United States
Environmental Protection
Agency
Office of Water
4301
EPA-820-B-95-001
March 1995
Water Quality
Guidance for the
Great Lakes
System:
Supplementary
Information
Document  (SID)

-------
                               DISCLAIMER

    This document has been reviewed  by the Office of Science and
    Technology,  U.S. Environmental Protection Agency, and approved
    for publication as a support document for the final Water
    Quality Guidance for the Great Lakes System.  Mention of trade
    names and commercial products does not constitute endorsement
    of their use.
                          AVAILABILITY NOTICE

    This document i's available for a  fee  upon  written request or
    telephone call to:
fA
             National Technical  Information Center  (NTIS)
                      U.S. Department of  Commerce
                          5285  Port  Royal  Road
                        Springfield,  VA  22161
                             (800) 553-6847
                             (703) 487-4650
                   NTIS Document Number:   PB95187266

                                  or

       Education Resources  Information Center/Clearinghouse  for
    Science, Mathematics, and  Environmental Education  (ERIC/CSMEE)
                     1200 Chambers  Road,  Room 310
                          Columbus,  OH  43212
                             (800) 276-0462
                             (614) 292-6717
                          ERIC Number:  D046
                                           U.S. Environmental Protectfon Aginey
                                           Region 5, Library (PL-12J)
                                           77 West Jackson Boulevard, 12th Fl§§f
                                           Chicago, IL 60604-3590

-------
                             FOREWORD
     On March 23', 1995, EPA published the final Water Quality
Guidance for the Great Lakes System in the Federal Register.  In
addition to the Guidance itself, the Federal Register publication
included: a brief summary of the process for developing the final
Guidance; a summary of the final Guidance; a summary of the
consultation with the U.S. Fish and Wildlife Service under the
Endangered Species Act; a summary of EPA's analyses pursuant to
the Regulatory Flexibility Act and the Paperwork Reduction Act;
and a summary of EPA's response to Executive Order 12866.

     The purpose of this Supplementary Information Document (SID)
is to provide a more detailed description of the final Guidance,
including a statement of the basis and purpose of the final
Guidance; a discussion of the major changes from the proposal;
and an analysis of issues raised in comments received on the
proposed Guidance and subsequent documents in the Federal
Register.

     The SID is divided into 12 major sections and further
divided within the sections into numerous subsections (see list
of major sections below and the table of contents).   Each
subsection presents a summary of the proposal for that issue,
significant comments received  (with EPA's response to each), and
the requirements for the final Guidance for that issue.
Responses to all comments received are provided in the Response
to Comment document with is available in the docket for the final
Guidance.
     Section I
     Section II
     Section III
     Section IV
     Section V
     Section VI
     Section VII
     Section VII-I
     Section IX
     Section X
     Section XI

     Section XII
Background
Regulatory Requirements
Aquatic Life
Bioaccumulation Factors
Human Health
Wildlife
Antidegradation
Implementation Procedures
Executive Order 12866
Regulatory Flexibility Act
Enhancing the Intergovernmental Partnership
under Executive Order 12875
Paperwork Reduction Act

-------
                        TABLE OF CONTENTS

I.  BACKGROUND	   1

A.   Description of Resource  	   1

B.   Environmental Problems in the Great Lakes System  ....   2
     1.   Historical Nature of Environmental Problems  in the
          Great Lakes System	   2
     2.   Current Trends of Environmental Problems in  the
          Great Lakes System	   3
          a.   Impacts From Chemical Contamination   	   3
          b.   Trends in Contaminant Levels 	   7
          c.   Fish Consumption Advisories  	   8

C.   History of the Great Lakes Water Quality Guidance  ...  11
     1.   The Great Lakes Toxic Substances Control
          Agreement	11
     2.   The Great Lakes Water Quality Initiative   	  11
     3.   The Great Lakes Critical Programs Act of 1990 ...  12
     4.   Principles Underlying the Final Water Quality
          Guidance for the Great Lakes System	12
          a.   Use the Best Available Science to Provide
               Protection to Human Health, Wildlife, and
               Aquatic Life	12
          b.   Recognize the Unique Nature of the Great
               Lakes Basin Ecosystem	13
          c.   Promote Consistency in Standards and
               Implementation Procedures While Allowing
               Appropriate Flexibility to States and Tribes  .  13
          d.   Establish Equitable Strategies to Control
               Pollution,Sources  	  16
          e.   Promote Pollution Prevention Practices  ....  17
          f.   Provide Accurate Assessment of Costs and
               Benefits	17

D.   Progress on Other Programs to Protect and Restore the
     Great Lakes System	17
     1.   The Great Lakes Toxic Reduction Effort  	  18
          a.   Pathway Track	18
          b.   Virtual Elimination Project	18
          c.   Lake Michigan Enhanced Monitoring Program. .  .  19
     2.   Clean Air Act Amendments	19
     3.   ARARs and the Superfund Program	  .  21
     4.   RAPs and LaMPs	21
          a.   RAPs	21
          b.   LaMPs	23
     5.   Sediments	24

E.   Science Advisory Board Review  	  26

F.   References	26

II.  REGULATORY REQUIREMENTS  	  29

                                ii

-------
A.   Scope and Purpose	29

B.   Definitions	29
     1.   Proposal	29
     2.   Comments	29
     3.   Final Guidance	30

C.   Adoption and Application of Criteria, Methodologies,
     Policies, and Procedures 	  32
     1.   Adoption of Tier I Criteria and Methodologies  ...  32
          a.   Proposal	  .  32
          b.   Comments	32
          c.   Final Guidance	37
     2.   Adoption and Application of Tier II Methodologies  .  37
          a.   Proposal	  37
          b.   Comments	37
          c.   Final Guidance	40
     3.   Application of Anti-backsliding Provisions of  the
          Clean Water Act	41
          a.   Proposal	41
          b.   Comments	41
          c.   Final Guidance	42
     4.   Basin-wide Application of Criteria and Values  ...  46
          a.   Proposal	46
          b.   Comments	46
          c.   Final Guidance	48
     5.   Pollutants Subject to Federal, State, and Tribal
          Requirements	48
          a.   Proposal	48
          b.   Comments	48
          c.   Final Guidance	56
     6.   Scientific Defensibility Exclusion   	  58
          a.   Proposal	58
          b.   Comments	59
          c.   Final Guidance	59
     7.   Wet Weather Exclusion	59
          a.   Proposal	59
          b.   Comments	60
          c.   Final Guidance	60
     8.   Bioaccumulative Chemicals of Concern  	  61
          a.   Proposal	61
          b.   Comments	62
          c.   Final Guidance	67
     9.   Potential Bioaccumulative Chemicals of Concern  .  .  68
          a.   Proposal	68
          b.   Comments	68
          c.   Final Guidance	69
     10.  Pollutants of Initial Focus 	  69
          a.   Proposal	69
          b.   Comments	70
          c.   Final Guidance	71

D.   Procedures for Adoption and EPA Review	71
     1.   Adoption Procedures 	  71

                               iii

-------
          a.   Proposal	71
          b.   Comments ...."'	71
          c.   Final Guidance	.  .  73
     2.   Interpretation of "Consistent With" 	  74
          a.   Proposal	74
          b.   Comments	,	>.  74
          c.   Final Guidance	77
     3.   Indian Tribes	•	79
          a.   Proposal	79
          b.   Comments	79
          c.   Final Guidance	  80

E.   Amendments to NPDES and Water Quality Standards Program
     Regulations	80
     1.   Proposal	80
     2.   Final Guidance	80

F.   Precedential Effects of Elements of the Guidance ....  81
     1.   Proposal	81
     2.   Comments	81
     3.   Final Guidance	82

G.   Implementation of Endangered Species Act	82
     1.   Proposal	82
     2.   Consultation with the U.S. Fish and Wildlife
          Service	83
          a.   Protection of Endangered Mussel Species   ...  83
          b.   Wildlife Criteria Methodology	85
          c.   Other Issues	86
     3.   Comments	88
     4.   Final Guidance	88

III.  AQUATIC LIFE	93

A.   Summary of Final Rule	93

B.   Final Tier I Methodology	93
     1.   Required Data	94
          a.   Proposal	94
          b.   Comments	94
          c.   Final Guidance	94
     2.   Commercially and Recreationally Important Species  .  95
          a.   Proposal	95
          b.   Comments	95
          c.   Final Guidance	  .  95
     3.   Ecologically Important Species  	  95
          a.   Proposal	95
          b.   Comments	95
          c.   Final Guidance	96
     4.   Elimination of Final Residue Value	  .  96
          a.   Proposal	96
          b.   Comments	96
          c.   Final Guidance	96
     5.   Acute-Chronic Ratios	96

                                iv

-------
          a.   Proposal	96
          b.   Comments	.  .  .  .	97
          c.   Final Guidance	97
     6.   Bioavailability	97
          a.   Proposal	97
          b.   Comments	98
          el1   Final Guidance	  102
     7.   Averaging Period/Frequency of Exceedance   	  102
          a.   Proposal	102
          b.   Comments	102
          c.   Final Guidance	103
     8.   Final Tier I Criteria	103
          a.   Proposal	103
          b.   Comments	••	104
          c.   Final Guidance	104
     9.   Tier I Criteria	109
          a.   Proposal	  .  109
          b.   Comments:   	109
          c.   Final Guidance	109
     10.  Potential Changes to National Guidelines   	  110
          a.   Proposal	110
          b.   .Comments	Ill
          c.   Final Guidance	Ill

C.   Final Tier II Methodology	Ill
     1.   Requirement for use in Interpreting the Narrative
          Toxics Criterion  	  Ill
          a.   Proposal	ill
          b.   Comments	112
          c.   Final Guidance	112
     2.   Data Requirements	113
          a.   Proposal	113
          b.   Comments	113
          c.   Final Guidance	113
     3.   Other Methods  for Tier II Values	113
          a.   Proposal	113
          b.   Comments	113
          c.   Final Guidance	114
     4.   Adjustment Factors	114
          a.   Proposal	114
          b.   Comments	114
          c.   Final Guidance	114
     5.   Assumed ACRs	117
          a.   Proposal	117
          b.   Comments	117
          c.   Final Guidance	117

D.   Comparison with the CWA and EPA's National Guidance   .  .  117

E.   Conformance* with the Great Lakes Water Quality
     Agreement	118
     1.   Tier I Aquatic Life Criteria and Methodology   .  .  .  118
     2.   Tier II Values and Methodology	119

-------
IV.  BIOACCUMULATION FACTORS	121

A.   Summary of Final Rule	-121

B.   Explanation of Final Provisions	  . 121
     1.   BAFs  •	121
     2.   Measured and Predicted BAFs	'.  . 122
          a.   Hierarchy of Methods	122
               i.   Proposal:	122
               ii.  Comments	123
               iii. Final Guidance	127
          b.   Field-Measured BAFs	127
               i.   Proposal	127
               ii.  Comments	127
               iii. Final Guidance	128
          c.   Field-Measured BSAFs	128
               i.   Proposal	128
               ii.  Comments	128
               iii. Final Guidance	129
          d.   Measured and Predicted BCFs	129
               i.   Proposal	129
               ii.  Comments	130
               iii. Final Guidance	131
          e.   Inorganic Chemicals	131
               i.   Proposal	131
               ii.  Comments	131
               iii. Final Guidance	131
     3.   Lipid Values	132
          a.   Lipid Value for Human Health BAFs	132
               i.   Proposal	132
               ii.  Comments	132
               i4.i. Final Guidance	133
          b.   Lipid Value for Wildlife BAFs	133
               i.   Proposal	133
               ii.  Comments	134
               iii. Final Guidance	134
     4.   FCMs	134
          a.   Proposal	134
          b.   Comments	134
          c.   Final Guidance	136
     5.   Accounting for the Effect of Metabolism  in
          Predicted BAFs	136
          a.   Proposal	136
          b.   Comments	137
          c.   Final Guidance	137
     6.   Bioavailability	137
          a.   Proposal	137
          b.   Comments	138
          c.   Final Guidance	139
     7.   Calculation of Baseline BAFs	139
          a.   Proposal	139
          b.   Final Guidance	139
     8.   Other Uses of BAFs	141
     9.   Individual BAFs	141

                                vi

-------
          a.   2,3,7,8-TCDD	141
               i.   Proposal   . .'.	141
               ii.  Comments	141
               iii. Final Guidance	141
          b.   PCBs	142
               i".   Proposal	•  142
               ii.  Comments	142
               iii. Final Guidance	•	143
          c.   Mercury	143
               i.   Proposal	143
               ii.  Comments	143
               iii. Final Guidance	144

C.   Conformance to the  CWA, Great Lakes Water Quality
     Agreement and Great Lakes Critical Programs Act  of
     1990	144

D.   Adoption of Water Quality Standards Consistent with  the
     Final Guidance	144

E.   References	145

V.  HUMAN HEALTH	147

A.   Summary of Final Rule	147

B.   Explanation of Final Provisions   	  147

C.   Criteria Methodologies  	  148
     1.   Endpoints Addressed	148
          a.   Proposal	148
          b.   Comments	148
          c.   Final Guidance	148
     2.   Mechanism of Action	148
          a.   Cancer	148
               i.   Proposal	148
               ii.  Comments	149
               iii. Final Guidance	149
          b.   Noncancer	150
               i.   Proposal	150
               id.  Comments	150
               iii. Final Guidance	150
     3.   Choice of Risk Level	151
          a.   Proposal	151
          b.   Comments	151
          c.   Final Guidance	151
     4.   Acceptable Dose	153
          a.   RAD	153
               i.   Biologically Relevant versus Sensitive
                    Species	153
                    (A) . Proposal	153
                    (B) . Comments  .  .•	153
                    (C) . Final Guidance	153
               ii.  Less than Lifetime Adjustment Factor.  .  .  154

                               vii

-------
               (A) .  Proposal	154
               (B) .  Comments	154
               (C).  Final Guidance 	  .....  154
          iii. Species Scaling Factor   	  154
               (A) .  Proposal	154
               (B).  Comments 	  .....  154
               (C) .  Final Guidance	  155
     b.   ADE	155
          i.   List of Deleterious Effects	155
               (A) .  Proposal	155
               (B) .  Comments	155
               (C) .  Final Guidance	155
          ii.  Uncertainty Factors	156
               (A) .  Proposal	156
               (B) .  Comments	  156
               (C) .  Final Guidance	156
     c.   IRIS	157
          i.   Proposal	157
          ii.  Comments	157
          iii. Final Guidance	157
5.   Exposure Assumptions	  .  158
     a.   Body Weight	158
          i.   Proposal	158
          ii.  Comments	158
          iii. Final Guidance	158
     b.   Duration of Exposure	158
          i.   Proposal	158
          ii.  Comments	159
          iii. Final Guidance	159
     c.   Incidental Exposure  	  159
          i.   Proposal	  .  159
          ii.  Comments	  159
          ij.i. Final Guidance	160
     d.   Drinking Water Consumption	160
          i.   Proposal	160
          ii.  Comments	161
          iii. Final Guidance	161
     e.   Fish Consumption	161
          i.   Proposal	161
          ii.  Comments	162
          iii. Final Guidance:	165
     f.   BAFs	165
     g.   Relative Source Contribution	166
          i.   Proposal	 .  .  166
          ii.  Comments	166
          iii. Final Guidance	167
6.   Minimum Data Requirements/Tier I and Tier II   ...  167
     a.   Carcinogens	168
          i.   Proposal	168
          ii.  Comments	168
          i-ii. Final Guidance	169
     b.   Non-carcinogens	170
          i.   Proposal	170
          ii.  Comments	170

                          viii

-------
               iii. Final Guidance	171

D.   Criteria Derivation	171
     1.   Proposed Criteria and Values   	  172
          a.   PCBs  (Human Cancer Value)	173
               i.   Proposal	173
               ii.  Comments	173
               iii. Final Guidance:	174
          b.   TCDD	175
               i.   TCDD - Noncancer Criterion	175
                     (A) . Proposal	175
                     (B) . Comments	175
                     (C) . Final Guidance	175
               ii.  TCDD - Cancer Criterion  .........  175
                     (A)  Proposal	175
                     (B) . Comments	175
                     (C) . Final Guidance	176
          c.   Mercury	176
               i.   Proposal	176
               ii.  Comments	176
               iii. Final Guidance	176

