
100







Control
CV (%)
21.5
20.9
8.57
10.4
0.716
0.00
0.00
9.81
0.00







Growth Information
NOEC
(%
sample)
Invalid"
Invalid0
100
100
100
100
100
100
100
7
100
100
100

0
0.00%
IC25
(%
sample)
Invalid0
Invalid'
>100
>100
>100
>100
>100
>100
>100
7
>100
>100
>100
>100
0
0.00%
Control
mean
(mg)
0.322
0.224
0.194
0.403
0.417
0.278
0.360
0.621
0.354







Control
CV (%)
21.9
45.4
16.1
18.1
7.68
8.75
7.86
43.9
10.7







Fecundity Information
NOEC
(%
sample)
Invalid0
-
-
-
100
-
100
100
100
4
100
100
100

0
0.00%
IC25
(%
sample)
Invalid1"
-
-
-
>100
.
>100
>100
>100
4
>100
>100
>100
>100
0
0.00%
Control
mean
(%)
85.7
-
-
-
72.6
.
96.9
77.1
85.4







Control
CV (%)
17.2
-
-
-
31.5
.
7.53
23.4
13.5








Flags"
3], d2, d3,
ds, e,, e2,
Si
a., a5, d5,
e,,g2
as, bt, d5,
e2, e,, g,
a5, ds
61
a5
£1
g2
e,, e2, gs







" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
                                                                           145

-------
Table 9.36.  Results for mysid chronic test method performed on reference toxicant samples.

LabID
Referee
22
22
36
39
39
77
77
101
101
181
238
417
420
Summary
Statistics

Sample
Code
9689
9652
9653
9657
9660
9661
9668
9669
9670
9671
9681
9685
9693
9695
N
Min
Max
Median
Mean
Within-
lab
Between-
lab
Total

Test date
02/22/00
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00
02/22/00
02/22/00
02/29/00
02/29/00
02/29/00
02/22/00





STD
cv%
STD
cv%
STD
cv%
Survival Information
NOEC
(%
sample)
50"
25
25
25
25
25
25
50
12.5
12.5
50
25
50
50
13
12.5
50
25







LC50
(%
sample)
66.9b
42.5
43.9
54.8
35.4
35.4
57.3
63.7
29.2
24.0
70.7
41.6
59.5
70.7
13
24.0
70.7
43.9
48.4
2.94
6.09%
14.5
30.0%
14.8
30.6%
Control
mean
(%)
87.1
97.5
95.0
97.5
82.5
80.0
100
100
100
100
87.5
95.0
100
100











Control
CV (%)
13.8
6.40
8.57
6.40
16.5
15.6
0.00
0.00
0.00
0.00
14.4
8.57
0.940
0.00











Growth Information
NOEC
(%
sample)
50b
25
6.25
25
12.5
25
25
25
12.5
12.5
50
25
25
50
13
6.25
50
25







IC25
(%
sample)
36.4b
30.8
26.4
38.2
19.5
26.7
39.3
38.4
21.0
21.4
60.9
36.4
27.6
61.8
13
19.5
61.8
30.8
34.5
3.00
8.69%
13.8
40.0%
14.1
40.9%
Control
mean
(nig)
0.180
0.211
0.228
0.304
0.419
0.415
0.259
0.251
0.275
0.255
0.173
0.194
0.283
0.338











Control
CV (%)
40.2
23.6
22.4
18.0
32.3
34.4
14.9
21.5
17.3
13.7
11.4
9.81
40.6
6.84











Fecundity Information
NOEC
(%
sample)
50"
-
-
-
-
-
-
-
12.5
12.5
-
25
-
50
4
12.5
50
18.75







IC25
(%
sample)
>50"
-
-
-
-
-
-
-
12.7
>25
-
46.4
-
>50
4
12.7
>50
35.7
33.5






Control
mean
(%)
59.5
-
-
-
-
-
-
-
85.4
84.6
-
80.6
-
81.3











Control
CV (%)
36.3
-
-
-
-
-
-
-
13.8
14.7
-
19.2
-
24.5












Flags"
di, e,,e2, g,,
82
as, b,, c5, d5,
e2, e,, g,,g2
a5, b,, c5, ds,
e2, e3, g,
35, «1, ^6
a5, d2, d3, d5,
e,
35, d2, d^, ds
as
as


as, g,
d5, e2, e.
as, ds, e,, g2
e,,e2











" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                         146

-------
Table 9.37. Results for mysid chronic test method performed on effluent samples.

LabID
Referee
22
36
36
39
73
73
77
125
125
181
181
238
238
417
420
Summary
Statistics

Sample
Code
9686
9651
9654
9655
9659
9662
9663
9667
9674
9675
9678
9679
9682
9683
9691
9694
N
Min
Max
Median
Mean
Within-
lab
Between-
lab
Total

Test date
02/22/00
02/22/00
02/23/00
02/23/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00
02/22/00





STD
cv%
STD
CV%
STD
cv%
Survival Information
NOEC
(%
sample)
25"
12.5
25
12.5
12.5
25
25
12.5
6.25
12.5
12.5
12.5
25
25
25
25
15
6.25
25
12.5







LC50
(%
sample)
33.4"
21.0
30.3
27.5
18.4
34.7
35.2
26.2
12.1
15.6
27.2
26.9
35.4
31.3
30.0
35.4
15
12.1
35.4
27.5
27.1
1.92
7.06%
6.67
24.6%
6.94
25.6%
Control
mean
(%)
95.0
87.5
87.5
100
82.5
100
100
100
90.0
90.0
95.0
95.0
100
100
95.0
100











Control
CV (%)
8.34
10.3
14.4
0.00
7.41
0.940
0.940
0.00
17.3
17.3
8.57
12.6
0.00
0.00
12.6
0.00











Growth Information
NOEC
(%
sample)
25b
12.5
12.5
12.5
6.25
25
25
12.5
6.25
12.5
12.5
12.5
25
25
25
25
15
6.25
25
12.5







IC25
(%
sample)
30.2"
16.5
19.5
16.9
9.32
31.1
30.8
18.3
10.3
8.66
20.2
21.5
31.3
30.1
28.1
30.0
15
8.66
31.3
20.2
21.5
1.13
5.26%
7.88
36.6%
7.96
37.0%
Control
mean
(nig)
0.323
0.175
0.251
0.289
0.408
0.338
0.430
0.292
0.622
0.622
0.193
0.202
0.206
0.205
0.288
0.339











Control
CV (%)
22.7
17.4
18.8
17.7
18.5
5.25
6.34
22.0
67.9
67.9
16.2
17.6
11.1
11.7
28.8
11.2











Fecundity Information
NOEC
(%
sample)
25b
-
-
-
-
25
25
-
6.25
12.5
-
12.5
25
25
-
25
8
6.25
25
25







IC25
(%
sample)
>25b
-
-
-
-
>25
>25
-
>12.5
>12.5
-
>25
>25
>25
-
8.45
8
8.45
>25
>25
19.8






Control
mean
(%)
77.1
-
-
-
-
71.9
71.9
-
92.7
92.7
-
69.4
93.8
89.6
-
83.3











Control
CV (%)
16.6
-
-
-
-
16.2
25.3
-
11.4
11.4
-
33.1
10.5
14.9
-
19.0












Flags2
di, e,, e2, g,
as, b,, ds, e2,
ei, gi
a5, b2, e,, e6
a5, b2> e6
a,, b|, d,
e.
e.
a,
d4, g?
d,, g,
a5, gi
gi
d5, e2, e3, g,,
g5
d,, e2, e,, g,
a<, d,, e.
e2, g5











" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                         147

-------
Table 9.38.  Results for mysid chronic test method performed on receiving water samples.

LabID
Referee
36
73
101
125
181
238
417
420
Summary
Statistics

Sample
Code
9688
9656
9665
9673
9677
9680
9684
9692
9697
N
Min
Max
Median
Mean
STD
cv%

Test date
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00
02/29/00







Survival Information
NOEC
(%
samplle)
25"
12.5
25
6.25
6.25
12.5
12.5
25
25
8
6.25
25
12.5



LC50
(%
sample)
33.6"
21.2
33.5
13.0
9.69
26.5
30.2
28.2
34.4
8
9.69
34.4
27.3
24.6
9.20
37.4%
Control
mean
(%)
86.9
97.5
100
97.5
92.5
97.5
97.5
100
100







Control
CV (%)
14.8
6.40
1.05
6.40
9.81
6.40
6.40
1.83
0.00







Growth Information
NOEC
(%
sample)
12.5"
6.25
12.5
6.25
6.25
12.5
12.5
12.5
25
8
6.25
25
12.5



IC25
(%
sample)
23.5"
14.6
25.3
10.2
8.02
23.2
31.3
15.2
31.3
8
8.02
31.3
19.2
19.9
9.12
45.9%
Control
mean
(mg)
0.337
0.323
0.494
0.350
0.621
0.195
0.209
0.235
0.342







Control
CV (%)
13.5
9.03
8.76
10.1
43.9
18.3
11.0
23.1
11.3







Fecundity Information
NOEC
(%
sample)
-
12.5
25
6.25
6.25
-
6.25
-
25
6
6.25
25
9.375



IC25
(%
sample)
-
>25
>25
>12.5
8.53
-
12.5
-
>25
6
8.53
>25
18.8
18.1


Control
mean
(%)
-
50.0
78.6
95.8
77.1
-
100
-
75.0







Control
CV (%)
-
38.6
32.9
10.5
23.4
-
6.86
-
29.5








Flags"
as> d2, d3,
BI, e2, g.
e*
e,

gi
a5, gi
b,, ds, e2,
e.
a5, ds, e.
e,,e2







" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                         148

-------
Table 9.39. Precision of point estimates from the mysid chronic test method.
Sample type


Reference toxicant
Effluent
Receiving water
Average
CV (%)
LC50
Within-lab •
6.09
7.06
-
6.58
Between-lab "
30.0
24.6
-
27.3
Total
30.6
25.6
37.4
31.2
IC25
Within-lab "
8.69
5.26
-
6.98
Between-lab "
40.0
36.6
-
38.3
Total
40.9
37.0
45.9
41.3
' Within and between-laboratory components of variability were not calculated for the receiving water sample type since no within-replication was provided for this sample type.
Table 9.40. Precision of NOEC values from the mysid chronic test method.
Sample type
Reference
toxicant
Effluent
Receiving water
Endpoint
Survival
Growth
Fecundity
Survival
Growth
Fecundity
Survival
Growth
Fecundity
NOEC Frequency
<6.25%
0
0
0
0
0
0
0
0
0
6.25%
0
1
0
1
2
1
2
3
3
12.5%
2
3
2
7
7
2
3
4
1
25%
7
7
1
7
6
5
3
1
2
50%
4
2
1
0
0
0
0
0
0
100%
0
0
0
0
0
0
0
0
0
Median
(% sample)
25
25
18.75
12.5
12.5
25
12.5
12.5
9.375
% within 1
cone, of
median
100
92.3
75.0
100
100
87.5
100
100
66.7
% beyond 1
cone, of
median
0.00
7.69
25.0
0.00
0.00
12.5
0.00
0.00
33.3
                                                                        149

-------
9.8     Sheepshead Acute Test Method Results

A total of seven participant laboratories conducted the sheepshead acute test method in the WET Study.  These
laboratories tested a total of seven blank samples, seven reference toxicant samples, seven effluent samples, and
seven receiving water samples.  For each sample tested, a 96-hour LC50 was generated as a test result. Results
of sheepshead acute testing are shown in Tables 9.41 - 9.44 for each sample type.

9.8.1    Successful Test Completion Rate

A total of 28 samples were tested at 7 participant laboratories.  All 28 tests were completed and met test
acceptability criteria. The resulting successful test completion rate calculated in the WET Study for the
sheepshead acute test method was 100%.  Testing on two samples (9586 and 9589) was repeated due to
laboratory error.  The laboratory terminated the initial tests after 48 hours rather than the required 96 hours. Due
to this error the laboratory retested samples the following week at their expense. Since this error was not a result
of the test method or test method performance, it was not included in the calculation of the successful test
completion rate.

9.8.2    False Positive Rate

A total of seven blank samples were analyzed by seven participant laboratories (Table 9.41). The LC50
calculated for all 7 blank samples was >100%, indicating no toxicity and no false positives.  The resulting false
positive rate calculated in the WET Study for the sheepshead acute test method was 0.00%.

9.8.3    Precision

Precision of the sheepshead acute test method was estimated by calculating the CV of LC50 values obtained for
the reference toxicant, effluent, and receiving water samples. Since no within-laboratory replication was provided
for this method, CVs were calculated based on total variability and no estimates for precision were available for
within-laboratory and between-laboratory components of variability.  All participant laboratory test data for the
reference toxicant, effluent, and receiving water samples were used in estimating precision. While results on two
test samples (9617 and 9600) were identified by ASTM h statistics as possible outliers,  these results were not
excluded from the analysis of precision due to the small size of the data set.  Exclusion of either result would
translate to 14% of the respective data set.

Table 9.45 summarizes the precision of point estimates from the sheepshead acute test method. CVs based on
total variance ranged from 19.4% to 32.5% with an average CV of 26.0% obtained for the sheepshead acute test
method in the WET Study.
                                                  150

-------
Table 9.41. Results for sheepshead acute test method performed on blank samples.

LabID
Referee
22
29
73
101
238
420
425
Summary
Statistics

Sample Code
9606
9586
9590
9594
9598
9602
9610
9614
N
Min
Max
Median
Mean

Test date
03/07/00
03/17/00
03/07/00
03/07/00
03/07/00
03/07/00
03/07/00
03/07/00





False positives
False positive rate
Survival Information
LC50
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
7
>100
>100
>100
>100
0
0.00%
Control mean
(%)
100
100
95.0
100
100
100
100
100







Control CV
(%)
0.00
0.00
8.66
0.00
0.00
0.00
0.00
0.00








Flags3
d5
U5, ^2' 3> 4' 5^
gs
d,,e2


e2,e3
e2
d5, e6







a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
Table 9.42. Results for sheepshead acute test method performed on reference toxicant samples.

LabID
Referee
22
29
73
101
238
420
425
Summary
Statistics

Sample Code
9609
9589
9593
9597
9601
9605
9613
9617
N
Min
Max
Median
Mean
STD
CV%

Test date
03/14/00
03/17/00
03/14/00
03/14/00
03/14/00
03/14/00
03/14/00
03/14/00







Survival Information
LC50
(% sample)
40.6b
37.6
46.7
37.2
35.4
35.4
39.2
66.0
7
35.4
66.0
37.6
42.5
11.1
26.0%
Control mean
(%)
100
100
100
100
100
100
100
100







Control CV
(%)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00








Flags"
d,, d2, d,
ds, e2, e3> e4> e5>
gg
d2, d,, ds, e2


62, 63, C4
62
d5, e6, g3







' Data qualifier flags are described in Table 8.3.
 Results from the referee laboratory were excluded from summary statistics.
                                                     151

-------
Table 9.43. Results for sheepshead acute test method performed on effluent samples.

LabID
Referee
22
29
73
101
238
420
425
Summary
Statistics

Sample Code
9608
9588
9592
9596
9600
9604
9612
9616
N
Min
Max
Median
Mean
STD
CV%

Test date
03/14/00
03/14/00
03/14/00
03/14/00
03/14/00
03/14/00
03/14/00
03/14/00







Survival Information
LC50
(% sample)
35.4"
33.4
35.4
35.4
18.3
35.4
35.4
34.2
7
18.3
35.4
35.4
32.5
6.30
19.4%
Control mean
(%)
100
100
100
100
100
100
100
100







Control CV
(%)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00








Flags"
d2,d5
d5, e2, e,, e4, e,
d2, e,

g3
e2, e,, e4, e6
e2
d5, e<,







a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
Table 9.44.  Results for sheepshead acute test method performed on receiving water samples.