E.   Relationship of the Great Lakes Initiative Guidelines
     to National Guidelines Revisions  	  176
     l.   Proposal	176
     2.   Comments	177
     3.   Final Guidance	177

F.   Comparison with the CWA and Great Lakes Water Quality
     Agreement	177
     1.   Tier I Criteria/Methodology	177
          a.   Comparison with the CWA	177
          b.   Conformance with the GLWQA	178
     2.   Tier IJ Criteria/Methodology	178
          a.   Comparison with the CWA	178
          b.   Conformance with the GLWQA	178

G.   References	178

VI.  WILDLIFE	181

A.   Introduction	181

B.   Scope of Methodology	182
     1.   Proposal	182
     2.   Discussion of Comments	182
     3.   Final Guidance	185

C.   Effect Component  	  185
     1.   Minimum Data Requirements	185
          a.   Proposal	185
          b.   Comments	186
          c.   Final Guidance	187
     2.   UFs	187

                                ix

-------
          a.   Proposal	187
          b.   Comments	187
          c.   Final Guidance	189

D.   Exposure Component 	 . 	 ..... 190
     1.   Representative Species   	 190
          a.   Proposal .	 190
          b.   Comments	191
          c.   Final Guidance	191
     2.   Exposure Parameters	192
          a.   Proposal	192
          b.   Comments	192
          c.   Final Guidance	194

E.   Protection of Individual Members of a Population .... 194
     1.   Proposal	194
     2.   Comments	194

F.   Wildlife Criteria	195
     1.   Proposal	195
     2.   Discussion	195
          a.   Mercury	195
          b.   DDT	196
          c.   PCBs	197
          d.   2,3,7,8 TCDD	197
     3.   Final Guidance	198

G.   Comparison of Wildlife Criteria and Methods to National
     Program and to Great Lakes Water Quality Agreement  .  .  .198

H.   References	198

VII.  ANTIDEGRADATION	203

A.   General Discussion/Background  	 203
     1.   History of the Great Lakes Antidegradation
          Guidance	203
     2.   Summary of the Proposed Guidance	204
          a.   The Antidegradation Standard 	 204
          b.   Antidegradation Implementation Procedures   .  .205
          c.   Antidegradation Demonstration  	 205
          d.   Antidegradation Decision 	 206

B.   Overview of. the Final Guidance	206

C.   Detailed Discussion of the Final Regulation   	 207
     1.   The Antidegradation Standard  	 207
     2.   Antidegradation Implementation Procedures 	 207
          a.   De Minimis Lowering of Water Quality	207
               i.   Background	207
               ii.  Discussion of  Significant Comments   .  .  . 208
               iii. The Final Guidance	209
          b.   High Quality Waters	210
               i.   Background	210

-------
               ii.  Discussion of Significant Comments   .  .  . 210
               iii. The Final Guidance	212
          c.   Lake Superior Basin - Outstanding
               International Resource Waters  	 212
          d.   Outstanding National Resource Water  	 212
               i.   Background	: 212
               ii.  Discussion of Significant Comments   .  .  . 212
               iii. The Final Guidance  . . .	213
          e.   Significant Lowering of Water Quality  .... 213
               i.   Background	213
               ii.  Discussion of Significant Comments   .  .  . 213
               iii. The Final Guidance	214
          f.   Deleted Definitions  	 214
          g.   Implementation Procedures  	 215
              . i.   Background	215
               ii.  Discussion of Significant Comments   .  .  . 216
               iii. The Final Guidance	218
     3.   Antidegradation Demonstration 	 219
          a.   Background	219
          b.   Discussion of Significant Comments 	 219
          c.   The Final Guidance	221
               i.   Identification of Cost-Effective
                    Pollution Prevention Alternatives to
                    Prevent or Reduce the Significant
                    Lowering of Water Quality 	 221
               ii.  Alternative or Enhanced Treatment to
                    Eliminate the Significant Lowering of
                    Water Quality	222
               iii. Important Social and Economic
                    Development	223
     4.   Antidegradation Decision  	 224
          a.   Background	224
          b.   Discussion of Significant Comments 	 224
          c.   The Final Guidance	225

VIII.  IMPLEMENTATION PROCEDURES	227

A.   Site-Specific Modifications to Criteria and Values  .  .  .227
     1.   General	227
     2.   Aquatic Life	227
          a.   Proposal	227
          b.   Comments	227
          c.   Final Guidance	228
     3.   Wildlife	  . 229
          a.   Site-specific Modifications  	 229
               i.   Proposal	229
               ii.  Comments	229
                    (A). Less Stringent Site-specific
                         Modifications  	 229
                    (B). More Stringent Site-specific
                         Modifications  	 230
               iii. Final Guidance	230
          b.   Protection of Endangered or Threatened
               Species	230

                               xi

-------
               i.   Proposal	230
               ii.  Comments  .	230
               iii. Final Guidance	 231
     4.   Bioaccumulation Factors 	 231
          a.   Proposal .	231
          b.   Comments	232
          c:   Final Guidance	233
     5.   Human Health	233
          a.   Proposal	233
          b.   Comments	233
               i.   Fish Consumption Rate	234
               i-i.  Percent Fish Lipid	234
               iii. BAF	235
          c.   Final Guidance	•	235

B.   Variances from Water Quality.Standards for Point
     Sources	 237
     1.   Applicability	237
          a.   Proposal	237
          b.   Discussion of Significant Comments  	 238
          c.   Final Guidance	239
     2.   Maximum Timeframe for Variances	239
          a.   Proposal	239
          b.   Discussion of Significant Comments  	 239
          c.   Final Guidance	239
     3.   Conditions to Grant a Variance	240
          a.   Proposal	240
          b.   Discussion of Significant Comments  	 240
          c.   Final Guidance	241
     4.   Submittal of Variance Application 	 242
          a.   Proposal	242
          b.   Discussion of Significant Comments  	 242
          c.   Final Guidance	242
     5.   Public Notice	242
          a.   Proposal	242
          b.   Discussion of Significant Comments  	 242
          c.   Final Guidance	243
     6.   Final Decision on Variance Request  	 243
          a.   Proposal	243
          b.   Discussion of Significant Comments  	 243
          c.   Final Guidance	244
     7.   Incorporating Variances into NPDES Permits   .... 244
          a.   Proposal	244
          b.   Discussion of Significant Comments  	 244
          c.   Final Guidance	244
     8.   Renewal of Variances	245
          a.   Proposal	245
          b.   Discussion of Significant Comments  	 245
          c.   Final Guidance	245
     9.   EPA Approval	245
          a.   Proposal	 245
          b.   Discussion of Significant Comments  	 246
          c.   Final Guidance	246
     10.  State or Tribal Water Quality Standards  Revisions  . 246

                               xii

-------
          a.   Proposal	246
          b.   Discussion of Significant Comments 	 246
          c.   Final Guidance	-246

C.   Total Maximum Daily Loads	247
     1.   Background	.  . 247
     2.   Overview of Proposed Procedures 3A and 3B . .  .  .''  . 249
          a..   Proposal	249
          b.   Comments	249
          c.   Final Guidance	250
     3.   General Conditions of Application 	 251
          a.   General Condition 1 - TMDLs Required 	 252
               i..   Proposal	252
               ii.  Comments	252
               iii. Final Guidance	252
          b.   General Condition 2 - Attainment of Water
               Quality Standards  	 253
               i.   Proposal	253
               ii.  Final Guidance	253
          c.   General Condition 3 - TMDL Allocations .... 254
               i.   Proposal	254
               id.  Comments	254
               iii. Final Guidance	254
          d.   General Condition 4 - WLA Values	255
               i.   Proposal	255
               ii.  Final Guidance	256
          e.   General Condition 5 - Margin of Safety .... 256
               i.   Proposal	256
               ii.  Comments	256
               iii. Final Guidance	256
          f.   General Condition 6 - More Stringent
               Requirements	257
          g.   General Condition 7 - Accumulation in
               Sediments	257
               i.   Proposal	257
               ii.  Comments	257
               iii. Final Guidance	258
          h.   General Condition 8 - Wet Weather Events  .  .  . 259
               i.   Proposal	259
               il.  Comments	259
               iii. Final Guidance	259
          i.   General Condition 9 - Background
               Concentrations of Pollutants 	 260
               i.   Choice of Data Set	260
                    (A)   Proposal	260
                    (B)   Comments	260
                    (C)   Final Guidance	261
               ii.  Geometric Mean	262
                    (A) . Proposal	262
                    (B) . Final Guidance	262
               iii. Data Points Above and Below Detection  .  . 263
                    (A)   Proposal	263
                    (B)   Comments	263
                    (C)   Final Guidance	264

                              xiii

-------
     j .    General Condition 10 - Effluent Flow	265
     k.    General Condition 11 - Reserved Allocations   . 266
          i.   Proposal	266
          ii.  Comments	266
          iii. Final Guidance	266
4.   Special Provisions for BCCs	266
     a.    Proposal	266
     b.    Comments	266
     c.    Final Guidance	268
5.   TMDLs for Open Waters of the Great Lakes	273
     a.    Mixing Zones for non-BCCs	'  . 273
          i.   Proposal	273
          ii.  Comments	274
          iii. Final Guidance	274
     b.    Calculating Load Allocations	  . 275
     c.    Protection from Acute Effects  	 275
          i.   Proposal	275
          ii.  Comments	275
          iii. Final Guidance	276
6.   TMDLs for Discharges to Tributaries 	 276
     a.    Steady State	277
          i\   Proposal	277
          ii.  Comments	277
          iii. Final Guidance	277
     b.    Stream Design Flows	277
          i.   Proposal	277
          ii.  Comments	278
          iii. Final Guidance	278
          iv.  Wildlife	279
               (A)  Proposal	279
               (B)  Comments	279
               (C)  Final Guidance	279
          v.   Chronic Aquatic Life	281
               (A)  Proposal	281
               (B)  Comments	281
               (C)  Final Guidance	281
          vi.  Acute Aquatic Life	281
               (A)  Proposal	281
               (B)  Comments	281
               (C)  Final Guidance	281
          vii. Human Health	282
               (A)  Proposal	282
               (B)  Comments	282
               (C)  Final Guidance	282
     c.    Mixing Zones for Non-BCCs	283
          i.   Proposal	283
          ii.  Comments	283
          iii. Final Guidance	284
7.   Procedures for High Background Concentrations  .  .  . 285
     a.    Proposal	285
     b.    Comments	285
     c.    Final Guidance .  . .  :	285
8.   Pollutant Degradation  	 286
     a.    Proposal	286

                          xiv

-------
          b.   Comments	286
          c.   Final Guidance .  .	286
     9.   Mixing Zone Studies	287
          a.   Proposal	287
          b.   Comments	287
          c.   Final Guidance	287
     10.  Pollution Trading Opportunities 	 288

D.   Additivity	289
     1.   Background	289
     2.   Additivity Considerations in Other EPA Programs .  . 290
     3.   Existing State Water Quality Standards  	 290
     4.   Proposed Guidance Overview  	 291
     5.   Aquatic Life	•	291
     6.   Human Health	291
          a.   Carcinogens	291
               i-.    Proposal	291
               ii.  Discussion of Significant Comments on
                    the Assumption of Additivity for
                    Carcinogens	292
               iii. Discussion of Significant Comments on
                    Application of Additivity Provision for
                    Carcinogens	293
               iv.  Final Guidance	295
          b.   Non-carcinogens	296
               i'.    Proposal	296
               ii.  Discussion of Significant Comments  . .  . 297
               iii. Final Guidance	298
     7.   Toxicity Equivalency Factors/Bioaccumulation
          Equivalency Factors 	 298
          a.   Proposal	298
          b.   Discussion of Significant Comments 	 299
          c.   Response for Application of TEFs/BEFs to
               Human Health	299
          d.   Response to Application of TEFs to Wildlife   . 300
          e.   Final Guidance:  	301
     8.   Wildlife	301
          a.   Proposal	301
          b.   Comments	302
          c.   Final Guidance:  	302
     9.   References	304

E.   Reasonable Potential for Exceeding Numeric Water
     Quality Standards	 .  . 307
     1.   Existing National Rules and Guidance  	 307
     2.   General Requirements of Procedure 5 	 309
          a.   Developing Preliminary Wasteload Allocations  . 311
               i.    Proposal	311
               ii.  Comments	311
               iii. Final Guidance	311
          b.   Developing Preliminary Effluent Limitations   . 312
               i,.    Proposal  .  .  .	312
               ii.  Comments	313
               iii. Final Guidance	315

                               xv

-------
     c.    Determining Reasonable Potential to Exceed
          the Preliminary Effluent Limitations Using
          Pollutant Concentration Data 	 315
          i.    Proposal	.316
          ii.  Comments	316
          iii.  Final Guidance:	318
     d.    Determining Reasonable Potential Using
          Pollutant Concentration Data Where the
          Effluent Flow Rate is Equal to or Greater
          than the 7Q10	324
          i.    Proposal:	324
          ii.  Comments	325
          iii.  Final Guidance	325
     e.    D.etermining Reasonable Potential in the
          Absence of Facility Specific Effluent
          Monitoring Data	.326
          i.    Proposal	326
          ii.  Comments	326
          iii.  Final Guidance	327
     f.    Determining Reasonable Potential for
          Pollutants When Tier II Values are Not
          Available	327
          i..    Proposal	327
          ii.  Comments	329
          iii.  Final Guidance	331
     g.    Determining Reasonable Potential Using Fish
          Tissue Data	333
          i.    Proposal	333
          ii.  Comments	333
          iii.  Final Guidance	334
     h.    Basis for Effluent Limitations	336
3.   Consideration of Pollutants in Intake Water .... 339
     a.    Introduction	339
     b.    Existing Mechanisms	340
     c.    Summary of Proposal	341
          i.    Proposed Guidance 	 341
          ii.  Other Options	 343
               (A)  Option 1	343
               (B)  Option 2	343
               (C)  Option 3	344
               (D)  Option 4	344
4.   Summary of Intake Pollutant Considerations in
     Final Guidance and Overall Rationale	345
     a.    Intake Pollutant Reasonable Potential
          Procedure	345
     b.    No Net Addition Interim Approach for Setting
          WQBELS	346
     c.    Consideration of Intake Pollutants from a
          Different Body of Water	349
     d.    Combined Approach for Multiple Intake
          Sources   	  ..... 352
5.   Legal Authority	353
                          xvi

-------
          a.   Discharge of Intake Pollutants is an Addition
               of Pollutants Under-the CWA	353
          b.   EPA's Authority and Rationale for
               Establishing Interim Permitting Procedures
               Allowing "No Net Addition" Limitations for
               Intake Water Pollutants  	 359
     6.   Continued Availability of Existing Mechanisms . .   . 362
     7.   Final Intake Credit Provision and Response to
          Major Comments	365
          a.   General Issues	365
               i.   Pollutant-by-Pollutant, Outfall-by-
                    Outfall Analysis	365
               ii.  Pollutants Covered  	 365
               iii. Required Demonstration  	 367
               iv.  Definition of Same Body of Water  .... 368
                    (A). General	369
                    (B). Intermediate Use/Public Water
                         Supply	372
                    (C) . Ground water	374
               v.   100 Percent Requirement	375
               vi.  No Increase in Concentration
                    Requirement	376
               vii. Chemical/Physical Alteration
                    Requirement	379
               viii. Timing/Location Requirement  	 380
               ix.  Relationship Between this Section and
                    Other Reasonable Potential Sections . .   .381
          b.   Issues Specific to the Intake Pollutant
               Reasonable Potential Procedure 	 382
               i..   No Mass Added Requirement	382
               ii.  Other Requirements	386
                    (A) . Documentation	386
                    (B) . Monitoring	387
                    (C) . Reopener	388
          c.   Issues Specific to Consideration of Intake
               Pollutants in the Derivation of WQBELs .... 388
               i.   Availability Only for Non-Attainment
                    Waters	388
               ii.  Partial Credit at Discretion of
                    Permitting Authority  	 390
               iii. Developing the Limit	392
                    (B). Expression of a Numeric Limit  . .   . 392