LabID
Referee
22
29
73
101
238
420
425
Summary
Statistics

Sample Code
9607
9587
9591
9595
9599
9603
9611
9615
N
Min
Max
Median
Mean
STD
CV%

Test date
03/07/00
03/17/00
03/07/00
03/07/00
03/07/00
03/07/00
03/07/00
03/07/00







Survival Information
LC50
(% sample)
25.9b
20.3
17.7
35.4
15.5
22.5
35.4
27.7
7
15.5
35.4
22.5
24.9
8.10
32.5%
Control mean
(%)
100
100
100
100
100
100
100
100







Control CV
(%)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00








Flags8
d2, d,
d5, e2, e3, e4, e5,
g8
d,, d2, e2


e2, e3
e2
d5, e6







a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                     152

-------
Table 9.45. Precision of point estimates from the sheepshead acute test method.
Sample type

Reference toxicant
Effluent
Receiving water
Average
CV (%)
Within-lab"
-
-
-
-
Between-lab"
-
-
-
-
Total
26.0
19.4
32.5
26.0
"Within and between-laboratory components of variability were not calculated for this method since no within-laboratory replication
was provided.
9.9     Sheepshead Chronic Test Method Results

A total of seven participant laboratories conducted the sheepshead chronic test method in the WET Study. These
laboratories tested a total of seven blank samples, seven reference toxicant samples, seven effluent samples, and
seven receiving water samples.  For each sample tested, a 7-day survival NOEC, a 7-day growth NOEC, a 7-day
LC50, and a 7-day growth IC25 were generated as test results.  As described in the WET method manual, growth
was measured as the total weight per replicate divided by the number of original organisms in that replicate. This
definition provides a combined growth and survival endpoint that is more accurately termed biomass. Also, test
concentrations above the survival NOEC were excluded from hypothesis testing conducted on the growth
endpoint. Results of sheepshead chronic testing are shown in Tables 9.46 - 9.49 for each sample type.

9.9.1    Successful Test Completion Rate

A total of 28 samples were tested at 7 participant laboratories.  All 28 tests were completed and met test
acceptability criteria.  The resulting successful test completion rate calculated in the WET Study for the
sheepshead chronic test method was 100%. Testing on sample 9618 was repeated due to laboratory error. The
laboratory inadvertently renewed the test on Day 6 with the wrong sample. Due to this error the laboratory
retested the sample the following week at their expense. Since this error was not a result of the test method or test
method performance, it was not included in the calculation of the successful test completion rate.

9.9.2    False Positive Rate

A total of seven blank samples were analyzed by seven participant laboratories (Table 9.46). No false positives
were observed for survival or growth endpoints. The survival NOEC was 100% for all 7 blank samples, and the
LC50 was >100% for all 7 blank samples. The NOEC for growth was 100% for all 7 blank samples, and the
IC25 was >100% for all 7 blank samples.  The resulting false positive rate calculated  in the WET Study for the
sheepshead chronic test method was 0.00%.
                                                  153

-------
9.9.3    Precision

Precision of the sheepshead chronic test method was estimated by calculating the CV of LC50 and IC25 values
obtained for the reference toxicant, effluent, and receiving water samples.  Since no within-laboratory replication
was provided for this method, CVs were calculated based on total variability and no estimates for precision were
available for within-laboratory and between-laboratory components of variability. All participant laboratory test
data for the reference toxicant, effluent, and receiving water samples were used in estimating precision. No test
results were identified by ASTM h statistics as possible outliers.

Table 9.50 summarizes the precision of point estimates from the sheepshead chronic test method. CVs for LC50
values were extremely low for effluent (2.33%) and receiving water (1.63%) sample types and higher for the
reference toxicant sample type (22.2%). This trend was also observed for CVs of IC25 values, which were
18.4%, 6.12%, and 7.15% for the reference toxicant, effluent, and receiving water sample types, respectively.
The higher variability associated with the reference toxicant sample could be due to the added step of
reconstituting the reference  toxicant ampule sample in each laboratory, or it could be an anomaly associated with
the small data set for this method. Averaging the CVs based on total variance for the three sample types, total
CVs of 8.72% for the LC50 and 10.6% for the IC25 were obtained for the sheepshead  chronic test method in the
WET Study.

The precision of NOEC values was determined by evaluating the range and distribution of NOEC values and the
percentage of values falling within and beyond one concentration from the median. Table 9.51 describes the
precision of NOEC values for the sheepshead chronic method.  For survival and growth endpoints, NOEC values
spanned only two concentrations or did not vary at all between  laboratories. All NOEC values (100%) were
within one concentration of the median.
                                                  154

-------
Table 9.46. Results for sheepshead chronic test method performed on blank samples.

LabID
Referee
22
73
101
181
238
273
420
Summary
Statistics

Sample
Code
9642
9618
9622
9626
9630
9634
9638
9646
N
Min
Max
Median
Mean

Test date
03/21/00
03/29/00
03/21/00
03/21/00
03/21/00
03/21/00
03/21/00
03/21/00





False positives
False nositive rate
Survival Information
NOEC
(% sample)
100"
100
100
100
100
100
100
100
7
100
100
100

0
0.00%
LC50
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
7
>100
>100
>100
>100
0
0.00%
Control
mean (%)
100
100
100
100
100
100
100
100







Control CV
(%)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00







Growth Information
NOEC
(% sample)
100b
100
100
100
100
100
100
100
7
100
100
100

0
0.00%
IC25
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
7
>100
>100
>100
>100
0
0.00%
Control
mean (mg)
1.21
0.825
0.751
0.908
0.913
0.669
1.61
1.92







Control CV
(%)
3.06
8.30
7.08
5.35
11.5
14.8
6.56
6.78








Flags3
di, e,
b2, d5, e2, e3,
e-i, gi, g9

e?, e,
e5
d5, e2
d,,d,
e2, e,







a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                        155

-------
Table 9.47. Results for sheepshead chronic test method performed on reference toxicant samples.

LabID
Referee
22
73
101
181
238
273
420
Summary
Statistics

Sample
Code
9645
9621
9625
9629
9633
9637
9641
9649
N
Min
Max
Median
Mean
STD
CV%

Test date
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00







Survival Information
NOEC
(% sample)
50"
25
50
25
25
50
50
25
7
25
50
25



LC50
(% sample)
64.4"
45.1
68.9
37.2
51.1
66.0
69.5
61.6
7
37.2
69.5
61.6
57.0
12.7
22 2%
Control
mean (%)
100
100
97.5
100
100
100
100
100







Control CV
(%)
0.00
0.00
5.94
0.00
0.00
0.00
0.00
0.00







Growth Information
NOEC
(% sample)
50"
25
25
25
25
25
25
25
7
25
25
25



IC25
(% sample)
54.3b
31.1
44.9
29.6
40.6
49.3
43.1
37.3
7
29.6
49.3
40.6
39.4
7.24
18 4%
Control
mean (mg)
0.889
0.844
0.654
1.26
0.760
0.641
1.60
2.10







Control CV
(%)
8.31
3.65
12.9
3.06
9.77
2.73
5.47
2.15








Flags'
d,, d5, e.
c5, d5, e2, e3,
6s. g]

4, e,, e,
e<
d,,e2
d,
e?, e5







' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                        156

-------
Table 9.48. Results for sheepshead chronic test method performed on effluent samples.