F.   Whole Effluent Toxicity  	 395
     1.   Background	395
     2.   Criteria for WET	396
          a.   Proposal	396
          b.   Comments	396
          c.   Final Guidance	397
     3.   Acute Toxicity Control	397
          a.   Proposal	397
          b.   Comments	397
          c.   Final Guidance	398
     4.   Chronic Toxicity Control  	 398

                              xvii

-------
          a.   Proposal	398
          b.   Comments	399
          c.   Final Guidance	 399
     5.   Numeric and Narrative Criteria	.  . 400
          a.   Proposal	400
          b.   Final Guidance	: 400
     6.   WET Test Methods	400
          a.   Proposal	•	400
          b.   Comments	400
          c.   Final Guidance	401
     7.   Permit Conditions 	 401
          a.   Data Indicates Reasonable Potential	401
               i.   Proposal	401
               ii.  Final Guidance	401
          b.   Insufficient Data to Determine Reasonable
               Potential	402
               i.   Proposal	402
               ii.  Comments	402
               iii. Final Guidance	403
          c.   Data Indicates No Reasonable Potential .... 403
     8.   Reasonable Potential Determinations 	 404
          a.   Proposal	404
               i.   Characterization of the Discharge .... 404
               ii.  Specific Acute WET Procedure  	 404
               iii. Specific Chronic WET Procedure  	 404
          b.   Comments	404
          c.   Final Guidance	406
               i.   Reasonable Potential Equation for Acute
                    WET	407
               id.  Reasonable Potential Equation for
                    Chronic WET	407

G.   Loading Limits	411
     1.   Proposal	411
     2.   Comments	411
          a.   Need for Loading Limits	411
          b.   Wet Weather Discharges	412
     3.   Final Guidance	412

H.   WQBELs Below the Level of Quantification 	 415
     1.   Expressing a WQBEL Below the Minimum
          Quantification Level  	 415
          a.   Proposal	415
          b.   Comments	415
          c.   Final Guidance	415
     2.   Compliance Issues 	 417
          a.   Proposal	417
          b.   Comments	417
          c.   Final Guidance	418
     3.   Compliance with the CEL	421
          a.   Proposal	421
          b.   Comments	:	422
          c.   Final Guidance	422
     4.   Pollution Minimization Program	422

                              xviii

-------
          a.   Proposal	422
          b.   Comments	423
          c.   Final Guidance	423
     5.   BCC Requirements	427
          a.   Proposal	427
          b.   Comments	427
          c:-   Final Guidance	427

I.   Compliance Schedules 	 429
     1.   Proposal	429
     2.   Final Guidance	429
          a.   Eligibility	429
          b.   Duration of Compliance Schedules . 	 431

IX.  EXECUTIVE ORDER 12866  	 435

A.   Introduction and Rationale for Estimating Costs and
     Benefits for the Great Lakes Water Quality Guidance  .  . 435

B.   Summary of Proposal	436
     1.   Costs	436
          a.   Method	436
          b.   Results	440
     2.   Cost-Effectiveness  	 441
          a.   Method	441
          b.   Results	442

C.   Revisions to Projected Costs	442
     1.   Summary of Revised Approach to Estimate Costs and
          Pollutant Load Reductions 	 442
          a.   Revised Criteria and Implementation
               Procedures	445
               i.   Dissolved Metals Water Quality Criteria  . 445
               ii.  Intake Credits	445
               iii. Additivity/TEFs	446
               iv.  Acilte Mixing Zones for WET	447
          b.   Criteria for Tier I and Tier II Pollutants .  . 447
          c.   Data for the Sample Facilities	450
          d.   Compliance Cost Decision Matrix  	 451
          e.   Pollution Prevention/Waste Minimization
               Costs	452
          f.   Indirect Dischargers 	 452
          g.   Revised Toxic Weights  	 453
     2.   Results	453
          a.   Estimated Compliance Costs 	 454
               i.   EPA's Low-End Estimate  	 454
               ii.  EPA's High-End Estimate 	 456
               iii. Comparison of Estimated Costs for the
                    Final Guidance to Costs of Proposed
                    Guidance	456
          b.   Estimated Pollutant Reductions 	 458

D-.   Major Is sues/Comments and Responses Related to
     Estimated Costs  	 462

                               xix

-------
          1.   Use of Pollution Prevention/Waste
               Minimization to Abhieve Guidance-Based
               Limits	.	-462
               a.   Comments	462
               b.   Response	463
          2.   Future Impact of Detection Levels	 463
               a.   Comments	.'  . 463
               b..   Response	464
          3.   Intake Credits	465
               a.   Comments	465
               b.   Response	465
          4.   Tier II Criteria	466
               a.   Comments	, .  . 466
               b.   Response	466
          5.   Wildlife Criteria/Mercury Criteria 	 466
               a.   Comments	466
               b.   Response	 . , .  . 466
          6.   Elimination of Mixing Zones for BCCs	467
               a.   Comments	467
               b.   Response	467
          7.   Antidegradation	468
               a.   Comments	468
               b.   Response	468
          8.   Additivity	469
               a.   Comments	469
               b-.   Response	469
          9.   Other Study Costs  	 470
               a.   Comments	470
               b.   Response	470

E.   Benefits	471
     1.   Summary of Proposal	471
          a.   Introduction	471
          b.   Qualitative Assessment of Benefits	471
          c.   Quantitative Assessment of Benefits Analysis  . 472
               i.   Economic Concepts Applicable to the
                    Benefits Analysis 	 472
               ii.  Benefits Methodologies   	 473
               iii. Results	474
     2.   Major Issues/Comments and Responses 	 477
          a.   Attribution of Benefits to the Guidance  , .  . 477
               i.   Comments	477
               ii.  Response	478
          b.   Risk Assessment	478
               i.   Comments	478
               ii.  Response	 478
          c.   Valuation Approaches Used for the Proposed
               Benefits Analysis  	 479
               i.   Comments	479
               ii.  Response	479
          d.   Limitations of the Case Study Approach and
               Specific Case Study Critiques  	 480
               i.   Comments	480
               ii.  Response	480

                                xx

-------
          e.   Economic Impacts of the Guidance	481
               i'.   Comments  . .".	481
               ii.  Response	482
     3.   Revised Benefits Estimates  	 482
          a.   Introduction	482
          b.   Updated Risk Assessments for Great Lakes
               Anglers	484
               i.   Sport Angler Risk Assessment  	 484
               ii.  Native American Risk Assessment	487
          c.   Revised Case Study Benefit-Cost Analyses for
               the Final Guidance	489
               i.   Fox River and Green Bay Case Study  .  .  . 489
               ii.  Saginaw River/Saginaw Bay Case Study   .  . 489
               iii. Black River Case Study	489
               iv.  Comparison of Benefits and Costs for the
                    Case Studies	490
               v.   Case Study Representativeness 	 490

X.  REGULATORY FLEXIBILITY ACT  	 493

XI.  ENHANCING THE INTERGOVERNMENTAL PARTNERSHIP UNDER
     EXECUTIVE ORDER 12875  	 495

XII.  PAPERWORK REDUCTION ACT	497
                               xxi

-------
                                Section I: Background
                         I.   BACKGROUND
A.    Description of Resource

      The Great Lakes are a unique natural resource that have played a vital
role  in the history and development of the United States and Canada.  The
Great Lakes consist of Lakes Superior, Huron, Michigan, Erie and Ontario and
their connecting channels  (i.e., the Saint Mary's River, Saint Clair River,
Detroit River, Niagara River and the Saint Lawrence River to the Canadian
Border).  The Great'Lakes plus all of the streams, rivers, lakes and other
bodies of water that are within the drainage basin of the Lakes collectively
comprise the Great Lakes System.

      The Great Lakes span over 750 miles across eight States--New York,
Pennsylvania, Ohio, Michigan, Indiana, Illinois, Wisconsin and Minnesota--and
the Province of Ontario.  The Lakes contain approximately 18 percent of the
world's and 95 percent of the United States' fresh surface water supply.  The
Great Lakes are a source of drinking water and energy, and are used for
recreational, transportation, agricultural and industrial purposes by the more
than  46 million Americans and Canadians who inhabit the Great Lakes region,
including 29 Native American tribes.  Over 1,000 industries and millions of
jobs  are dependent upon water from the Great Lakes.  The Great Lakes System
also  supports hundreds of species of aquatic life, wildlife and plants along
more  than 4,500 miles of coastline which boast six National Parks and
Lakeshores, six National Forests, seven National Wildlife Refuges, and
hundreds of State parks, forests and sanctuaries.  The prominent features of
the Great Lakes can be seen when the Earth is viewed from outer space.

      Although each of the Lakes is an interrelated and interdependent
component of the Great Lakes System, the Lakes vary in size, retention times,
and existing water quality.  Lake Superior, the northernmost Lake, is the
largest, deepest, coldest and most pristine of the Great Lakes, while Lake
Erie  is the smallest, shallowest, and the most susceptible of all of the Great
Lakes to the effects of urban and agricultural activities.  Lake Superior also
has the longest retention time--the average time it takes for one molecule of
water to exit the system--of 173 years, while Lake Erie has the shortest at
2.7 years.  Lake Huron alone acts as a drain for 51,700 of the 201,000 square
miles comprising the Great Lakes basin.  Lake Ontario, the easternmost Lake,
eventually receives all of the outflow from the other Great Lakes.  The waters
of four of the Great Lakes--Superior, Huron, Erie and Ontario--are shared by
the United States and Canada; only Lake Michigan is located wholly within the
United States.

      State and provincial Natural Heritage programs,  which have been
established throughout the Great Lakes basin for the purpose of identifying
elements (i.e., natural ecological communities or individual species)  of
biological diversity requiring protection, have identified 131 elements of
global significance within the Great Lakes basin that are crucial to
maintaining the region's biological diversity (The Nature Conservancy,  1994).
Of these,  31 are natural ecological community types.  The remaining 100
consist of plants,  insects, mollusks, fish, birds, reptiles and one mammal.
All of these species are either critically imperiled,  imperiled or rare on a
world-wide basis.  Nearly half (47 percent) are present either exclusively or
predominantly within the basin,  or are the best examples of their respective
species.  All 131 elements provide some of the most visible examples of the
biological character of the Great Lakes System.   The continued existence of
these species depends upon their ability to survive in the Great Lakes basin.

-------
2      Water Quality Guidance for the Great Lakes System — Supplementary Information Document

B.    Environmental Problems in the Great Lakes System

1.    Historical Nature of Environmental Problems in the Great Lakes System

      The Great Lakes Basin Ecosystem--the interacting components of air,
land, water and living organisms, including humans, that live within the Great
Lakes drainage basin--is a very young,  yet remarkable, ecosystem in geological
terms due to the physical and biological characteristics discussed above.  The
species which inhabit the Great Lakes basin reside in a wide range of
habitats.  Many of the features and processes that provided the environment
for these species to initially migrate to this region, adapt to its physical
characteristics, and evolve into even more unique and diverse species are also
susceptible to the multiple environmental stressors that humans have placed,
and continue to place, upon this important ecosystem.

      The Tribal relationship with the Great Lakes has been well documented
since being noted by the earliest Europeans to come into contact with the
indigenous people of the area.  The Great Lakes System played an integral role
in Tribal sustenance, stories, culture and spirituality to an extent unmatched
by any other group.

      Early settlement and related economic activities drastically changed
portions of the Great Lakes System.  The clearing of land for agricultural
purposes, commercial fishing, and industrialization placed a combination of
physical, biological and chemical stresses upon the Great Lakes Basin
Ecosystem, some of which continue today.

      By the late 1800s, all of the available agricultural lands within the
basin had been cleared and settled.  This harvesting and clearing of the Great
Lakes basin landscape imposed the first significant human stress condition
upon the Great Lakes Basin Ecosystem.  Temperatures in tributary streams rose
and entire forest systems, as well as wetlands and prairies, were frequently
lost as a result of these early logging and agricultural practices.  Also lost
were thousands of acres of critical habitat for wildlife.  The loss of  this
habitat caused many species to migrate to other areas or simply to perish due
to this physical alteration of the Great Lakes Basin Ecosystem  (The Nature
Conservancy, 1994).  In addition  to the above factors, dams and watercourse
modifications, human disturbance  of breeding and nesting locations, soil
erosion and the silting of spawning areas, primarily due to urbanization and
shoreline development pressures,  were responsible for reductions in suitable
fish, avian and wildlife habitat.

      The construction of  several canals from the beginning of the 1800s and
ending with the construction of the St. Lawrence Seaway in the 1950s
connecting the various Great Lakes, several  of their waterways and the  St.
Lawrence River during this same time period also impacted the Great Lakes
System.  Although these canals provided a valuable form of transportation for
the industries and populations located within and around the Great Lakes
basin, they also provided  a mechanism for the unintentional introduction of
exotic  (i.e., non-indigenous) species such  as the alewife  (Alosa
psuedoharengus), the  sea lamprey  (Petromyzon marinus), the Euroasian river
ruffe  (Gymnocephalus  cernuus), and more recently, the zebra mussel  (Dreissena
polymorpha).  Once these exotic  species entered the Great Lakes System, they
altered the natural biotic system by not only competing with native  species
for sustenance, but also by  serving as those species' prey.

        Great Lakes' fish were an important  resource valued by both  the  early
Native American  and European settlers.  Commercial fishing flourished
throughout  the mid- to  late-1800s.  Near the end of the century, however,  the
commercial  fishing industry  experienced a multitude of  stresses from the
introduction of  exotic  species,  overfishing, loss  of  habitat, and pollution.
This combination of physical, biological and chemical stressors eventually
impacted  the commercial fishing  industry in the Great Lakes beginning  in the
late nineteenth  century.

-------
                                Section I: Background
      Despite the amount of general social and economic changes, wild rice
harvesting, hunting, gathering, inland'shore and landlocked fishing continue
to be viable economic and cultural pursuits for modern Great Lakes Native
American communities.  There are presently more than 100 Tribal commercial
fishermen operating-on the Great Lakes.  In 1992, the Tribal commercial fish
harvest on U.S. waters of Lake Superior was more than 600 tons.
                  •
      Additionally, traditional wild rice harvesting of landlocked, flowing,
and inland Lake Superior shore waters is an economic, subsistence and cultural
mainstay of the annual cycle of life for Native American  (and non-Native
American) people in Minnesota, Wisconsin and Michigan.  Hunting of waterfowl
and small and large mammals is also very important to Tribal peoples for
subsistence and cultural reasons.

        The wastes that resulted from industrial and other activities that
prevailed in the basin during the late 1800s and early to mid-1900s placed
additional stresses upon the Great Lakes.  Iron and steel manufacturing, salt
brine production, copper and iron mining, paper making, power generation and
chemical manufacturing facilities utilized the natural resources and waterways
of the Great Lakes in their industrial and transportation processes.  Metals,
organic compounds and other chemical substances were introduced into the Great
Lakes System as a result of this industrialization.  In addition, the growth
of the human population in the Great Lakes basin and the corresponding
disposal of domestic wastes also contributed to the overall degradation of
Great Lakes water quality.

      By the mid-1900s, both the fishing quality and the water quality in the
Great Lakes basin were succumbing to these physical, biological and chemical
stresses.  For example, increased loadings of nutrients to the Great Lakes had
dramatically stimulated the growth of green plants and algae.  When these
materials decomposed, the Lakes experienced decreased levels of dissolved
oxygen in bottom waters, which in turn, caused the displacement of certain
species of insects and fish from the affected areas of the Great Lakes Basin
Ecosystem.  Environmental managers determined that a lakewide approach to
limit the loadings of phosphorous was the key to controlling the excessive
algal growth in the Great Lakes.

      The presence of environmentally persistent, bioaccumulative, chlorinated
organic contaminants was observed and identified as a serious environmental
threat in the 1960s  (Tanabe, 1988; Stevens et al., 1989).  Beginning in 1963,
adverse environmental impacts in the form of poor reproductive success and
high levels of DDT/DDE were observed in herring gulls in Lake Michigan
(Environment Canada, 1991) .  Expanded pollution control programs were
instituted in an attempt to stem and reverse the adverse effects posed by the
presence of these chemicals.  Although individual and concerted actions taken
by citizens, industries, governments and private organizations helped to
reduce the threats posed by these stressors on the Great Lakes Basin
Ecosystem, human activities still adversely impact this important natural
resource.