LabID
Referee
22
73
101
181
238
273
420
Summary
Statistics

Sample
Code
9644
9620
9624
9628
9632
9636
9640
9648
N
Min
Max
Median
Mean
STD
cv%

Test date
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00
03/28/00







Survival Information
NOEC
(% sample)
25b
25
25
25
25
25
25
25
7
25
25
25



LC50
(% sample)
34.1"
33.6
34.2
33.7
34.3
35.4
35.4
35.4
7
33.6
35.4
34.3
34.5
0.805
2.33%
Control
mean (%)
100
100
100
100
97.5
100
100
100







Control CV
(%)
0.00
0.00
0.00
0.00
5.94
0.00
0.00
0.00







Growth Information
NOEC
(% sample)
25"
12.5
12.5
25
25
25
12.5
12.5
7
12.5
25
12.5



IC25
(% sample)
29.4b
27.1
26.8
28.7
29.7
30.5
27.9
25.6
7
25.6
30.5
27.9
28.0
1.72
6.12%
Control
mean (rag)
0.994
0.896
0.653
1.23
0.768
0.626
1.56
2.02







Control CV
(%)
4.11
6.51
13.7
10.0
15.5
2.62
2.26
10.1








Flags"
di, e5
c5, d5, e2, e3,
e,,g.

d2, e,, e,
e,
b,, d,, e2
d,
d^, e2, e,







a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                         157

-------
Table 9.49. Results for sheepshead chronic test method performed on receiving water samples.

LabID
Referee
22
73
101
181
238
273
420
Summary
Statistics

Sample
Code
9643
9619
9623
9627
9631
9635
9639
9647
N
Min
Max
Median
Mean
STD
CV%

Test date
03/21/00
03/21/00
03/21/00
03/21/00
03/21/00
03/21/00
03/21/00
03/21/00







Survival Information
NOEC
(% sample)
25"
25
25
25
25
25
25
25
7
25
25
25



LC50
(% sample)
35.4"
35.4
35.3
34.2
34.2
35.4
35.3
35.4
7
34.2
35.4
35.3
35.0
0.569
1 63%
Control
mean (%)
100
92.5
100
100
100
100
100
100







Control CV
(%)
0.00
11.3
0.00
0.00
0.00
0.00
0.00
0.00







Growth Information
NOEC
(% sample)
6.25"
25
12.5
25
12.5
12.5
12.5
12.5
7
12.5
25
12.5



IC25
(% sample)
27.5"
28.7
27.9
29.8
27.3
27.8
25.4
24.0
7
24.0
29.8
27.8
27.3
1.95
7 15%
Control
mean (mg)
1.20
1.08
0.841
0.933
1.05
0.637
1.61
2.08







Control CV
(%)
1.12
4.51
6.93
6.63
3.90
5.79
3.44
5.03








Flags3
di, e,, es, g6
c5, d5, e2, e3,
e<, g,

e?, e,
e5
b,, d,, e2
d,,d,
e2, e5







" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                        158

-------
Table 9.50. Precision of point estimates from the sheepshead chronic test method.
Sample type


Reference toxicant
Effluent
Receiving water
Average
CV (%)
LC50
Within-lab"
-
-
-
-
Between-Iab"
-
-
-
-
Total
22.2
2.33
1.63
8.72
IC25
Within-laba
-
-
-
-
Between-laba
-
-
-
-
Total
18.4
6.12
7.15
10.6
' Within- and between-laboratory components of variability were not calculated for this method since no within-laboratory replicates was provided.
Table 9.51. Precision of NOEC values from the sheepshead chronic test method.
Sample type
Reference
toxicant
Effluent
Receiving water
Endpoint
Survival
Growth
Survival
Growth
Survival
Growth
NOEC Frequency
<6.25%
0
0
0
0
0
0
6.25%
0
0
0
0
0
0
12.5%
0
0
0
4
0
5
25%
4
7
7
3
7
2
50%
3
0
0
0
0
0
100%
0
0
0
0
0
0
Median
(% sample)
25
25
25
12.5
25
12.5
% within 1
cone, of
median
100
100
100
100
100
100
% beyond 1
cone, of
median
0.00
0.00
0.00
0.00
0.00
0.00
                                                                       159

-------
9.10    Silverside Acute Test Method Results

A total of nine participant laboratories conducted the silverside acute test method in the WET Study.  These
laboratories tested a total of 6 blank samples, 12 reference toxicant samples, 12 effluent samples, and 6 receiving
water samples.  For each sample tested, a 96-hour LC50 was generated as a test result. Results of silverside
acute testing are shown in Tables 9.52 - 9.55 for each sample type.

Precision estimates for the silverside acute test were not obtained from the reference toxicant sample type since
greater than 20% (100%) of test results were outside of the test concentration range. All LCSOs for the reference
toxicant sample were >100% sample.  This was caused by precipitation of the spiked copper in the liquid ampule
sample (see Section 5.3).

9.10.1  Successful Test Completion Rate

A total of 36 silverside acute tests were initiated by 9 participant laboratories. Of the 36 tests initiated, 2 tests
(samples 9507 and 9506) failed to meet test acceptability criteria for control survival.  The resulting successful
test completion rate calculated in the WET Study for the silverside acute test method was 94.4%.  The two
invalid tests were initiated in the same laboratory (33) on the same day and were likely due to poor health of test
organisms supplied on that day. In addition, the referee laboratory failed to meet the test acceptability criteria for
control survival in three tests (samples 9531, 9532, and 9533). The referee laboratory successfully repeated
these tests, and results from the repeated tests are shown in Tables 9.52  - 9.54.

9.10.2  False Positive Rate

A total of six blank samples were analyzed by six participant laboratories (Table 9.52). The LC50 calculated for
all 6 blank  samples was >100%, indicating no toxicity and no false positives. The resulting false positive rate
calculated in the WET Study for the silverside acute test method was 0.00%.

9.10.3  Precision

Precision of the silverside acute test method was estimated by calculating the  CV of LC50 values obtained for the
effluent and receiving water samples. Precision estimates were not obtained from the reference toxicant sample
type since greater than 20% (100%) of test results were outside of the test concentration range. Within-
laboratory, between-laboratory, and total CVs were calculated for the effluent samples. Only a total CV was
calculated for the receiving water samples since no within-laboratory replication was provided for this sample
type. All participant laboratory test data for the effluent and receiving water samples were used in estimating
precision. No results were identified by ASTM h statistics as possible outliers.
                                                   160

-------
Table 9.56 summarizes the precision of point estimates from the silverside acute test method.  As expected, the
majority of variability was due to the between-laboratory component, with a within-laboratory CV of 9.91% and
between-laboratory CV of 49.7%. Total CVs were much higher for the effluent sample type (50.7%) than the
receiving water sample type (26.3%). Averaging the CVs based on total variance for the two sample types, a
total CV of 38.5% was obtained for the silverside acute test method in the WET Study.
Table 9.52. Results for silverside acute test method performed on blank samples.

LabID
Referee
29
36
209
244
425
459
Summary
Statistics

Sample code
9532
9504
9512
9518
9528
9536
9538
N
Min
Max
Median
Mean

Test date
12/03/99
11/09/99
11/09/99
1 1/02/99
1 1/09/99
1 1/09/99
1 1/03/99





False positives
False rjositive rate
Survival Information
LC50
(% sample)
>100b
>100
>100
>100
>100
>100
>100
6
>100
>100
>100
>100
0
0.00%
Control mean
(%)
100
100
100
100
100
95.0
100







Control CV
(%)
0.00
0.00
0.00
0.00
0.00
8.66
0.00








Flags"
b2, b3, e2, e4, g,,
gl2
e., e2
d2, d6, e4

64
db e6
b2,d2







! Data qualifier flags are described in Table 8.3.
' Results from the referee laboratory were excluded from summary statistics.
                                                  161

-------
Table 9.53.  Results for silverside acute test method performed on reference toxicant samples.

LabID
Referee
33
29
29
36
36
125
209
209
221
244
244
425
Summary
Statistics

Sample code
9531
9507
9502
9503
9510
9511
9515
9520
9521
9523
9526
9527
9535
N
Min
Max
Median
Mean

Test date
12/03/99
1 1/02/99
1 1/02/99
1 1/02/99
11/02/99
11/02/99
11/02/99
11/09/99
11/09/99
11/02/99
11/02/99
1 1/02/99
1 1/02/99





Survival Information
LC50
(% sample)
>100"
Invalid'
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
11
>100
>100
>100
>100
Control mean
(%)
100
55.0
100
100
95.0
95.0
100
95.0
90.0
95.0
95.0
95.0
100





Control CV
(%)
0.00
8.52
0.00
0.00
8.66
8.66
0.00
8.66
0.00
8.66
8.66
8.66
0.00






Flags"
b2, b3, e2, e4, g,,
g]2
a.
b,,e2
t>i, e2
c5, d2, d5, d6
d2, d5, d6




e,,e4
e4






a Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
                                                      162

-------
Table 9.54.  Results for silverside acute test method performed on effluent samples.

LabID
Referee
29
33
33
36
125
125
221
221
244
425
459
459
Summary
Statistics

Sample code
9533
9505
9508
9509
9513
9516
9517
9524
9525
9529
9537
9540
9541
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
12/03/99
11/09/99
11/09/99
1 1/09/99
1 1/09/99
11/09/99
1 1/09/99
1 1/09/99
1 1/09/99
1 1/09/99
1 1/09/99
11/09/99
1 1/09/99





STD
cv%
STD
cv%
STD
cv%
Survival Information
LC50
(% sample)
28.6"
42.0
20.4
17.9
40.8
23.9
18.3
68.2
60.3
84.7
37.7
38.2
33.0
12
17.9
84.7
38.0
40.4
4.01
9.91%
20.1
49.7%
20.5
50.7%
Control mean
(%)
90.0
100
90.0
90.0
90.0
100
100
90.0
95.0
100
100
100
100











Control CV
(%)
0.00
0.00
0.00
17.1
0.00
0.00
0.00
0.00
8.66
0.00
0.00
0.00
0.00












Flags"
b2, b3, e2, e4, g,,
g]2
d2, e2


d2, d7, e4


d,, e4
d,,e4
e4

d2
d2











' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                     163

-------
Table 9.55. Results for silverside acute test method performed on receiving water samples.

LabID
Referee
33
125
209
221
425
459
Summary
Statistics

Sample code
9530
9506
9514
9519
9522
9534
9539
N
Min
Max
Median
Mean
STD
CV%

Test date
11/02/99
11/02/99
11/02/99
11/02/99
11/02/99
1 1/02/99
1 1/03/99







Survival Information
LC50
(% sample)
47.3"
Invalidc
21.5
28.7
44.3
35.4
39.0
5
21.5
44.3
35.4
33.8
8.90
26.3%
Control mean
(%)
95.0
65.0
100
100
90.0
100
100







Control CV
(%)
8.66
24.0
0.00
0.00
17.1
0.00
0.435








Flags'
e2, g.
a.


d,
d,
b2, d2, e,







" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results from invalid tests were excluded from summary statistics.
Table 9.56. Precision of point estimates from the silverside acute test method.
Sample type

Reference toxicantb
Effluent
Receiving water
Average
CV (%)
Within-lab"
-
9.91
-
9.91
Between-laba
-
49.7
-
49.7
Total
-
50.7
26.3
38.5
* Within- and between-laboratory components of variability were not calculated for the receiving water sample type since no within-
replication was provided for this sample type.
b Precision estimates were not calculated for the reference toxicant sample type since the majority of results for this sample type were
outside of the test concentration range (ie., >100%).
                                                        164

-------
9.11    Silverside Chronic Test Method Results

A total of 10 participant laboratories conducted the silverside chronic test method in the WET Study. These
laboratories tested a total of 7 blank samples, 13 reference toxicant samples, 13 effluent samples, and 7 receiving
water samples.  For each sample tested, a 7-day survival NOEC, a 7-day growth NOEC, a 7-day LC50, and a 7-
day growth IC25 were generated as test results.  As described in the WET method manual, growth was measured
as the total weight per replicate divided by the number of original organisms in that replicate.  This definition
provides a combined growth and survival endpoint that is more accurately termed biomass. Also, test
concentrations above the survival NOEC were excluded from hypothesis testing conducted on the growth
endpoint.  Results of silverside chronic testing are shown in Tables 9.57 - 9.60 for each sample type.

9.11.1  Successful Test Completion Rate

A total of 40 samples were tested at 10 participant laboratories. All 40 tests were completed and met test
acceptability criteria. The resulting successful test completion rate calculated in the WET Study for the silverside
chronic test method was 100%.  The growth endpoint results for sample 9545 were reported as inconclusive
based on an evaluation of the concentration-response relationship (Table 8.4). If this test is considered
unsuccessful (since the test would be repeated in a regulatory context), the successful test completion rate
becomes 97.5%.

9.11.2  False Positive Rate

A total of seven blank samples were analyzed by seven participant laboratories (Table  9.57).  No false positives
were observed for survival or growth endpoints.  The survival NOEC was 100% for all 7 blank samples, and the
LC50 was >100% for all 7 blank samples.  The NOEC for growth was 100% for all 7  blank samples, and the
IC25 was >100% for all 7 blank samples.  The resulting false positive rate calculated in the WET Study for the
silverside chronic test method was 0.00%.

A participant laboratory reported a growth NOEC of less than 100% for sample 9556,  indicating a false positive.
This sample exhibited an interrupted concentration-response curve, with only the 12.5% treatment producing a
significant difference from the control.  Based on EPA guidance for evaluating concentration-response
relationships, the growth NOEC for sample 9556 was re-calculated and reported as 100% (Table 8.4).

9.11.3  Precision

Precision of the silverside chronic test method was estimated by calculating the CV of LC50 and IC25 values
obtained for the reference toxicant, effluent, and receiving water samples. Within-laboratory, between-
laboratory, and total CVs were calculated for the reference toxicant and effluent samples.  Only a total CV was
calculated for the receiving water samples since no within-laboratory replication was provided for this sample
type.  Results from two samples (9582  and  9583) were excluded from the analysis of precision. Samples  9582

                                                 165

-------
and 9583 were identified by ASTM h statistics as possible outliers, and review of data qualifier flags revealed a
possible cause (Table 9.1).

Table 9.61 summarizes the precision of point estimates from the silverside chronic test method. For LC50
values, within-laboratory CVs ranged from 9.17% to 12.2%, between-laboratory CVs ranged from 32.2% to
46.8%, and total CVs ranged from 33.5% to 48.4%. CVs were generally higher for the effluent sample (48.4%)
than for reference toxicant (33.5%) or receiving water (40.0%) samples. As expected, the majority of variability
was due to the between-laboratory component, with within-laboratory CVs averaging 10.7% and between-
laboratory CVs averaging 39.5%.  Averaging the CVs for the LC50 based on total variance for the three sample
types, a total CV of 40.6% was obtained for the silverside chronic survival endpoint.

CVs for IC25 values were slightly higher than those for the survival endpoint (LC50). Within-laboratory CVs
ranged from 7.24% to 22.0%, between-laboratory CVs ranged from 29.1% to 55.5%, and total CVs ranged from
36.4% to 56.0%.  As expected, the majority of variability was due to the between-laboratory component, with
within-laboratory CVs averaging 14.6% and between-laboratory CVs averaging 42.3%.  Averaging the CVs for
the IC25 based on total variance for the three sample types, a total CV of 43.8% was obtained for the silverside
chronic growth endpoint.

The precision of NOEC values was determined by evaluating the range and distribution of NOEC values and the
percentage of values falling within and beyond one concentration from the median.  Table 9.62 describes the
precision of NOEC values for the silverside chronic method. For the survival endpoint, NOEC values spanned
four concentrations for the reference toxicant sample type, five concentrations for the effluent sample type, and
three concentrations for the receiving water sample type. The percentage of values within one concentration of
the median was 90.9%, 84.6%, and 85.7% for the reference toxicant, effluent, and receiving water sample types,
respectively.  Growth endpoints spanned four concentrations for the reference toxicant and effluent sample types
and spanned three concentrations for the receiving water sample type.  The percentage of values within one
concentration of the median was 90.9%, 91.7%, and 85.7% for the reference toxicant, effluent, and receiving
water sample types, respectively. Of the four results (for samples 9561, 9545, 9564, and 9562) that were beyond
one concentration from the median, two (samples 9564 and 9562) were from the same laboratory (125).   During
tests on both samples, the pH in control replicates and lower test concentrations exceeded 9.0.
                                                 166

-------
Table 9.57.  Results for silverside chronic test method performed on blank samples.