2.     Current Trends of Environmental Problems in the Great Lakes System

      a.    Impacts From Chemical Contamination.  In spite of the fact that
the Great Lakes contain 5,500 cubic miles of water that cover a total surface
area of 94,000 square miles, the Lakes are sensitive to the effects of a wide
range of pollutants from all sources, both point and nonpoint.  In particular,
the Great Lakes are susceptible to relatively non-degradable, bioaccumulative
chemicals because of the Lakes' unique physical, chemical and biological
characteristics.  Examples of these characteristics include:  (1) long
hydraulic retention times (i.e., relatively closed systems);  (2) low
biological productivity;  (3) low suspended solids concentrations;  (4) great
depth; and  (5) the presence of self-contained fish and wildlife populations
dependent on the Great Lakes System for water, habitat and food.  Taken
together, these characteristics result in such pollutants remaining in the

-------
4      Water Quality Guidance for the Great Lakes System — Supplementary Information Document

system for long periods of time and bioaccumulating in fish and  wildlife
tissue at concentrations which are orders of magnitude above the ambient
concentrations in the water column.  Once these pollutants are released into
the Great Lakes System, they will cycle within the System for decades,
exerting biological-effects and presenting relatively high levels of  risk  to
aquatic life, wildlife and humans which inhabit the Great Lakes  basin.

      The internal responses and processes that operate in the Great  Lakes
because of their depth and long hydraulic residence times mean that pollutants
recycle between biota, sediments and the water column.  As a result,  the Great
Lakes may still be experiencing the effects of historic discharges.   In
certain circumstances, for example, water column concentrations  may be  more
affected by the total mass of pollutants in the Lakes rather than the loadings
from point sources.  Additional discussion of the characteristics and
processes affecting contaminant levels in the Great Lakes is provided in the
preamble to the proposed Guidance  (58 PR at 20807-20817) .

      Scientists have detected 362 contaminants in the Great Lakes System.  Of
these, approximately one third have toxicological data showing that they can
have acute or chronic toxic effects on aquatic life, wildlife and/or  human
health (IJC, 1991).  Chemicals that have been found to bioaccumulate  at levels
of concern in the Great Lakes include, but are not limited to, PCBs,  mercury,
DDT, dioxin, chlordane, and mirex.  The main route of exposure to these
chemicals for humans is through the consumption of Great Lakes fish  (Colborn
et al., 1990).

      Potential adverse human health effects resulting from the  consumption of
contaminated fish include both the increased risk of cancer and  the potential
for systemic or noncancer risks such as kidney damage.  EPA has  calculated
health risks to populations in the Great Lakes System from consumption  of
contaminated fish based on exposure to eight bioaccumulative pollutants:
chlordane, DDT, dieldrin, hexachlorobenzene, mercury, PCBs, 2,3,7,8-
tetrachloro-p-dioxin  (TCDD), and toxaphene.  These chemicals were chosen based
on their potential to cause adverse human health effects  (i.e.,  cancer  or
disease)  and the availability of information on fish tissue contaminant
concentrations from the Great Lakes.

      Based on this data, EPA estimates that the lifetime cancer risks  for
Native Americans in the Great Lakes System due to ingestion of contaminated
fish at current concentrations range from 1.8 times 10"3 (1.8 in  one thousand)
for Lake Superior, to 3.7 times 10'2 (3.7 in one hundred) for Lake Michigan.
Estimated risks to low income minority sport anglers range from  2.5 times  10'3
for Lake Superior, to 1.2 times 10'2 for Lake Michigan.  Estimated risks for
other sport anglers range from 9.7 times 10"4 for Lake Superior,  to 4.5  times
10'3 for Lake Michigan (Regulatory Impact Analysis of the Final Great  Lakes
Water Quality Guidance. 1995. EPA  820-B-95-011)  (RIA), Chapter 6).  In
comparison, EPA has- long maintained that one times 10~*  (one in ten thousand)
to 1 times 10"6 (one in one million) is an appropriate range of risk to  protect
human health.

      EPA also estimates a high potential risk of systemic injury to
populations in the Great Lakes basin due to ingestion of fish contaminated
with these pollutants at current concentrations  (RIA, Chapter 6).  The
systemic adverse health effects associated with the assessed contaminants  are
described in appendix E of the RIA.

      While it is npt possible to  state conclusively that low-level exposure
to these chemicals  is or is not associated with adverse human health
reproductive effects, evidence from wildlife studies  (Fox, 1992; Flint  and
Vena, 1991) and epidemiological investigations of occupational exposures  to
these chemicals indicates that they may be able to alter human reproduction.
Epidemiological studies that have  addressed adverse pregnancy outcomes  in
populations in the Great Lakes have shown some potential effects of  concern

-------
                                Section I: Background
 (Swain, 1991; Fein et al., 1984; Smith, 1984), while other studies have  shown
 little or no effects  (Dar et al., 1992).

      Additionally, new research into the ability of several organochlorine
 compounds to mimic hormones has  recently been published  (Environment Canada,
 1991).  Studies have shown that  PCBs  (Korach et al., 1988) and DDE and DDT
 (McLachlan et al., 1987) have been found to have weak estrogenic abilities.
 There has been speculation among researchers that an increased incidence of
 breast cancer may occur among women exposed to organochlorines  (Manz et  al.,
 1991; Falck et al., 1992) .  However, the link between these compounds and
 breast cancer has not been clearly established.

      As stated earlier, adverse effects observed in fish, birds and mammals
 in the Great Lakes have been attributed to contaminants.  These effects
 include death, eggshell thinning, reduced hatching  success, abnormal behavior,
 deformities  (such as crossed beaks and club feet) and population declines.
 For several species of birds (e.g., double=crested  cormorants, Caspian terns,
 Forster's terns, common terns, ring-billed gulls, herring gulls) in the  Great
 Lakes, a suite of effects including embryo mortality, edema, growth
 retardation, and deformities have been reported and considered similar to
 chick edema syndrome associated  with some classes of PCBs  (Gilbertson et al.,
 1991; Fox et al., 1991; Tillitt  et al., 1992; Environment Canada, 1991;  Mora
 et al., 1993; Kurita and Ludwig, 1988; Yamashita et al., 1993).  More subtle
 effects, such as abnormalities in the thyroid, endocrine and liver systems,
 have also been observed in organisms within these populations  (Gilbertson et
 al., 1991; Colborn, 1991) .  Table 1-1 below summarizes the fish and wildlife
 species thought to be affected by contaminants in the Great Lakes and is
 available in the docket for this rulemaking.

      Of the eleven wildlife species which showed evidence of contaminant
 impacts in the past, three of these species (bald eagles, cormorants and
 herring gulls) are providing some evidence of recent improvements  (Giesy et
 al., 1994).  Current concentrations of organochlorine compounds in Great Lakes
 fish and fish-eating birds are less than during the 1960s and 1970s  (Giesy et
 al., 1994).  Although these trends for these species are encouraging,
persistent toxic chemicals still exist at levels that continue to produce
 adverse effects on fish-eating wildlife and fish  (Giesy et al., 1994) .

      With respect to Great Lakes fish populations  and fish communities, some
 researchers argue that the effects of contaminants  on fish populations and
 communities are difficult to separate from the effects of overfishing, habitat
 loss and the introduction of exotic species (Environment Canada, 1991) .  Great
Lakes lake trout populations, for example, have been devastated by
overharvesting and the introduction of the sea lamprey, as well as subjected
to a variety of impairments from toxic pollutants (Environment Canada, 1991).

      EPA recognizes that chemical contaminants are only one of the threats to
the health of the Great Lakes Basin Ecosystem.  The continued decline of
physical habitat and the presence of exotic (i.e.,  non-indigenous) species,
for example, are also of concern.  The final Guidance provides a consistent
approach for reducing the threat from chemical stressors to the Great Lakes
Basin Ecosystem.  Other programs and activities are currently being
implemented by EPA and other Federal and State agencies to address biological,
chemical and physical problems in the Great Lakes (see section I.D below).

-------
i
8
I I
V f
o J5
tt CJ
|
1 1
8><2
11
IE
I «
Is
l|
oo 13
W P

1
o
s
e
o
•8
3

a-!
* w



•1 I
1 2




§ .
•^ eo
~ §

Qd i?
0 S
fiu Q
GO
!


w


u



u
z

z

X







X




-





X




1


w


w



w

a














-





X







X


g





X









X




-





X




Double-crested
Cormorant |


X


X





X









X




C"





X




Black-crowned Night
Heron |


W


w





X

u







X




-





X




Bald Eagle


X


X



X

X

X







X















Herring Oull


W
z


X







X























Ring-billed Oull |


W


X



W
z











X




X










Caspian Tern


X


X





X









X















Common Tern


X


X



X











X















Forster's Tern j


a


X



a
z

<

u







X




w





z




.1
CO


X








Z









X















Lake Trout


X








Z

























•o
1
e
CO


X


X





Z

























White Sucker






















1
1
0
•s
•2
1
i §
1?
f ^
u
5 e-





= effects documented
A = not applicable


-------
                                Section I: Background
      b.    Trends in Contaminant Levels.  Concentrations of certain
bioaccumulative contaminants, such as PCBs, DDT, dieldrin, and oxychlordane,
have declined significantly over the past 20-30 years, as evidenced by basin-
wide decreases of these pollutants in water, fish, bird eggs, and  sediments
 (DeVault, 1993a; DeVault 1993b; Environment Canada, 1991).  These  declines are
believed to be attributable to bans and restrictions that were placed on  the
manufacture and use*of these chemicals from the late 1960s through the mid-
1970s.  Decreased chemical levels in the Great Lakes are also attributable, in
part, to existing regulatory controls, industrial source controls, and the
Lakes' ability to respond to changes in loads following remedial actions
 (Deposition of Air Pollutants to the Great Waters: First Report to Congress,
EPA 453-R-93-055.D.S. EPA, 1994) (Great Waters Report).

      Recent water column data for both Lakes Superior and Michigan indicate
declines for PCBs are continuing.  Pollutant levels for PCBs, for  example,
declined from 1.73 ng/I> to 0.18 /ig/L in Lake Superior between 1978 and 1992
 {DeVault, 1993).  Between 1980 and 1993, PCB levels in southern Lake Michigan
also declined, from 1.8 /xg/L to 0.2 ^g/L.  Similar results were observed  for
DDT, 2,3,7,8-TCDD and 2,3,7,8-tetrachloro-dibenzofuran  (TCDF) across the  Great
Lakes basin, and in levels of mirex found in Lake Ontario.  Significant
declines in both net loadings and environmental concentrations of  lead and
mercury have also occurred since the mid-1970s.

      Due to difficulties with reliably measuring chemicals at extremely  low
concentrations in water, scientists have often relied upon fish tissue
residues in some species, such as lake trout and walleye, to provide an
indication of chemicals in the water column as well as to indicate any changes
in water column pollutant concentrations.  In the preamble to the  proposed
Guidance, EPA provided environmental information regarding several
bioaccumulative chemicals (e.g., PCBs, DDT) in game fish  (58 FR 20809-20816).
EPA stated that contaminant levels in fish had showed dramatic declines
through the mid-1980s, but were currently fluctuating around a lower level.
This conclusion was based on data showing that earlier declines may have  been
leveling off in recent years (DeVault, 1993).

      The lake trout data which were presented in the preamble to  the proposed
Guidance illustrated that existing PCB levels in such fish were too high  to
ensure adequate projection of human health and that the rate of decline in
fish tissue levels had slowed.   Similarly, the data for coho salmon showed
that PCB levels in this species remained above the fish tissue concentrations
of concern and that the trend of PCBs in this species between 1984 through
1990 had leveled off in both Lakes Michigan and Erie.

      In addition to the data presented in the preamble to the proposed
Guidance, lake trout data for Lakes Superior and Michigan, and walleye data
for Lake Erie, indicate that concentrations of PCBs and DDT initially declined
rapidly, but leveled off in the mid-1980s.  Similarly, concentrations of  PCBs
and DDT in Lake Erie walleye showed no significant changes since 1982.  PCB
and DDT concentrations in lake trout from Lakes Michigan and Superior had not
changed significantly since 1986 (DeVault, 1993).

      Current trends indicate that many water quality objectives and fish
tissue criteria for the protection of human health are still being exceeded
(DeVault et al.,  1994).  In Lake Erie, for example, water column
concentrations of both PCBs and DDT declined from 1980 through 1984, and  have
not changed significantly through 1992, with levels of PCBs still
substantially above the fish tissue concentrations of concern (DeVault et al.,
1994).   PCB and DDT concentrations in coho salmon filets from Lake Michigan
showed rapid declines from 1980 through 1983, but increased between 1983  and
1992.  See Figures i-1 and 1-2.  Thus, even though water column concentrations
of PCBs continued to decline in Lakes Superior and Michigan through the early
1990s,  an increase or lack of decline of PCBs has been observed in Lake
Superior and Lake Michigan fish tissue residues since the mid-1980s.

-------
8      Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      The difference between the declining rates of PCBs in the water column
concentrations and the fish tissue residues in Lakes Superior and Michigan
indicate that these two systems have not reached an equilibrium of existing
loads and bottom sediments.  One possible explanation provided by DeVault et
al.  (1994) is that several recent changes to the zooplankton and forage fish
communities in the Great Lakes System may be responsible.  These changes were
observed concurrently with the slowing and reversals in contaminant declines
discussed above.

      Despite the trends discussed above, the fish tissue concentrations of
PCBs described above show that the acceptable fish tissue concentration of 0.1
pig/g for the protection of biological resources is still being exceeded across
most of the Great Lakes basin  (DeVault et al., 1994).  These contaminant
concentration levels continue to result in exceedances of State and Provincial
human health criteria, potential risks to human health from cancer and
noncancer systemic injuries, and fish consumption advisories for PCBs in each
of the Great Lakes.  Based on this information, further decreases in loadings
to the Great Lakes are necessary in order to meet both existing and proposed
future water quality objectives and criteria.  EPA and the Great Lakes States
and Tribes are currently addressing residual pollutant problems in the Great
Lakes through a variety of regulatory and voluntary programs, some of which
are discussed in section I.D below.

      c.    Fish Consumption Advisories.  Due to the presence of contaminants
at levels above current standards or guidelines, elevated contaminant levels
in wildlife, and contaminated sediment hot spots (DeVault et al., 1994), fish
consumption advisories exist in all of the Great Lakes States, including
various waters located in the Great Lakes basin.  The Great Lakes States issue
these fish contaminant advisories based on a system incorporating and weighing
such factors as the type of contaminants found in Great Lakes fish flesh,
contaminant levels in fish of various sizes and species, the typical
consumption rates of sensitive populations such as sport anglers, nursing and
pregnant women, and an evaluation of the human health risks of the potential
impacts of these substances.

      Pollutants for which these fish advisories are issued include eight of
the 22 bioaccumulative chemicals of concern  (BCCs)  identified in the final
Guidance.  High-risk groups, which fish consumption advisories are established
to protect, include mothers who breastfeed their young, due to the presence of
pollutants in human tissue to which babies continue to be exposed after they
are born.  Also at risk of greater exposure are those sport anglers, Native
Americans, and the urban poor who may consume more Great Lakes fish than the
average consumption of the basin population as a whole.  The Federal
government is examining the impacts to these higher-risk populations through a
number of studies initiated during 1993 and 1994 by the Agency for Toxic
Substances and Disease Registry, within the Department of Health and Human
Services.  As a result of the factors mentioned above, impacts from fish
consumption to higher-risk populations are being taken into consideration in
environmental regulation activities,  including the final Guidance.

-------
                      Section I: Background
0)
   (0

   §
 O '4=
o
 •
I
o
CO
<

o
i
i
|

i
o
           w
                                 =   '    '- L
                                 E  /*    if
                                 —  .  \   -" '
o>

OJ



O)

o
O)

O)
00

00
00

r-
00

(0
00

in
oo


oo

«
00

01
00


00

o
00

o>
              €0
                •
              o
                                  o
                                   »
                                  o
                                        tr
                                        <
                                        Hi
        6/Bn    NOIlVaiN30NOO 1QQ
FIGURE 1-1  (Devault et al. 1986; Devault, 1993)

-------
10
Water Quality Guidance for the Great Lakes System - Supplementary Information Document
i*
"5 3
CO
    CO
0  £
-S -2
O -M
o  2
   *-•
»  £
4)  

O) o> o 00 0> 00 00 00 N. 00 • oo 00 CO 00 CM 00 00 o 00 0> q o B/Bn NOIlVaiNBONOO 3Od FIGURE 1-2 (Devault et al. 1986; Devault, 1993)


-------
                                Section I: Background                             11
C.    History of the Great Lakes Water Quality Guidance

l.    The Great Lakes Toxic Substances Control Agreement

      Against this background of environmental concerns, the Governors of the
eight Great Lakes States signed the Great Lakes Toxic Substances Control
Agreement (Governors' Agreement) pledging the States' cooperation in studying,
managing and monitoring the Great Lakes as an integrated ecosystem in 1986.
The purpose of the Governors Agreement was to establish a framework for
coordinated regional action in controlling toxic substances entering the Great
Lakes System; to further the understanding and control of toxic contaminants;
and to develop common goals, management practices and control strategies for
toxics to ensure a cleaner and healthier Great Lakes Basin Ecosystem.
      The Governors believed that maintaining and improving the quality of
Great Lakes waters would sustain water supply systems as well as commercial,
manufacturing and recreation industries, while creating new economic
development opportunities.  They agreed to preserve the value of the Great
Lakes by striving to maintain a high standard of water quality when
establishing regulatory standards, and committed to continue reducing toxics
in the Great Lakes System to the maximum extent possible.  For further
discussion of the principles outlined in the Governors' Agreement, see the
preamble to the proposed Guidance (58 PR 20820) .
      The Great Lakes Governors also realized that a number of toxic
substances control programs and measures had been responsible for achieving
significant reductions in certain organic pesticides and metals.  These
reductions are discussed in more detail in section I.B.2 above.  However, the
Governors recognized that the complexity of the pollutant problem in the Great
Lakes System called for continuing action.  In many cases, the Governors
believed new and creative initiatives would be required to achieve the goal of
prohibiting the discharge of toxic pollutants in toxic amounts in order to
protect the water quality of the Great Lakes.
      To implement the goals of the Governors' Agreement, the signatory States
directed their environmental administrators to jointly develop an agreement
for coordinating the control of toxic releases within the Great Lakes System.
This coordinated effort between the Great Lakes States contributed to the
development of the Great Lakes Water Quality Initiative.

2.    The Great Lakes Water Quality Initiative

      EPA and the Great Lakes States initiated the Great Lakes Water Quality
Initiative (Initiative) in 1989 to further address the environmental concerns
identified in the Governors Agreement.  The Initiative was intended to provide
a forum for Great Lakes States and EPA to develop uniform water quality
criteria and implementation procedures for the Great Lakes basin.  The
participants planned to use the results of this effort as a basis for revising
State water quality standards and permit programs pursuant to sections 303(c)
and 402 of the Clean Water Act  (CWA).     Three committees were formed under
the Initiative.  A Steering Committee, consisting of directors of water
programs from EPA's National and Regional offices and the Great Lakes States'
environmental agencies, as co-regulators of CWA water quality programs,
discussed policy, scientific and technical issues and directed the work of the
Technical Work Group.  The Technical Work Group, consisting of technical staff
from the Great Lakes States environmental agencies, EPA, the U.S. Fish and
Wildlife Service and the U.S. National Park Service, prepared proposals for
submission to the Steering Committee.  The Public Participation Work Group,
consisting of representatives from environmental groups, municipalities,
industry and academia, observed the deliberations of the other two groups,
advised them of the public's concerns, and kept its various constituencies
apprised of Initiative activities.  These committees are collectively known as
the Initiative Committees.

      Of particular' concern to the Steering Committee were those pollutants
which exhibit the potential to produce System-wide impacts.  Based upon
observed impairments to the Great Lakes, the States believed that these

-------
12     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

pollutants pose one of the greatest threats to the health of the Great  Lakes
Basin Ecosystem.  The Steering Committee believed that further reductions  in
loadings of such pollutants from all sources should be pursued.  The Steering
Committee wanted actions to be taken to ensure that problems with pollutants
which could potentially cause impairment of beneficial uses in the Great Lakes
Basin Ecosystem would not develop in the future.  Therefore, the Steering
Committee charged EPA and State staff on the Technical Work Group to define
the pollutants that,warrant more stringent controls, and to draft additional
control approaches for those pollutants.  Further discussion on the factors
chosen for identifying and controlling such pollutants is provided in section
II.C.8 of this document.

3.     The Great Lakes Critical Programs Act of 1990

      The enactment of the Great Lakes Critical Programs Act  (CPA) of 1990
(Public Law 101-596, November 16, 1990) codified the ongoing Initiative effort
into the CWA.  Section 101 of the CPA  (CWA section 118(c)(2)) requires  EPA to
publish proposed and final water quality guidance for the Great Lakes System
which conforms with-the objectives and provisions of the Great Lakes Water
Quality Agreement (GLWQA) and is no less restrictive than provisions of the
CWA and National water quality criteria and guidance.  The final Guidance must
specify minimum requirements for the waters of the Great Lakes System in three
areas: (a) water quality criteria, including numerical limits on pollutants in
ambient Great Lakes waters to protect human health, aquatic life and wildlife;
(b) antidegradation policies; and (c) implementation procedures.  This  section
also requires Great Lakes States to adopt water quality standards,
antidegradation policies and implementation procedures consistent with  the
final Guidance within two years of EPA's publication of the final Guidance, or
be subject to EPA promulgation of the provisions within the same two-year
period.  (CWA section 118(c)(2)(C)).  Further discussion of the CPA is provided
in the preamble to the proposed Guidance (58 FR 20823).

      The substantive scope of the final Guidance is consistent with the
initial 1989 expectations for the original Initiative.  The final Guidance
includes water quality criteria and methodologies, an antidegradation policy,
and implementation procedures that were initially developed through an  open,
collaborative process by staff from the Great Lakes States and EPA with input
from other stakeholders in the basin.  The principal changes in the scope  of
the final Guidance that have occurred since the initial planning for the
Initiative resulted, from the enactment of the CPA.

4.     Principles Underlying the Final Water Quality Guidance for the Great
      Lakes System

      The final Guidance is the culmination of a six-year cooperative effort
that included participation by the eight Great Lakes States, the environmental
community, academia, industry, municipalities and EPA Regional and National
offices.  EPA's development of the final Guidance continued the work begun by
the Initiative Committees and was guided by the following six general
principles:

      a.    Use the Best Available Science to Provide Protection to Human
Health. Wildlife, and Aquatic Life.  EPA and the Initiative Committees  have
been committed to using the best available science to develop programs  to
protect the Great Lakes System since the beginning of this ambitious project.
In the 1986 Governors' Agreement, the Governors of the Great Lakes States
recognized that the problem of persistent toxic substances was the foremost
environmental issue confronting the Great Lakes.  They also recognized  that
the regulation of toxic contaminants was scientifically  complex because the
pollutants are numerous, their pathways into the Lakes are varied, and  their
effects on the environment, aquatic life and human health were not completely
understood.  Based  on the  importance  of the Great Lakes  Basin Ecosystem and
the documented adverse  effects from toxic contamination,  however, the
Governors directed  their environmental administrators  to jointly  develop an

-------
                                Section I: Background                             13
agreement and procedure for coordinating the control of toxic releases and
achieving greater uniformity of regulations governing such releases within the
Great Lakes basin. •

      As discussed further above, the Initiative was subsequently created to
begin work on these goals.  EPA and the Great Lakes States, with input from
interested parties- in the basin, began collecting and analyzing data,
comparing regulatory requirements and technical guidance in their various
jurisdictions, and drafting specific methodologies and prpcedures to control
the discharge of toxic contaminants.  The provisions of the final Guidance
were based in large part on these prior efforts of the Initiative Committees,
and incorporate the best available science to protect human health, wildlife
and aquatic life intthe Great Lakes System.  For example, the final Guidance
includes new criteria and methodologies developed by the Initiative Committees
to specifically protect wildlife; incorporates recent data on the
bioavailability of metals into the aquatic life criteria and methodologies;
incorporates Great Lakes-specific data on fish consumption rates and fish
lipid contents into the human health criteria; and provides a better
methodology to determine the bioaccumulation properties of individual
pollutants.

      b.    Recognize the Unique Nature of the Great Lakes Basin Ecosystem.
The final Guidance also reflects the unique nature of the Great Lakes Basin
Ecosystem by establishing special provisions for BCCs.  The special provisions
for BCCs initially developed by the Initiative Committees and incorporated
into the final Guidance include more stringent antidegradation procedures, to
ensure that these problems are minimized; phase out and elimination of mixing
zones for existing discharges of BCCs after 12 years, with limited exceptions,
to reduce overall loadings to the Lakes;  more extensive data generation
requirements to ensure that they are not underregulated for lack of data; and
the development of water quality criteria that will protect wildlife that feed
on aquatic prey.  As discussed in sections I.A and I.B above, it is reasonable
and appropriate to establish special provisions for these chemicals because of
the physical, chemical and biological characteristics of the Great Lakes
System, and the documented environmental harm to the ecosystem from the past
and continuing presence of these types of pollutants.

      The final Guidance is not only designed to address existing problems,
but also to prevent emerging and potential problems posed by additional
chemicals in the future which may damage the overall health of the Great Lakes
System.  The experience with such pollutants as DDT and PCBs indicates that it
takes many decades to overcome the damage to the ecosystem caused by even
short-term discharges, and that prevention would have been dramatically less
costly than cleanup.   The elements of the final Guidance provide a coordinated
ecosystem approach for addressing possible pollutant problems before they
produce adverse and long-lasting basinwide impacts, rather than waiting to see
what the future impacts of the pollutants might be before acting to control
them.  The comprehensive approach used in the development of the final
Guidance provides regulatory authorities with both remedial and preventive
ways of gauging the actions and potential effects of chemical stressors upon
the Great Lakes Basin Ecosystem.  The methodologies, policies and procedures
contained in the final Guidance provide mechanisms for appropriately
addressing both pollutants that have been or may in the future be documented
as chemicals of concern.  Additional discussion of the characteristics and
processes affecting contaminant levels in the Great Lakes and special controls
for BCCs is also provided in the preamble to the proposed Guidance (58 PR at
20807-23 and 20844-45) and in section II.C.8 of this document.

      c.    Promote Consistency in Standards and Implementation Procedures
While Allowing Appropriate Flexibility to States and Tribes.  Promoting
consistency in standards and implementation procedures while providing for
appropriate State and Tribal flexibility was• the third principle in EPA's
development of the final Guidance.  A primary impetus for the Governors'
Agreement, the Initiative, and the requirements set forth in the CPA was a

-------
14     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

recognition of the need to promote consistency in the minimum water quality
standards, antidegradation policies, and implementation procedures adopted by
the Great Lakes States to protect human health, aquatic life and wildlife.
Although provisions in the CWA provide for the adoption of and periodic
revisions to State water quality criteria, such provisions do not necessarily
ensure that water quality criteria of adjoining States are consistent within a
shared waterbody.  For example, State acute ambient water quality criteria in
place in six of the.eight Great Lakes States include a range of 1.79 to 15.0
/ig/L for cadmium in order to protect against acute effects for aquatic  life,
and from 0.21 to 1.33 /zg/L for dieldrin.  Other examples of variations  in
acute ambient water quality criteria include: nickel, which ranges from 290.30
jjg/L -to 852.669 jig/L; lindane, with a range of no criteria in place to  1.32
fjg/L; and mercury, ranging from 0.5 M9/L to 2.4 jig/L.  Similar ranges and
disparities exist in the chronic ambient water quality criteria and human
health criteria adopted by the Great Lakes States.

      Disparities also exist among State procedures to derive individual
discharge permits from water quality criteria.  Wide variations exist,  for
example, in procedures for granting mixing zones, interpretation of background
levels of pollutants, consideration of pollutants present in intake waters,
controls for pollutants present in concentrations below the level of
detection, and determination of appropriate levels for the discharge of
multiple pollutants.  Additionally, when calculating exposure factors in fish
that will be consumed by humans and wildlife, some States consider
accumulation through multiple steps in the food chain  (bioaccumulation) while
others consider only the single step of concentration from the water column
(bioconcentration).  Further disparities exist in different translator
methodologies in deriving numeric values for implementing narrative water
quality criteria; different assumptions when calculating total maximum  daily
loads (TMDLs) and wasteload allocations, including different assumptions about
background concentrations, mixing zones, receiving water flows, or
environmental fate; and different practices in deciding what pollutants need
to be regulated in a discharge, what effect detection limits have on
compliance determinations, and how to develop whole effluent toxicity
limitations.

      These inconsistencies in State standards and implementation procedures
have resulted in the disparate regulation of point source discharges, which
may have led to disputes in the past among the Great Lakes States.  In  the
Governors' Agreement, however, the Governors recognized that the water
resources of the basin transcend political boundaries and committed to  taking
steps to manage the Great Lakes as an integrated ecosystem.  The Great  Lakes
States, as part- of the Initiative Committees, recommended provisions that were
ultimately included in the proposed Guidance for coordinated review and
development based on their extensive experience in administering State  water
programs and knowledge of the significant differences in these programs within
the basin.  The final Guidance incorporates the work begun by the Initiative
Committees to identify these disparities and improve consistency in water
quality standards and permit procedures in the Great Lakes System.

      Although improved consistency in State water programs is a primary goal
of the final Guidance, it is also necessary to provide appropriate flexibility
to States and Tribes in the development and implementation of water programs.
In overseeing States' implementation  of the CWA, EPA has found that reasonable
flexibility is not only necessary to  accommodate site-specific situations and
unforeseen circumstances, but is also appropriate to enable innovation  and
progress as new approaches and information become available.  Many commenters
urged EPA to evaluate the appropriate level of flexibility provided to  States
and Tribes in the proposed Guidance provisions.  EPA reviewed all sections  of
the proposed Guidance and all comments received to determine the appropriate
level of flexibility needed to address these concerns  while still providing a
minimum level of  consistency between  the  State and Tribal programs.  Based  on
this review, the  final Guidance provides  flexibility for State and Tribal

-------
                                Section I: Background                             15
adoption and implementation of the final Guidance provisions in many areas,
including the following:

          Antidegradation:  Great Lakes States and Tribes may develop their
own approaches for implementing the prohibition against deliberate actions of
dischargers that increase the rate of mass loading of BCCs without an approved
antidegradation demonstration.  Furthermore, States and Tribes have
flexibility in adopting antidegradation provisions regarding non-BCCs.

          TMDLs:  Great Lakes States and Tribes may use assessment and
remediation plans for the purposes of appendix F of part 132 if the State or
Tribe certifies that the assessment and remediation plan meets certain TMDL-
related provisions in the final Guidance and public participation requirements
applicable to TMDLs, and if EPA approves such plan.  Thus, States have the
flexibility in many cases to use Lakewide Management Plans (LAMPs), Remedial
Action Plans (RAPs) and State Water Quality Management Plans in lieu of TMDLs.

          Intake Credits:  Great Lakes States and Tribes may consider the
presence of intake water pollutants in establishing water quality-based
effluent limits (WQBELs) in accordance with procedure 5 of appendix F to part
132, as discussed further in section VIII.E of this document.

          Site-Specific Modifications: Great Lakes States and Tribes may adopt
either more or less'stringent modifications to human health,  wildlife, and
aquatic life criteria and BAFs based on site-specific circumstances specified
in procedure 1 of appendix F to part 132, as discussed further in section
VIII.A of this document.  All criteria, however, must not be likely to cause
jeopardy to threatened or endangered species listed or proposed to be listed
under the Federal Endangered Species Act  (ESA) .

          Variances:  Great Lakes States and Tribes may grant variances from
water quality standards based on the factors identified in procedure 2 of
appendix F to part 132, discussed further in section VIII.B of this document.