LabID
Referee
22
36
39
124
221
333
421
Summary
Statistics

Sample
code
9581
9544
9550
9556
9558
9570
9576
9584
N
Min
Max
Median
Mean

Test date
10/26/99
10/26/99
10/19/99
10/26/99
10/19/99
10/19/99
10/26/99
10/27/99





False positives
False oositive rate
Survival Information
NOEC
(% sample)
100"
100
100
100
100
100
100
100
7
100
100
100

0
0.00%
LC50
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
7
>100
>100
>100
100
0
0.00%
Control
mean (%)
97.5
97.5
97.5
97.5
100
95.0
95.0
97.5







Control CV
(%)
5.94
5.94
5.94
5.94
0.00
7.07
7.07
5.94







Growth Information
NOEC
(% sample)
100b
100
100
100
100
100
100
100
7
100
100
100

0
0.00%
IC25
(% sample)
>100b
>100
>100
>100
>100
>100
>100
>100
7
>100
>100
>100
100
0
0.00%
Control
mean (mg)
0.647
0.566
1.28
0.783
2.33
0.896
1.46
1.07







Control CV
(«/.)
5.66
8.50
12.0
5.75
5.48
16.2
7.76
14.4








Flags3
£5
b,, c,, d,, e5

g5
es. e«
d,, d2, e5
di, e,, e6, g,
d,, d5, e6, g,







' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                         167

-------
Table 9.58. Results for silverside chronic test method performed on reference toxicant samples.

LabID
Referee
22
33
36
36
39
39
124
125
209
333
333
421
421
Summary
Statistics

Sample code
9578
9543
9547
9552
9553
9554
9555
9561
9563
9567
9574
9575
9582
9583
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
10/19/99
10/19/99
10/19/99
10/26/99
10/26/99
10/19/99
10/19/99
10/26/99
10/19/99
10/19/99
10/19/99
10/19/99
10/20/99
10/20/99





STD
cv%
STD
cv%
STD
CV%
Survival Information
NOEC
(% sample)
6.25b
12.5
12.5
12.5
12.5
12.5
12.5
<6.25
6.25
12.5
12.5
25
25°
100C
11
<6.25
25
12.5







LC50
(% sample)
17.1b
28.1
23.5
24.3
21.1
27.3
23.7
11.0
10.0
26.7
34.6
37.0
65.6°
>100Q
11
9.96
37.0
24.3
24.3
2.23
9.17%
7.82
32.2%
8.13
33 5%
Control
mean (%)
97.5
97.5
100
100
97.5
95.0
95.0
100
80.0
100
100
100
100
95.0











Control CV
(%)
5.94
5.94
0.00
0.00
5.94
7.07
7.07
0.00
9.48
0.00
0.00
0.00
0.00
7.07











Growth Information
NOEC
(% sample)
6.25b
12.5
12.5
12.5
12.5
12.5
12.5
<6.25
6.25
12.5
12.5
25
25C
100°
11
<6.25
25
12.5







IC25
(% sample)
15,3"
18.1
17.8
26.1
17.3
17.7
16.9
6.95
8.02
20.6
30.5
25.6
53.7C
>100C
11
6.95
30.5
17.8
18.7
4.10
22.0%
5.43
29.1%
6.80
36.4%
Control
mean (ing)
0.603
0.951
0.941
1.41
1.36
1.06
1.06
2.81
1.34
1.57
1.92
2.09
1.25
0.915











Control CV
(%)
13.1
28.7
14.2
3.36
4.08
13.9
13.9
10.4
8.50
13.5
6.56
3.06
14.9
8.44












Flags"
bb e,, e6
bb c5, ds, e5
Si
b,,e6
bi, e6


e5
d4, e5

Si
gi
b2, d,, e6, g,,
gi
b2, d,, e6, g,,
81











" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results were identified as outliers, a probable cause was identified, and therefore results were excluded from summary statistics.
                                                                            168

-------
Table 9.59. Results for silverside chronic test method performed on effluent samples.

LabID
Referee
22
33
33
39
124
125
125
209
209
221
221
333
421
Summary
Statistics

Sample code
9580
9545
9548
9549
9557
9560
9564
9565
9568
9569
9572
9573
9577
9585
N
Min
Max
Median
Mean
Within-lab
Between-lab
Total

Test date
10/26/99
10/26/99
10/27/99
10/27/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/26/99
10/27/99





STD
cv%
STD
cv%
STD
cv%
Survival Information
NOEC
(% sample)
12.5b
100
12.5
25
25
25
6.25
12.5
25
12.5
50
50
25
50
13
6.25
100
25







LC50
(% sample)
28.9"
>100
57.7
75.1
43.2
48.2
16.4
20.9
35.6
32.0
66.1
70.7
47.2
>100
13
16.4
>100
48.2
54.9
6.70
12.2%
25.7
46.8%
26.5
48 4%
Control
mean (%)
92.5
97.5
100
97.5
97.5
100
100
100
92.5
100
92.5
100
95.0
100











Control CV
(%)
11.3
5.94
0.00
5.94
5.94
0.00
0.00
0.00
6.32
0.00
16.1
0.00
11.4
0.00











Growth Information
NOEC
(% sample)
12.5"
Inconclusive'
12.5
25
25
25
6.25
12.5
25
12.5
50
50
25
50
12
6.25
50
25







IC25
(% sample)
27.8b
Inconclusive0
53.7
56.9
32.5
36.8
8.74
16.1
31.1
30.8
63.0
67.3
40.4
>100
12
8.74
>100
38.6
44.8
3.24
7.24%
24.9
55.5%
25.1
56 0%
Control
mean (mg)
0.561
0.613
0.867
1.02
0.783
2.86
0.984
0.984
1.29
1.19
0.602
0.607
1.58
0.921











Control CV
(%)
11.1
24.8
21.7
7.85
5.75
7.31
13.0
13.0
5.27
14.9
7.37
13.3
2.01
21.8












Flags"
e5
b,, cs, d5, e5,
g2
bi, b2, g,
bi, b2, g,
b,,d7
e5
d4, e5
d4, e5
d7
d,,d,
di, e5
di, e5
d,,g,
di, d7, e6, g,,
Si











' Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
c Results were excluded from summary statistics. Based on EPA guidance for evaluating concentration-response relationships (EPA, 2000a) the test result was inconclusive and the
sample should be retested to obtain a reliable result.
                                                                              169

-------
Table 9.60. Results for silverside chronic test method performed on receiving water samples.

LabID
Referee
22
33
36
124
125
209
221
Summary
Statistics

Sample
code
9579
9542
9546
9551
9559
9562
9566
9571
N
Min
Max
Median
Mean
STD
CV%

Test date
10/19/99
10/19/99
10/19/99
10/19/99
10/19/99
10/19/99
10/19/99
10/19/99







Survival Information
NOEC
(% sample)
12.5"
25
25
12.5
12.5
6.25
25
25
7
6.25
25
25



LC50
(% sample)
26.2"
52.7
43.8
38.4
26.5
11.5
55.8
39.2
7
11.5
55.8
39.2
38.3
15.3
40.0%
Control
mean (%)
85.0
95.0
100
100
100
80.0
97.5
100







Control CV
(%)
6.95
7.07
0.00
0.00
0.00
9.48
5.94
0.00







Growth Information
NOEC
(% sample)
12.5"
25
25
12.5
12.5
6.25
25
25
7
6.25
25
25



IC25
(% sample)
19.1b
36.0
36.0
29.3
18.2
9.35
41.6
32.8
7
9.35
41.6
32.8
29.0
11.4
39.1%
Control
mean (mg)
0.644
0.840
0.941
1.34
2.33
1.34
1.58
1.18