          Compliance Schedules:  Great Lakes States and Tribes may allow
existing Great Lakes dischargers additional time to comply with permit limits
in order to collect data to derive new or revised Tier I criteria and Tier II
values in accordance with procedure 9 of appendix F to part 132, as discussed
further in section VIII.I of this document.

          Mixing Zones:  Great Lakes States and Tribes may authorize mixing
zones for existing discharges of BCCs after the 10-year phase-out period in
accordance with procedure 3.B of appendix F to part 132, if the permitting
authority determines, among other things, that the discharger has reduced its
discharge of the BCC for which a mixing zone is sought to the maximum extent
possible.  Water conservation efforts that result in overall reductions of
BCCs are also allowed even if they result in higher effluent concentrations.
These provisions are discussed further in section VIII.C.4 of this document.

          Scientific Defensibility Exclusion:  Great Lakes States and Tribes
may apply alternate procedures consistent with Federal, State, and Tribal
requirements upon demonstration that a provision in the final Guidance would
not be scientifically defensible if applied to a particular pollutant in one
or more sites.   This provision, in § 132.4(h) of the final Guidance, is
discussed further in section II.C.6 of this document.

          Reduced Detail:  In many instances, EPA has revised the proposed
Guidance to reduce the amount of detail in the provisions without sacrificing
the objectives of the provisions.  Examples of such revisions include
simplification of procedures for developing TMDLs in procedure 3 of appendix F
to part 132,  discussed further in section VIII.C of this document, and
simplification of procedures for determining reasonable potential to exceed
water quality standards in procedure 5.B of appendix F to part 132, discussed
further in section VIII.E of this document.

-------
16     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

          Other Provisions:  Flexibility is also present in provisions for the
exercise of best professional judgment toy the Great Lakes States and Tribes
when implementing many individual provisions in the final Guidance including:
determining the appropriate uncertainty factors in the human health and
wildlife criteria methodologies; selection of data sets for establishing water
quality criteria; identifying reasonable and prudent measures in
antidegradation provisions; and specifying appropriate margins of safety when
developing TMDLs.  In all cases, of course, State and Tribal provisions would
need to be scientifically defensible and consistent with all applicable
regulatory requirements.

      d.    Establish Equitable Strategies to Control Pollution Sources.  Many
commenters argued that the proposed Guidance unfairly focused on point source
discharges.  They asserted that nonpoint sources or diffuse sources of
pollution, such as air emissions, are responsible for most of the loadings of
some pollutants of concern in the Great Lakes, that increased regulation of
point sources will be inequitable and expensive, and that the final Guidance
will not result in any environmental improvement given the large, continuing
contribution of toxic pollutants by nonpoint sources.

      EPA recognizes that regulation of point source discharges alone cannot
address all existing or future environmental problems from toxic pollutants in
the Great Lakes.  In addition to discharges from point sources, toxic
pollutants are also,contributed to the Great Lakes from industrial and
municipal emissions to the air, resuspension of pollutants from contaminated
sediments, and urban and agricultural runoff, hazardous waste and Superfund
sites, and spills.  Restoration and maintenance of a healthy ecosystem will
require significant efforts in all of these areas.  As discussed further in
section I.D below, EPA, Canada and the Great Lakes States and Tribes are
currently implementing or developing many voluntary and regulatory programs to
address these and other nonpoint sources of environmental contaminants in the
Great Lakes.

      Additionally, EPA intends to use the scientific data developed in the
final Guidance and fiew or revised water quality criteria subsequently adopted
by Great Lakes States and Tribes in evaluating and determining appropriate
levels of control in other environmental programs.  For example, EPA's future
biennial reports under section 112(m) of the Clean Air Act will consider the
extent to which air discharges cause or contribute to exceedances of water
quality criteria in assessing whether additional air emission standards or
control measures are necessary to prevent serious adverse effects.  Similarly,
once the provisions of the final Guidance are adopted by the Great Lakes
States or Tribes, they will serve as applicable or relevant and appropriate
requirements  (ARARs) for on-site responses under the Comprehensive
Environmental Response, Compensation and Liability Act  (CERCLA).  EPA will
also consider the data and criteria developed for the final Guidance,
including the information on bioaccumulative pollutants, in developing or
evaluating LaMPs and RAPs under  section 118  of the CWA and Article VI, Annex 2
of the GLWQA; determination of corrective action requirements under sections
3004(u), 3008(h), or 7003 of the Solid Waste Disposal Act; new or existing
chemical reviews under the Toxic Substances  Control Act  (TSCA); pesticide
reviews under the Federal  Insecticide, Fungicide and Rodenticide Act  (FIFRA) ,-
and reporting requirements for toxic releases under the Emergency Planning and
Community Right-to-Know Act  (EPCRA).

      The final  Guidance also  includes provisions to address the contribution
of pollutants by nonpoint  sources.  First,  the water quality criteria  to
protect human health, wildlife and aquatic  life, and the antidegradation
provisions  apply to the waters  in  the Great Lakes System regardless of whether
discharges  to the water are  from point or nonpoint sources.  Accordingly,  any
regulatory programs for nonpoint sources that require compliance with  water
quality standards would also be  subject to  the  criteria and antidegradation
provisions  of the final Guidance once they  are  adopted  into State or Tribal
standards.

-------
                                Section I: Background                             17
      Second, several elements of the final Guidance would, after State,
Tribal or Federal promulgation, require or allow permitting authorities to
consider the presence of pollutants in ambient waters--including pollutants
from nonpoint source dischargers--in establishing WQBELs for point sources.
For example, permit'authorities may consider the presence of other point or
nonpoint source discharges when evaluating whether to grant a variance from
water quality .criteria under procedure 2 of appendix F to part 132.
Additionally, the provisions for TMDLs in procedure 3 of appendix F to part
132 address nonpoint sources by requiring allocation of the available load
capacity of receiving waters that do not meet water quality criteria among all
sources of the pollutant, including nonpoint sources.  The development of
TMDLs is the preferred mechanism for addressing equitable division of the
loading capacities of these nonattained waters.  Because TMDLs have not been
completed for most nonattained waters, however, the final Guidance promotes
the development of TMDLs through a phased approach, where appropriate, and
provides for short-term regulatory relief to point source dischargers in the
absence of TMDLs through intake credits, variances, and other water quality
permitting procedures.  These provisions are discussed in sections VIII.B,
VTII.C, and VIII.F of this document.

      e.    Promote Pollution Prevention Practices.  The final Guidance also
promotes pollution prevention practices consistent with EPA1 s National
Pollution Prevention Strategy and the Pollution Prevention Action Plan for the
Great Lakes.  The Pollution Prevention Act of 1990 declares as National policy
that reducing the sources of pollution is the preferred approach to
environmental protection (Pub. L. 101-508, section 6601-6610 104 Stat. 1388,
codified at section 13101-13109 West Supp. 1991) .  When source reductions are
not possible, however, recycling, treating and properly disposing of
pollutants in an environmentally safe manner complete the hierarchy of
management options designed to prevent pollution from entering the
environment.

      Consistent with the goals of the Pollution Prevention Act, EPA developed
the Great Lakes Pollution Prevention Action Plan (April, 1991) .  The Great
Lakes Pollution Prevention Action Plan highlights how EPA, in partnership with
the Great Lakes States, will incorporate pollution prevention into actions
designed to reduce'the use and release of toxic substances in the Great Lakes
basin.  The Great Lakes Pollution Prevention Action Plan is discussed in more
detail in the preamble to the proposed Guidance (58 PR 20827-28).

      The final Guidance builds upon these two components of the Great Lakes
program by promoting the development of pollution minimization programs in the
level of detection,  mixing zone elimination, and antidegradation sections of
the final Guidance.   Also,  the decision to provide special provisions for BCCs
implements EPA's commitment to pollution prevention by reducing the discharge
of these pollutants in the future.  This preventive step not only makes good
environmental management sense, but is appropriate based on the documented
adverse effects that the past and present discharge of these pollutants have
produced in the Great Lakes System.

      f.    Provide Accurate Assessment of Costs and Benefits.  In developing
the final Guidance,  EPA identified and carefully evaluated the anticipated
costs and benefits of the major provisions.  EPA received many comments on the
draft cost and benefit studies conducted as part of the proposed Regulatory
Impact Analysis (RIA)  required by Executive Order 12291, and its successor,
Executive Order 12866.  Based upon consideration of those comments and further
analysis, EPA has revised the RIA.  The results of this analysis are
summarized in section IX of this document.

D.    Progress on Other Programs to Protect and Restore the Great Lakes System

      The final Guidance is only one component of many Federal, State and
Tribal programs designed to address environmental contamination in the Great
Lakes System.  Unacceptably high concentrations of bioaccumulative pollutants,

-------
18     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

including PCBs, currently exist in fish tissue and sediments through the
basin.  As discussed further above, these pollutants have the potential to
cause significant adverse human health effects, including increased risk of
cancer and systemic injuries, as well as to adversely impact the aquatic life
and wildlife in this ecosystem.

      EPA recognizes that implementation of the final Guidance alone will not
solve the existing environmental problems from chemical pollutants in the
Great Lakes.  Complete restoration  of the Great Lakes and achievement of
water quality criteria will require significant time, resources, and the
scientific expertise of many parties.  Other regulatory and voluntary programs
are currently underway, however, to prevent or reduce future pollutant
loadings and to remediate the averse effects associated with past pollutant
discharges to the Great Lakes System.  Implementation of the provision in the
final Guidance, in conjunction with the activities described below, will
provide EPA and the"Great Lakes States and Tribes with a comprehensive
framework for reducing toxic loadings.  In addition, these activities will
enable EPA and the Great Lakes States and Tribes to develop cost-effective
strategies to further the goals and requirements of the CWA, Great Lakes Toxic
Reduction Effort (GLTRE)  and GLWQA, and attain the level of water quality
necessary to fully protect human health, wildlife and aquatic life in the
Great Lakes System.

1.    The Great Lakes Toxic Reduction Effort

      EPA and the Great Lakes States and Tribes have established a multi-media
strategy called the GLTRE to achieve further reductions in the use and release
of toxic substances to the Great Lakes System.  The GLTRE emphasizes
addressing nonpoint sources and wet weather point sources of pollution, and is
consistent with the Great Lakes 5-year Strategy discussed in the preamble to
the proposed Guidance  (58 FR 20826).

      The GLTRE has three multi-media tracks.  The Pathway track focuses on
the primary nonpoint source paths and wet weather sources through which BCCs
enter the Great Lakes System.  The Virtual Elimination Project focuses on
identifying the ongoing uses and sources of the BCCs and identifies specific
actions designed to* achieve further reductions of these pollutants.  The third
track, the Lake Michigan Enhanced Monitoring Program, being pursued as part of
the Lake Michigan LaMP, is designed to develop a sound scientific basis to
guide future pollution prevention and reduction efforts to address toxic
pollutants  in the Great Lakes.

      a.    Pathway Track. ' The first track of the GLTRE has identified five
primary wet weather and nonpoint  source pathways for BCCs: air deposition;
Combined Sewer Overflow  (CSO)/stormwater/runoff; contaminated sediments;
storage, handling and transport  (spills); and  leaking waste storage sites.
The GLTRE will address any gaps or barriers  in existing regulatory and non-
regulatory programs designed to prevent and reduce the introduction of BCCs
through these five sources.  The  final product of the GLTRE Pathway Track will
be a menu of actions and recommendations aimed at focusing current and
emerging program authorities on preventing,  controlling and reducing loadings
of BCCs; improving reporting, education and outreach; improvements in
monitoring  and modeling; and risk communication techniques.  Actions and
recommendations that enhance media-specific  or multi-media regulatory gaps and
eliminate barriers to effective regulation,  including use of the ambient water
quality criteria in the final Guidance  in other media regulatory programs,
will also be addressed.  The Pathway track will encourage the prevention of
BCC use, recycling and proper disposal  of BCCs, as well as the  replication of
exemplary or innovative prevention and  reduction programs.  Prevention
measures may also include  recommendations for  bans or sunsets for  certain
BCCs, if appropriate.

      b.    Virtual Elimination Pro-ject.  The  goal of the second component of
the GLTRE is the virtual elimination of bioaccumulative, persistent toxic

-------
                                Section I: Background                             19
substances from the Great Lakes basin.  The project identifies the use and
release of specific toxic substances in the Great Lakes basin and examines the
existing regulatory framework that applies to each substance.  The project, is
designed to provide, recommendations for voluntary and incentive-based changes
to increase the pace and level of reductions of bioaccumulative toxics.  This
project is initially focusing on PCBs and mercury for the purpose of finding
opportunities to achieve virtual elimination of discharges through voluntary
source reductions.

      c.    Lake Michigan Enhanced Monitoring Program.  As discussed in
section I.D.4.b below, this program will develop a sound, scientific base of
information to guide toxic pollutant load reduction efforts at the State and
Federal levels.  The Lake Michigan Enhanced Monitoring Program will help
determine: loadings of contaminants from tributaries, the atmosphere and open
lake sediments; concentrations and fluxes of toxic chemicals in the food web;
and the magnification of toxic chemicals concentrations through representative
food chains.  This work will result in a better understanding of the relative
sources and fate of toxic pollutants in Lake Michigan, and will be of use in
addressing pollutant concerns in all of the Great Lakes.  The results of this
program will also be used to make recommendations on regulatory and
nonregulatory changes needed to fully achieve water quality standards.

      2.    Clean Air Act Amendments

      Implementation of the major provisions of the Clean Air Act Amendments
of 1990 (CAAA) is ah integral part of EPA's broader program to protect and
restore the water quality of the Great Lakes.  Specifically, in response to
information indicating that air pollution contributes significantly to water
pollution, Congress included section 112(m), referred to as the Great Waters
program, in the CAAA.  The purpose of the Great Waters program is to evaluate
the atmospheric deposition of air pollutants to the Great Lakes, Lake
Champlain, Chesapeake Bay, and coastal waters.  As required by this provision,
EPA completed the Great Waters Report in May, 1994.  The Great Waters Report
includes information on the contribution from atmospheric deposition of
pollutant loadings, the environmental or public health effects of such
pollution, the source or sources of such pollution, and a description of
regulatory steps EPA plans to initiate under applicable Federal laws aimed at
the protection of human health and the environment.

      The Great Waters Report observed that water quality conditions in the
Great Lakes, among other waterbodies, are greatly improved compared to a few
decades ago.  Although these improvements are credited to environmental
regulatory programs and public and industrial cleanup efforts that are
addressing primarily waterborne pollution, the Great Waters Report cautions
that the Great Waters ecosystems are far from fully recovered.  The Great
Waters Report adds that it is now necessary to address the more diffuse
sources of pollution, including the air component, in order to attain water
quality goals and to ensure protection of human health and the environment.

      The Great Waters Report concluded that although specific data may be
available for some waterbodies, such as Lake Superior, insufficient data are
available to generalize the atmospheric loadings to all waters.  Overall,
scientists estimate that from 35 to 50 percent of current yearly inputs of a
variety of toxic chemicals to the Great Lakes may be from the air.  Adverse
effects of the chemicals of concern addressed in the Great Waters Report are
evident and studies of selected waters show significant proportions of toxic
pollution coming from the atmosphere.  Although the Great Waters Report noted
that uncertainties in current information are significant, and further
research is needed to better characterize the information for decisionmakers,
adequate information is available to lead EPA to the conclusion that some
actions are justified and necessary at this time.  Because the linkage between
specific sources and subsequent deposition and effects has yet to be
demonstrated,  however, the kinds of action recommended in the Great Waters

-------
20     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

Report focus on the chemicals of concern   (e.g., mercury, PCBs,  dioxins)
rather than on specific sources.

      EPA considered the implications of action and of inaction, while
recognizing that section 112(m) of the CAAA mandates that EPA should act to
"prevent" adverse effects and to "assure protection of human health and the
environment."  EPA's recommendation is that specified reasonable actions are
justified, based on evaluation of the scientific information currently
available, and should now be taken, and that research should continue.

      Most of the initial actions EPA is undertaking focus on utilizing
regulatory mechanisms in the CAAA that are intended to address the most
hazardous chemicals.  The characteristics of toxicity, persistence, and
tendency to bioaccumulate warrant special treatment of the Great Waters
pollutants of concern.  This is consistent with Congressional intent for those
regulatory mechanisms and for section 112(m) of the CAAA.

      The Great Waters Report recommended that for specific source 'categories
of particular concern in the Great Lakes area, CAAA section 112(d) maximum
achievable control technology  ("MACT") standards should be issued ahead of
schedule, where possible.  It also recommended evaluating whether lesser
quantity emission rates, as defined by section 112(a) of the CAAA, should be
set for pollutants of particular concern in the Great Lakes System, and
considering setting area source standards for some source categories.
Following the initial Great Waters Report,  the CAAA requires additional
reports on a biennial basis.  Based on the reports required by section  112(m)
of the CAAA, EPA is required by November 15, 1995, to promulgate further
emission standards or control measures under section 112 of the  CAAA, if
necessary to prevent serious adverse effects, or recommend necessary
regulatory changes under other applicable Federal legislation.

      Between 1992 and 2000, EPA must promulgate technology-based emission
standards for all categories of major sources emitting the 189 toxic air
pollutants listed in section 112(b) of the CAAA.  Area sources will be
regulated in those cases that the Administrator determines are justified in
order to protect health and the environment.  Within eight years of the
promulgation of such standards, EPA will evaluate whether the health and
environmental risk remaining after the application of the control requirements
is significant enough to warrant additional regulatory requirements.  To date,
EPA has promulgated final technology-based standards for Resource Conservation
and Recovery Act (RCRA) treatment, storage and disposal facilities, the
synthetic organic chemical manufacturing industry, coke ovens, industrial
cooling towers, the dry cleaning industry,  halogenated solvent cleaning,
magnetic tape production, commercial ethylene oxide sterilizers, and has
proposed several more standards including those covering epoxy and polymer
resin production, the pulp and paper industry, marine vessels, off-site waste
treatment, petroleum refineries, secondary lead smelters, ship building, wood
furniture manufacturing, municipal waste combusters, chromium electroplating,
and gasoline distribution.    New source performance standards have been
proposed for several industry categories including synthetic organic
chemicals, wastewater and cold-cleaning digesters.  The standard for the
synthetic organic chemical manufacturing industry alone is estimated to reduce
toxic emissions by 510,000 tons a year  (59 FR 19410, April 22, 1994).
Proposed new source performance standards and emissions guidelines have been
published for municipal waste combusters sources greater than 35 MG/day (56 FR
5514), and are scheduled for final promulgation in September, 1995.

      Under section 112 of the CAAA, EPA may add additional substances  to  the
list of hazardous air pollutants, including pollutants of concern in the Great
Lakes, provided the criteria for additions as identified in CAAA section
112(b) are met.  EPA also has discretion to add pollutants when  scientific
information dictates -additions are warranted:  Once a pollutant  is listed, EPA
must set technology-based standards for sources of that pollutant by the year
2000 or within two years of  listing, whichever  is  later.

-------
                                Section I: Background                              21
      Section 112(c)(6) of the CAAA provides for a reduction  in  sulfur  dioxide
 (S02) emissions from utilities of approximately 10 million  tons  by 2010.   The
 first phase of these reductions goes  into effect in  1995.   Not only will  a
 portion of this reduction occur within  the Great Lakes, but many of the
 control technologies likely to be used  to achieve these reductions may  reduce
 toxic air pollutants, such as mercury,  that are specifically  of  concern in the
 Great Lakes System.  In addition, under section 112(n)(1)(B)  of  the CAAA,  EPA
 is conducting-a study of mercury emissions from utilities and various other
 sources, which will include an assessment of the health and environmental
 effects of such emissions.

      EPA will continue ongoing efforts to implement  section  112 of the CAAA
 and other sections of the CAAA and use  the results of the Great  Waters  Report
 in the development of policy and regulations that will reduce emissions of
 Great Waters pollutants of concern.   EPA also recognizes the  need for an
 integrated multimedia approach to the problem of air  deposition  and will
 utilize authorities beyond the CAAA to  reduce human and environmental exposure
 to pollutants of-concern.

 3.    ARARs and the Superfund Program
                   *
      Section 121 of the CERCLA  (or Superfund) establishes  cleanup standards
 for remedial actions.  Under CERCLA,  remedial actions must  attain a degree of
 cleanup that assures protection of human health and the environment.  For  any
 hazardous substances remaining on-site,  the remedy must attain any applicable
 or relevant and appropriate standard, requirement, criteria,  or  limitation
 promulgated under any Federal or State  environmental  law, or  any promulgated
 State standard requirement, criteria  or limitation that is  more  stringent  than
 the Federal regulations.  These requirements are commonly referred to as
 ARARs.  Under CERCLA, a requirement is  applicable if  the environmental  law or
 regulation directly addresses the circumstances at a  site.  If not applicable,
 then a requirement may be relevant and  appropriate if circumstances at  a site
 are sufficiently similar to the problems or situations regulated by the
 requirement that it is well-suited to the site.

      Among the CWA regulations that  may be ARARs for CERCLA  actions are the
 requirements governing discharges of  pollutants to surface  waters,  including
 State, Tribal or Federal water quality  standards.  To be an ARAR,  a standard
 must be promulgated and legally enforceable.  The final Guidance is not, by
 itself, enforceable, and is, therefore,  not an ARAR,  although it could  be
 treated as a nonbinding policy to be  considered in CERCLA actions.   Provisions
 consistent with the,final Guidance will  become enforceable  only  when adopted
 by a State or Tribe as part' of its National Pollutant Discharge  Elimination
 System (NPDES) or water quality standards programs, promulgated  by EPA  in  the
 absence of State or Tribal action, or when included in an NPDES  permit.  When
 such adoption or promulgation occurs, those provisions will be ARARs, as
 appropriate,  for on-site CERCLA discharges.  Off-site CERCLA  discharges must
 comply fully with all applicable Federal, State or Tribal requirements.

 4.    RAPS and LaMPs

      The United States and Canadian  Governments agreed to  develop and
 implement RAPs and LaMPs pursuant to Article VI, Annex 2 of the  GLWQA.  The
 development of RAPs and LaMPs is also required by section 118(c)(3) and (4)  of
 the CWA.   RAPs,  which are designed to address impairments to  any one of 14
 specified beneficial uses, are being  developed for each of  43 designated Areas
 of Concern (AOCs).  LaMPs, which are  designed to address loadings of critical
pollutants that are currently impairing or have the potential to impair the
 beneficial uses of the open waters of the Great Lakes, are  being developed by
 EPA in conjunction with the States, Tribes and Canada on a  phased basis.

      a.     RAPs.   The RAP process consists of three  stages:  problem
 definition and identification of sources and causes of contamination;
 identification of remedial and preventive actions designed  to restore uses;

-------
22     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

and implementation of the RAP and the actual restoration of beneficial  uses.
RAPs are: currently under development in: all 26 U.S., 12 Canadian and five
binational AOCs.  To date, 40 of the 43 AOCs have established stakeholder
groups, coordinating committees, public advisory councils or some other forum
representative of local societal, economic, and environmental concerns.  These
infrastructures help to facilitate public participation in the RAP process, as
well as to coordinate RAP development, implement an ecosystem approach, and
build the institutional capacity to restore beneficial uses.

      Each RAP is tailored to address the specific pollutant problems that are
or may be present at the AOC.  The Grand Calumet River/Indiana Harbor Ship
Canal and nearshore Lake Michigan, for example, is the only AOC in which all
14 beneficial uses are impaired.  Several of these use impairments are  caused
by the presence of millions to cubic meters of contaminated sediments in the
waterway.  RAP efforts to eliminate the causes of these use impairments
include projects designed to clean up individual segments of the waterway and
to prevent their recontamination.  The RAP process has developed and obtained
funds for a Toxic Pollution Prevention Program on the waterway and the  Gary,
Hammond and East Chicago Sanitary Districts have formally adopted the RAP's
Common Policy on Toxic Pollution Prevention.

      Restoration of impaired uses in the Grand Calumet River\Indiana Harbor
Ship Canal has progressed mostly by implementing cleanup, pollution
prevention, and restoration projects outside of the waterway proper.  The RAP
process has involved the Indiana Department of Environmental Management's
(IDEM) pollution prevention staff, local industry and the general public in
implementing a Household Hazardous Waste Collection Project, which began in
April of 1994.  Agricultural clean sweeps to collect and properly dispose of
pesticides were also conducted in 1992 and 1993 in several Northern Indiana
counties.  Local educators have helped IDEM develop an Enviromobile which
stops at area schools to educate school children on ways to prevent pollution
while increasing their environmental consciousness.  All of these projects
directly reduce the causes of several use impairments in the AOC, and some
will help to preserve and protect the AOC's globally endangered habitats and
their outstanding biodiversity which is among the highest in the Great  Lakes.

      The Niagara River AOC is located in Erie and Niagara Counties in  western
New York.  Past municipal and industrial discharges and waste disposal  sites
have been a source of contaminants to the Niagara River.  A long history of
development has also changed the Niagara shoreline along much of the river,
affecting fish and wildlife habitat.  Habitat impairment and survival of
aquatic life in the.AOC have been attributed to PCBs, mirex, chlordane,
dioxin, dibenzofuran, hexachlorocyclohexane, polyaromatic hydrocarbons  (PAHs),
and other pesticides.  Contaminated sediments contain metals and cyanides and
are a source of use impairments  to the AOC.

      The Niagara River RAP represents a comprehensive and focused corrective
action strategy to: remediate contaminated sediments and hazardous waste
sites; continue and enhance monitoring activities; continue point and nonpoint
source control programs; and improve fish and wildlife habitat.  In
particular, EPA and the New York State Department of Environmental
Conservation (NYSDEC) are overseeing remediation at three sites along the
Niagara River that are considered sources of contaminants causing use
impairments in the river.  Other efforts currently being pursued under  this
RAP process include stream water quality monitoring to estimate pollutant
loadings, investigations of inactive hazardous waste sites, developing  a
nonpoint source loading estimation methodology for surface runoff, groundwater
migration and atmospheric deposition in conjunction with Environment Canada
and the Ontario Ministry of the  Environment and Energy  (OMEE), developing a
water quality enhancement and protection policy to include discharge
restriction categories, antidegradation and substance bans; working with the
Buffalo Sewer Authority to continue the model development process for system
sub-basins; and developing a comprehensive  inventory of fish and wildlife
populations and habitat.  Pursuant to the Niagara River RAP, a. large number of

-------
                                Section I: Background                             23
sediment remediations/removals, waste site remediations and enforcement
actions on the U.S. side of the Niagara River have resulted in significant
reductions (i.e., 98 percent of priority pollutants) of loadings to the river.
For further examples of RAP activities, see "Progress in Great Lakes Remedial
Action Plans"  (EPA 905-R-24-020, 1994) available in the docket for the final
Guidance.

      b.    LaMPs.  The LaMP process integrates Federal, State and local
programs that address loadings of critical pollutants, assess whether these
programs ensure attainment of beneficial uses, and recommend media-specific
program enhancements to reduce loadings of critical pollutants to the open
waters of the Great Lakes as necessary to attain beneficial uses.  LaMPs also
address pollutants which might impair waters that currently attain beneficial
uses.

      As part of the Lake Michigan LaMP process, the Lake Michigan States
 (Wisconsin, Illinois, Indiana, Michigan) are evaluating the beneficial use
impairments due to all stressors, including toxics, habitat quantity and
quality, exotic species, and human influences.  Multi-media pollution
prevention projects have recently been completed or are ongoing in greater
Chicago, western Michigan, and Milwaukee.  Sediment assessment and remediation
projects are proceeding in Illinois, Michigan and Indiana.  Agricultural
"clean sweeps" to properly collect and dispose of unused pesticides are
continuing in Indiana, Michigan and Wisconsin, and urban "clean sweeps" are
taking place in Northwest Indiana.  The Lake Michigan Forum (i.e., the public
participation component of the Lake Michigan LaMP), along with the Federal and
State governments, also continues to foster public participation.  The Forum
has finalized an action plan for enhancing public outreach, education, and
participation in the LaMP process.

      Implementation of the Lake Michigan Enhanced Monitoring and Mass Balance
Project is also proceeding in the Lake Michigan basin. This project is a
cooperative effort among EPA, the U.S. Geological Survey, U.S. Fish and
Wildlife Service, National Oceanic and Atmospheric Administration, National
Biological Survey and the four Lake Michigan States, and supports the Lake
Michigan LaMP process, GLTRE, and CAAA requirements by determining the
relative loadings of several LaMP Critical Pollutants from both air and water
sources to the Lake.  Modeling these pollutant inputs, and determining their
fate and transport, will allow the participating agencies to predict Lake
Michigan system responses to various load reduction alternatives as well as to
target limited resources to those pollutant sources posing the greatest risk
to the health of the Lake's ecosystem, and will provide important basinwide
information for use in other Great Lakes efforts.  A revised Lake Michigan
LaMP will be published in 1995.

      On February 15, 1994, EPA published a notice of availability in the
Federal Register of•the proposed Stage 1 Lake Superior LaMP (50 FR 7252).  The
draft Lake Superior LaMP puts forward a framework for zero discharge that will
further the goals of virtual elimination in the Lake Superior Basin.  EPA and
the Lake Superior Basin States have developed a Pollution Prevention Strategy
that focuses on all sources  (i.e., point and nonpoint) of the nine zero
discharge pollutants  (PCBs, dieldrin, chlordane, DDT and metabolites, mercury,
dioxin (2,3,7,8-TCDD), toxaphene, hexachlorobenzene, and octachlorostyrene),
and that identifies pollution prevention opportunities, implementation
methods, and measures of success.  Participating agencies have agreed to
develop a Binational Pollution Prevention Strategy, using the EPA's National
Pollution Prevention Strategy as a template.  A final Lake Superior LaMP will
be complete in 1995*.

      The workplan for the Lake Ontario Lakewide Management Plan for Critical
Pollutants was finalized in November 1993.  Target dates include the
completion of the development of this LaMP in three stages: Stage I is
anticipated to be completed in April 1995; Stage II by December 1995; and
Stage III will be complete by the end of December 1996.  Also, a work group

-------
24	Water Quality Guidance for the Great Lakes System - Supplementary Information Document

has been formed to develop a Lake Ontario LaMP Public Forum.  The Coordination
Committee  (water division managers frofli EPA, NYSDEC, Environment Canada and
the OMEE) and Secretariat  (technical staff from these same agencies) members
are discussing options for the LaMP management structure and incorporating
other agencies into the LaMP process when dealing with issues beyond Critical
Pollutants.

      In the spring of 1993, EPA and Environment Canada formed a temporary
Binational Implementation Committee to develop a concept .paper that would
provide a framework for the development of the Lake Erie LaMP, and to organize
a LaMP Management Committee to oversee LaMP development.  Representatives of
the U.S. and Canadian Federal, State, and Provincial governments met in
September 1994 to formally convene a binational Management Committee to
oversee development of the Lake Erie LaMP.  The Management Committee approved
the Lake Erie LaMP concept paper, which describes the proposed focus, scope,
and management structure of the Lake Erie LaMP.  As outlined in the concept
paper, the Lake Erie LaMP will address a number of environmental issues above
and beyond toxic substances, including habitat destruction, exotics, and
nutrient loadings.  The Management Committee also charged a staff-level work
group to initiate a variety of LaMP activities.

      In addition, EPA expects any new loadings data obtained during the LaMP
process will be incorporated by the States and Tribes when establishing or
revising TMDLs and wasteload allocations  (WLAs) in the Great Lakes System.
These new TMDLs and. WLAs will then be appropriately reflected in subsequent
revisions to NPDES permits.

      5.    Sediments

      Contaminated sediments are a significant source of loadings of toxic
pollutants at harbors and river mouths throughout the Great Lakes System.  To
address this source of toxic pollutant loadings, the 1987 CWA Amendments
authorized a five-year demonstration program to identify and develop
assessment and treatment technologies for contaminated sediments in the Great
Lakes basin (CWA section 118(c) (7)) .  This program, known as the Assessment
and Remediation of Contaminated Sediments  (ARCS) Program, was designed  to
evaluate appropriate assessment and treatment methodologies for the cleanup of
toxic pollutants in Great Lakes contaminated sediments.  The information
developed through the ARCS Program will help support implementation of  RAPs by
providing guidance to effectively address the contaminated sediment problems
in designated AOCs.