Control CV
(%)
9.60
21.3
14.2
8.12
5.48
8.45
9.46
12.2








Flags"
bi, e5, efi
b,, cs, d5, e5
bi,gi

e5, e6
d4, e5

d,, d2, es







" Data qualifier flags are described in Table 8.3.
b Results from the referee laboratory were excluded from summary statistics.
                                                                         170

-------
Table 9.61. Precision of point estimates from the silverside chronic test method.
Sample type


Reference toxicant
Effluent
Receiving water
Average
CV (%)
LC50
Within-lab '
9.17
12.2
-
10.7
Between-lab '
32.2
46.8
-
39.5
Total
33.5
48.4
40.0
40.6
IC25
Within-lab "
22.0
7.24
-
14.6
Between-lab a
29.1
55.5
-
42.3
Total
36.4
56.0
39.1
43.8
' Within- and between-laboratory components of variability were not calculated for the receiving water sample type since no within-replication was provided for this sample type.
Table 9.62. Precision of NOEC values from the silverside chronic test method.
Sample type
Reference
toxicant
Effluent
Receiving water
Endpoint
Survival
Growth
Survival
Growth
Survival
Growth
NOEC Frequency
<6.25%
1
1
0
0
0
0
6.25%
1
1
1
1
1
1
12.5%
8
8
3
3
2
2
25%
1
1
5
5
4
4
50%
0
0
3
3
0
0
100%
0
0
1
0
0
0
Median
(% sample)
12.5
12.5
25
25
25
25
% within 1
cone, of
median
90.9
90.9
84.6
91.7
85.7
85.7
% beyond 1
cone, of
median
9.09
9.09
15.4
8.33
14.3
14.3
                                                                         171

-------
9.12    Results Summary

According to the study plan for the WET Study, performance of the WET test methods was evaluated by
determining successful test completion rates, false positive rates, and precision (e.g., CVs for point estimates and
ranges for NOEC values).

9.12.1  Successful Test Completion Rate

The successful test completion rate for each WET test method was calculated as the percentage of initiated and
properly terminated tests that met the test acceptability criteria as specified in the WET method manuals (see
Section 9.1.2).  Table 9.63 summarizes the successful test completion rates achieved in the WET Study for each
of the test methods. Successful test completion rates were 100% for 4 of the 10 test methods, indicating that no
invalid tests were conducted for those methods. The successful test completion rate was above 94% for 8 of the
10 methods. Only the Ceriodaphnia chronic and Selenastrum chronic test methods achieved successful test
completion rates below 90%.  The 82.0% successful test completion rate for the Ceriodaphnia chronic test
method resulted from a small number of laboratories that failed 50% or more of the tests that they conducted (see
Section 9.3.1).  Successful test completion rates for the Selenastrum chronic test method were 63.6% and 65.9%
for tests conducted with and without EDTA, respectively. The lower successful test completion rates for the
Selenastrum chronic test method appeared to be partially attributed to organism culturing methods and the
presence or absence of EDTA in the culture medium (see Section 9.6.1).
Table 9.63.  Successful test completion rates for test methods evaluated in the WET Study.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic (with EDTA)
Selenastrum chronic (without EDTA)
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
N
104
122
107
101
44
44
44
28
28
36
40
No. of invalid tests
5
22
0
2
16
15
1
0
0
2
0
Successful test
completion rate
(%)
95.2
82.0
100
98.0
63.6
65.9
97.7
100
100
94.4
100
                                                  172

-------
9.12.2  False Positive Rate

False positive rates were calculated as the percentage of valid tests showing toxicity in blank samples (see Section
9.1.3). Table 9.64 summarizes the false positive rates observed for each test method evaluated in the WET
Study. For survival endpoints, false positive rates were 0.00% for all test methods evaluated.  False positive
rates for reproduction endpoints were 3.70% for the Ceriodaphnia chronic test method and 0.00% for the mysid
chronic test method.  For the growth endpoints, false positive rates were 0.00% for all test methods except the
fathead chronic test method and the Selenastrum chronic test method conducted without EDTA.  For the fathead
chronic test method, false positive rates were 4.17% for the growth IC25 and 4.35% for the growth NOEC. For
the Selenastrum chronic test method conducted without EDTA, false positive rates were 33.3% for the growth
IC25 and 20.0% for the growth NOEC. These relatively high false positive rates for the Selenastrum  chronic test
method may be due in part to a small sample size. The false positive rate of 33.3% reflects only 2 false positives
out of 6 valid test results, and the 20.0% false positive rate reflects only 1 false positive out of 5 valid tests.
When the growth  IC50 was calculated for the same 6 tests, no false positives were observed.  Also, no false
positives were observed for the Selenastrum chronic test performed with the addition of EDTA.  In summary,
false positives were observed for only 3 of the 10 test methods (Ceriodaphnia chronic, fathead chronic, and
Selenastrum chronic performed without EDTA), and the rate of false positives was below 5% for all test methods
except for the Selenastrum chronic test method performed without EDTA.
Table 9.64. False positive rates for test methods evaluated in the WET Study.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
(with EDTA)
Selenastrum chronic
(without EDTA)
Mysid chronic
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
N
33
27
27
24
5
6
7
7
7
6
7
False positive rate (%)
Survival endpoint
LC50
0.00
0.00
0.00
0.00
-
-
0.00
0.00
0.00
0.00
0.00
NOEC
-
0.00
-
0.00
-
-
0.00
-
0.00
-
0.00
a N for the growth NOEC endpoint was 23.
b N for the growth NOEC endpoint was 5.
c N for the fecundity endpoints was 4.
Growth endpoint
IC25
-
-
-
4.17
0.00
33.3
0.00
-
0.00
-
0.00
NOEC
-
-
-
4.35a
0.00
20.0"
0.00
-
0.00
-
0.00

Reproduction endpoint
IC25
-
3.70
-
-
-
-
o.ooc
-
-
-
-
NOEC
-
3.70
-
-
-
-
o.ooc
-
-
-
-

                                                  173

-------
9.12.3   Precision

The precision of test methods evaluated in the WET study was estimated by calculating a CV for point estimates
(i.e., LC50s and IC25s). For test methods that included within-laboratory replication of sample types, CVs were
calculated based on within-laboratory, between-laboratory, and total variance (see Section 9.1.4).  These CVs
were calculated independently for each sample type and averaged to obtain final estimates of within-laboratory,
between-laboratory, and total variability.  Table 9.65 shows the CVs calculated for each test method based on
within-laboratory, between-laboratory, and total variability. As expected, the within-laboratory variability
observed for most test methods was much lower than between-laboratory variability. Within-laboratory CVs
ranged from 6.58% to 12.1% for survival endpoints and from 6.98% to 25.2% for
Table 9.65. Within-laboratory, between-laboratory, and total variability observed for test methods
evaluated in the WET Study.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
(with EDTA)
Selenastrum chronic
(without EDTA)
Mysid chronic0
Sheepshead acute1
Sheepshead chronicf
Silverside acute
Silverside chronic
CV (%)"
Survival endpointb
Within-
laboratory
12.1
7.09
8.96
7.91
-
-
6.58
-
-
9.91
10.7
Between-
laboratory
24.0
21.8
19.5
11.3
-
-
27.3
-
-
49.7
39.5
Total"
29.0
21.5
20.0
13.4
-
-
31.2
26.0
8.72
38.5
40.6
Sublethal endpointc
Within-
laboratory
-
18.5
-
14.6
25.2
23.3
6.98
-
-
-
14.6
Between-
laboratory
-
27.3
-
15.1
14.6
72.0
38.3
-
-
-
42.3
Total"
-
35.1
-
20.9
34.3
58.5
41.3
-
10.6
-
43.8
d Within-laboratory, between-laboratory, and total CVs presented are averaged across sample types.
b CVs for the survival endpoint are based on LC50 values.
c CVs for the sublethal endpoint are based on IC25 values.
A CVs based on total variance may not necessarily be greater than CVs based on within and between-laboratory variance because the
CVs presented are averaged across sample types. No within-laboratory replication was provided for the receiving water sample type,
so CVs based on within and between-laboratory variance are averaged across only the reference toxicant and effluent sample types;
CVs based on total variance are averaged across the reference toxicant, effluent, and receiving water sample types. See Sections 9.2 -
9.11 for precision estimates calculated independently for each sample type and variance component.
c For the mysid chronic test method, sublethal enpdoint CVs are for the growth endpoint.
r Within and between-laboratory components of variability were not estimated for the sheepshead acute and chronic test methods
because no within-laboratory replication was provided for these methods.
                                                      174

-------
sublethal endpoints. Within-laboratory variability was highest for the Selenastrum test method, with calculated
IC25 CVs of 25.2% with EDTA and 23.3% without EDTA.  The within-laboratory variability for this method
was much lower when calculated for the growth IC50. CVs for the Selenastrum growth IC50 endpoint were
5.83% with EDTA and 14.5% without EDTA.  Within-laboratory CVs calculated in the WET Study represent
the variability of results between replicate samples tested simultaneously in a given laboratory. Therefore, these
within-laboratory CVs are expected to be lower than previously reported within-laboratory CVs based on
reference toxicants tested over time in a given laboratory (as in USEPA, 2000b).

Between-laboratory variability was higher than within-laboratory variability for all test methods with the
exception of the Selenastrum chronic test method performed with EDTA. Between-laboratory variability
observed in the WET Study ranged from 11.3% to 49.7% for survival endpoints and from 14.6% to 72.0% for
sublethal endpoints. Similarly to within-laboratory CVs, the growth IC50 endpoint for the Selenastrum chronic
test method was less variable than the growth IC25 endpoint. The between-laboratory CV for the  growth IC50
endpoint was 13.3% (versus 14.6% for the IC25) when the Selenastrum chronic test was conducted with EDTA
and 43.9% (versus  72.0% for the IC25) when the Selenastrum chronic test was conducted without EDTA.

The CVs calculated based on total variance were used to summarize the precision of test methods evaluated in the
WET Study. These precision estimates are averaged across sample types and incorporate both within-laboratory
and between-laboratory components of method variability. Summarized precision estimates based on total
variance are presented in Table 9.66. CVs for acute WET test methods ranged from 20.0% to 38.5%. CVs for
chronic  WET test methods ranged from 8.72% to 40.6% for survival endpoints and from 10.6% to 58.5% for
sublethal endpoints. These CVs are well within the range of CVs previously reported for WET test methods.
USEPA (1988) reported multilaboratory precision (CVs) of 22-167% (with a weighted mean of 50%) for acute
WET methods testing reference toxicants. USEPA (1991) reported interlaboratory precision (CVs) of acute
methods as 34.6% to 50.1% for the sheepshead acute method and 22.3% to 59.5% for the mysid acute method.
Interlaboratory precision (CVs) of chronic methods was previously reported as 20.5% to 41.1% for the
Ceriodaphnia chronic test method, 31% for the fathead chronic method, and 44.2% for the sheepshead chronic
method  (USEPA, 1991).
                                                 175

-------
Table 9.66.  Summarized precision estimates (CVs) for test methods evaluated in the WET Study.
Test method
Ceriodaphnia acute
Ceriodaphnia chronic
Fathead acute
Fathead chronic
Selenastrum chronic
(with EDTA)
Selenastrum chronic
(without EDTA)
Mysid chronicd
Sheepshead acute
Sheepshead chronic
Silverside acute
Silverside chronic
CV (%)"
Survival endpointb
29.0
21.5
20.0
13.4
-
-
31.2
26.0
8.72
38.5
40.6
Sublethal endpoint'
-
35.1
-
20.9
34.3
58.5
41.3
-
10.6
-
43.8
1 CVs presented are based on total variance and averaged across sample types.
b CVs for the survival endpoint are based on LC50 values.
c CVs for the sublethal endpoint are based on IC25 values.
d For the mysid chronic test method, sublethal enpdoint CVs are for the growth endpoint.
                                                      176

-------
10.0 REFERENCES

American Society for Testing and Materials.  1997.  Standard practice for conducting an interlaboratory study to
        determine the precision of a test method. E 691-92. In ASTM Standards on Precision and Bias for
        Various Applications, 5th Ed.  West Conshohocken, PA, pp.309-328.

DynCorp. 1996. Directory of Environmental Testing Laboratories. S.J. Kolb, D.J. Lesko, B.W. Schorer, B.P.
        Moser, B.E. Lane, R.K. Holder, and K. Lee, eds., DynCorp, Alexandria, VA.

SAS Institute.  2000.  SAS Procedure Guide. Version 8, Gary, NC.

Tidepool Scientific Software.  1996. ToxCalc Toxicity Data Analysis Software, User's Guide. Version 5.0,
        McKinleyville, CA.

U.S. Environmental Protection Agency. 1988. Availability, Adequacy, and Comparability of Testing
        Procedures for the Analysis of Pollutants Established Under Section 304(h) of the Federal Water
        Pollution Control Act. EPA/600/9-87/030.  U.S. Environmental Protection Agency, Environmental
        Monitoring Systems Laboratory (currently, National Exposure Research Laboratory), Cincinnati, OH.

U.S. Environmental Protection Agency. 1991. Technical Support Document for Water Quality-Based Toxics
        Control. EPA/505/2-90/001.  U.S. Environmental Protection Agency, Office of Water Enforcement and
        Permits, and  Office of Water Regulations and Standards, Washington, DC.

U.S. Environmental Protection Agency. 1993. Methods for Measuring the Acute Toxicity of Effluents and
        Receiving Waters to Freshwater and Marine Organisms, 4th ed., EPA/600/4-90/027F. U.S.
        Environmental Protection Agency, Environmental Monitoring Systems Laboratory (currently, National
        Exposure Research Laboratory), Cincinnati, OH.

U.S. Environmental Protection Agency. 1994a. Short-term Methods for Estimating the Chronic Toxicity of
        Effluents and Receiving Waters to Freshwater Organisms, 3rd ed., EPA/600/4-91/002. U.S.
        Environmental Protection Agency, Environmental Monitoring Systems Laboratory (currently, National
        Exposure Research Laboratory), Cincinnati, OH.

U.S. Environmental Protection Agency. 1994b. Short-term Methods for Estimating the Chronic Toxicity of
        Effluents and Receiving Waters to Marine and Estuarine Organisms, 2nd ed., EPA/600/4-91/003.  U.S.
        Environmental Protection Agency, Environmental Monitoring Systems Laboratory (currently, National
        Exposure Research Laboratory), Cincinnati, OH.
                                                 177

-------
U.S. Environmental Protection Agency. 1996. Clarifications Regarding Flexibility in 40 CFR Part 136 Whole
        Effluent Toxicity (WET) Test Methods, April 10, 1996, memorandum from Tudor Davies, U.S.
        Environmental Protection Agency, Office of Science and Technology, Washington D.C.

U.S. Environmental Protection Agency. 1999. Errata for Effluent and Receiving Water Toxicity Test Manuals:
        Acute Toxicity of Effluents and Receiving Waters to Freshwater and Marine Organisms; Short-term
        Methods for Estimating the Chronic Toxicity of Effluents and Receiving Waters to Freshwater
        Organisms; and Short-term Methods for Estimating the Chronic Toxicity of Effluents and Receiving
        Waters to Marine andEstuarine Organisms.  January 1999.  EPA/600/R-98/182. U.S. Environmental
        Protection Agency, Office of Research and Development, Duluth, MN.

U.S. Environmental Protection Agency. 2000a. Method Guidance and Recommendations for Whole Effluent
        Toxicity (WET) Testing (40 CFR Part 136). EPA/821/B-00/004.  U.S. Environmental Protection
        Agency, Office of Water, Washington, D.C.

U.S. Environmental Protection Agency. 2000b. Understanding and Accounting for Method Variability in
        Whole Effluent Toxicity Applications Under the National Pollutant Discharge Elimination System
        Program. EPA/833/R-00/003.  U.S. Environmental Protection Agency, Office of Wastewater
        Management, Washington, D.C.
                                                178

-------