      Contaminated sediments have been identified as environmental problems in
42 of the 43 AOCs.  The 1987 CWA Amendments also specified five AOCs for
priority consideration in conducting the demonstration projects.  These AOCs
are Saginaw Bay, Michigan; Sheboygan Harbor, Wisconsin; Grand Calumet River,
Indiana; Ashtabula River, Ohio; and Buffalo River, New York.   (Id.)

      EPA successfully completed the ARCS Program in 1994.  Approximately 40
documents will be published by early 1995 which discuss the findings of the
sediment assessment work, risk assessment studies, mass balance modeling
results, and results of the pilot demonstrations.  Three guidance documents on
risk assessment, sediment assessment, and treatment technologies have been
published.  Technology transfer workshops, which covered such topics as
sediment assessments, mass balance modeling, and determining contaminant
losses from dredging and disposal projects, have been held at a  number  of
locations throughout the Great Lakes basin.

      A key component of the ARCS Program was  the public outreach and
involvement that was maintained throughout the demonstration process.
Information on ARCS Program  activities has been widely distributed  to the
public in the form of ARCS Update fact sheets, news releases, a  slide show,
and public meetings.  The ARCS Program has also made all written documentation

-------
                                Section I: Background                             25
accessible by setting up repositories for ARCS Program materials in local
libraries at the five ARCS priority locations.

      The final summary report on ARCS program results was presented to
Congress in the fall of 1994 and is available in the docket for the final
Guidance.  The conclusions contained in the final summary report discuss the
state-of-the-art methods the ARCS Program demonstrated for the assessment of
contaminated Sediments, especially in the area of toxicity testing, and the
new ground broken by the ARCS Program in the application of the mass balance
modeling approach.  The ARCS Program made significant contributions to the
knowledge base on contaminated sediment remediation by selecting promising
treatment technologies, taking them out into the field, and demonstrating
their effectiveness on site.

      The major findings of the ARCS Program consisted of the need to perform
thorough, integrated sediment assessments; the importance of mass balance
modeling in the evaluation of remediation scenarios; the identification and
demonstration of several feasible sediment treatment technologies; and the
recognition and success of public involvement and active participation in
sediment assessment and remediation projects.  The field assessment,
contaminant fate modeling, risk assessment, and remediation technology
techniques demonstrated in the ARCS Program have improved the knowledge base
and will enable Federal, State and local officials to make scientifically
sound remediation decisions.

      The products of the ARCS Program will not, by themselves, eliminate the
problems posed by contaminated sediments, nor do they propose one "cure all"
treatment technology for their remediation.  They do, however, provide
guidance for the selection of sediment assessment and treatment technologies
as well as recommendations for future full-scale applications.  The results
and products from the ARCS Program will have far-reaching implications for the
remediation of contaminated sediments within the Great Lakes as well as on a
nationwide basis.

      At the time the proposed Guidance was published, EPA was also preparing
to publish for public comment proposed sediment quality criteria for five
organic chemicals.  EPA's proposed sediment quality criteria are intended to
protect benthic organisms from unacceptable effects of chemicals associated
with sediments.  The proposed sediment quality criteria, however, do not
protect against additive, synergistic or antagonistic effects of contaminants,
or bioaccumulative effects to aquatic life or humans.  Implementation of the
final Guidance will complement the proposed sediment quality criteria in these
areas.

      In January 1994, EPA announced in the Federal Register the availability
of the draft sediment quality criteria for three polycyclic hydrocarbons
(acenaphthene, fluoranthene and phenanthrene) and two pesticides (endrin and
dieldrin).  The public comment period for these criteria documents ended in
June 1994, and EPA is now in the process of developing final sediment
criteria.

      At the time of the proposal, EPA was also in the process of developing a
proposed National Contaminated Sediment Management Strategy.  On August 30,
1994, EPA announced in the Federal Register the availability of the proposed
comprehensive, multimedia Strategy.  The proposed Strategy describes specific
actions that EPA will take to reduce environmental and human health risks
associated with contaminated sediments.  The Strategy also defines a plan of
action to address the National problem of contaminated sediments and to
streamline decision-making within and among EPA programs.  EPA accepted
comments on the Strategy through October 31, 1994, and is currently reviewing
the comments.   The proposed Strategy is available in the docket for the final
Guidance.

-------
26     Water Quality Guidance for the Great Lakes System - Supplementary Information Document

      Implementation of EPA's National Strategy has been proceeding in  the
Great Lakes basin, with emphasis on the Great Lakes AOCs.  Programs in  EPA and
the States have been assessing sediment contamination and utilizing a rangg of
regulatory tools to' address the most contaminated sites.  Some examples of
remediation activity are the sediment cleanups that have taken place at
Waukegan Harbor in Illinois, the Black River in Ohio, Cedar Creek near
Milwaukee, Wisconsin, and in East Chicago, Indiana.  Settlements have also
been reached in enforcement actions with responsible parties for sediment
cleanup for sites in Gary and East Chicago, Indiana.  Among the actions being
pursued are those for: Manistique River/Harbor in Manistique, Michigan; the
River Raisin near Monroe, Michigan; Ashtabula River and Fields Brook near
Ashtabula, Ohio; Grand Calumet River in Hammond, Indiana; and Sheboygan
River/Harbor near Sheboygan, Wisconsin.

E.    Science Advisory Board Review

      Information concerning EPA's Science Advisory Board  (SAB) and their
review of the draft Water Quality Guidance for the Great Lakes System prior to
proposal is contained in the preamble to the proposed Guidance  (58 FR at
20826) .  During the process of developing the final Guidance, EPA met with the
SAB from April 27-29, 1994, to discuss the development of a National
methodology for developing wildlife criteria and bioaccumulation factors.  EPA
considered and addressed the SAB's comments pertaining to the Great Lakes
System in preparing the final Guidance.  Some of these comments are discussed
in the applicable sections of this document that follow.  The final SAB report
is also included in the docket for the final Guidance.

F.    References

      Colborn, T.E., Davidson, A., Green S.N., Hodge, R.A., Jackson, C.I., and
Liroff, R.A.  1990.  Great Lakes, Great Legacy?  Washington, D.C.: The
Conservation Foundation; and Ottawa, Ontario: The Institute for Research on
Public policy; 174.

      Colborn, T.I.  1991.  Epidemiology of Great Lakes bald eagles.  J.
Environ. Health Toxicol. 4:395-453.

      Dar, E., Kanarek, M., Anderson, H., and Sonzogni, W.  1992.  Fish
consumption and reproductive outcomes  in Green Bay, Wisconsin.  Environ. Res.
59: 189-201.

      DeVault, D.S., W.A. Willford, R.J. Hesselberg, D.A. Nortrupt, D.A.
Rundberg, A.K. Alwan, and C. Bautista.  1986.  Contaminant trends  in lake
trout  (Salvelinus namaycush) from the upper Great Lakes. Arch. Environ.
Contain. Toxicol. 15:349-356.

      DeVault, D.S., J.M. Clark, G. Lahvis, and J. Weishaar.  1988.
Contaminants and trends  in fall run coho salmon.  J. Great Lakes Res.   14:23-
33.

      DeVault, D.S., USEPA, GLNPO, Personal Communications.

      DeVault, D.S.  1993a.  Data on  contaminant trends  in Lake  Trout
 (Unpublished).  Contains data for 1984-1990.

      DeVault, D.S.  1993b.  Data on  contaminant trends  in fall  run  coho
salmon  (Unpublished).   Contains data  for 1986-1990.

      DeVault, David S., and R. Hesselberg.   1994.   Contaminant  trends  in lake
trout and walleye from  the St. Lawrence Great Lakes.   (Final Draft)

      Environment Canada.   1991.  Toxic Chemicals  in the Great  Lakes and
Associated  Effects", Vols.  I and II.   Environment Canada,  Department  of
Fisheries and Oceans, Health and  Welfare  Canada.

-------
                                Section I: Background                             27
      Falck, F., A. Ricci, M. Wolff, J. Godbold, and P. Deckers.   1992.
Pesticides and polychlorinated biphenyl residues in human breast  lipids  and
their relation to breast cancer.  Arch. Environ. Health 47(2):  143-146.

      Fein, G.G., J.L. Jacobson, S.W. Jacobson, P.M. Schwartz,  and J.K.
Dowler.  1984b.  Prenatal exposure to polychlorinated biphenyls:  Effects on
birth size and gestational age.  J. Pediatr. 105(2): 315-320.

      Flint, R.W. and J. Vena  (eds).  1991.  Human Health Risks from Chemical
Exposure: The Great Lakes Ecosystem.  Chelsea, Michigan:  Lewis Publishers,
Inc.

      Fox, G.A.  1991.  Practical causal inference for epidemiologists.  J.
Toxicol. Env. Health.  33: 359-373.

      Fox, G. A., B. Collins, E. Hayakawa, D.V. Weseloh, J.P. Ludwig, T.J.
Kubiak and T.C Erdman.  1991.  Reproductive outcomes in colonial  fish-eating
birds: A biomarker for developmental toxicants in Great Lakes food chains.
II.  Spatial variation in the occurrence and prevalence of bill defects  in
young double-crested cormorants in the Great Lakes, 197901987.  J. Great Lakes
Res: 17(2): 158-167.

      Fox, G.A.  1992.  In: Chemically-Induced Alterations in Sexual and
Functional Development: The Wildlife/Human Connection  (Colborn, T., and
Clement, C., eds.).  Princeton, NJ: Princeton Scientific Publishing Co.,  Inc.,
1992; 147-158.

      Giesy, J.P., J.P. Ludwig, and D.E. Tillitt.   (In press).  In: Dioxins
and Health.  A. Scheckter (Ed).  Penum: New York.

      Giesy, J.P., J.P. Ludwig, and D.E. Tillitt.  1994.  Deformities in birds
of the Great Lakes region: Assigning causality.  Environ. Sci.  Technol.,
28(3): 128A-135A.

      Gilbertson, M., T. Kubiak, J. Ludwig, and G. Fox.  1991.  Great Lakes
embryo mortality, edema, and deformities syndrome  (GLEMEDS) in  colonial  fish-
eating birds: similarity to chick-edema disease.  J. Toxicol. Environ. Health,
33: 455-520.

      International Joint Commission.  1991.  Cleaning Up the Great Lakes.
Windsor, Ontario: International Joint Commission.

      Korach, K.S., P. Sarver, K. Chae, J.A. McLachlan, and J.D. McKinney.
1988.  Estrogen receptor binding activity of polychlorinated hydroxybiphenyls:
conformationally restricted structural probes.  Mol. Phar. 33:120-126.

      Kubiak, T.J., H.J. Harris, L.M. Smith, T.R. Schwartz, D.L.  Stalling,
J.A. Trick, L. Sileo, D.E. Docherty, and T.C. Erdman.  1989.
Microcontaminants and reproductive impairment of the Forster's  tern on Green
Bay, Lake Michigan-:1983.  Arch. Environ. Contain. Toxicol.  18: 706-727.

      Kurita, H. and J.P. Ludwig.  1988.  Embryonic teratologies and
abnormalities assessed in naturally-incubated eggs and double-crested
cormorants (Phalacrocorax auritus) and Caspian terns (Hydroprogne  caspia) from
Michigan Great Lakes colonies in 1988.  Reports to the Michigan Audubon
Society on the 1986-1988 Findings of the Michigan Colonial Waterbird
Monitoring Project.  Dec. 15, 1988, Ecological Research Services,  Inc.

      Manz, A., J. Berger, J.H. Dwyer, D. Flesch-Janys, S. Nagel,  and H.
Waltsgott.  1991.  Cancer mortality among workers in chemical plant
contaminated with dioxin.  Lancet. 338:959-964.

      McLachlan, J.A., R. Newbold, K.S. Korach, and M. Hogan.   1987.  Risk
assessment considerations for reproductive and developmental toxicity of

-------
28     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

oestrogenic xenobiotics.  In: Human Risk Assessment: The Roles of Animal
Selection and Extrapolation  (Roloff, M.V. and A.W. Wilson, eds.).  London:
Taylor and Francis, Ltd., 1987: 187-193.

      Mitchell, C.A. and T.W. Custer.  1986.  Hatching success of Caspian
terns nesting in the lower Laguna Madre, Texas, USA.  Colonial Waterbirds.   9:
86-89..

      Mora, M.A. , H.J. Auman, J.P. Ludwig, J.P. Giesy, D.A- Verbrugge,  and
M.E. Ludwig.  1993.  Polychlorinated biphenyls and chlorinated insecticides  in
plasma of Caspian terns: relationships with age, productivity, and colony site
tenacity in the Great Lakes.  Arch. Environ. Contam. Toxicol.  24: 320-331.

      Nature Conservancy Great Lakes Program.  1994.  The Conservation  of
Biological Diversity in the Great Lakes Ecosystem: Issues and Opportunities.
Chicago, Illinois.

      Soikkeli, M. -1973.  Breeding success of the Caspian tern  in Finland.
Bird-Banding.  44:196-204.

      Stevens, R.J., and M.A. Neilson.  1989.  Inter- and intralake
distributions of trace organic contaminants in surfacewaters of  the Great
Lakes.  J. Great Lakes Res. 15(3): 377-393.

      Swain, W.R.  1991.  Effects of organchlorine chemicals on  the
reproductive outcomes of humans who consumed contaminated Great  lakes fish:  an
epidemiologic consideration.  J. Toxicol. Environ. Health 33: 587-639.

      Tanabe, S.  1988.  PCB problems in the future: foresight from current
knowledge.  Environmental Pollution  (50) 5-28.

      Tillitt, D.E., G.T. Ankley, J.P. Giesy, J.P. Ludwig, H. Kurita-Matsuba,
D.V. Weseloh, P.S. Ross, C.A. Bishop, L. Sileo, K.L. Stromborg,  J. Larson, and
T. Kubiak.  1992.  Polychlorinated biphenyl residues and egg mortality  in
double-crested cormorants from the Great Lakes.  Env. Toxicol. Chem. 11: 1281-
1288.

      U.S. EPA.  1994.  Assessment and Remediation of Contaminated Sediments
(ARCS) Program, EPA. 905-S-94-001.

      U.S. EPA.  1994.  Deposition of Air Pollutants to the Great Waters:
First Report to Congress, EPA 453-R-93-055.

      U.S. EPA.  1994.  EPA's Contaminated Sediment Management Strategy, EPA
823-R-94-001.

      U.S. EPA.  1994.  Progress  in Great Lakes Remedial Action  Plans,  EPA
905-R-24-020.

      Yamashita, N.', S. Tanabe, J.P. Ludwig, H. Kurita, M.E. Ludwig, and R.
Tatsukawa.  1993.  Embryonic abnormalities and organochlorine contamination  in
double-crested cormorants  (Phalacrocorax Auritus) and Caspian terns
(Hydroprogne caspia) from the upper Great Lakes in 1988.  Environ. Poll.
79:163-173.

-------
                          Section II: Regulatory Requirements                        29
          H.   REGULATORY REQUIREMENTS


A.    Scope and Purpose

      Section 118(c)(2) of the Clean Water Act  ("CWA")  (Pub. L. 92-500 as
amended by the Great Lakes Critical Programs Act  of  1990, Pub. L. 101-596)
requires EPA to publish final water quality guidance on minimum water quality
standards, antidegradation policies,  and implementation procedures for the
Great Lakes System.  Part 132, including appendixes  A through F, constitutes
the Water Quality Guidance for the Great Lakes  System required by section
118 (c) (2).

      The Guidance fulfills the requirements of section 118(c)(2) and is
generally organized as follows:

          Water quality standards; The Guidance contains numerical water
quality criteria for 29 pollutants, listed in Tables 1 through 4 of part 132.
The Guidance also contains methodologies for the  development of water quality
criteria and water quality values for the  protection of aquatic life, human
health,  and wildlife,  and a methodology for development of bioaccumulation
factors, in appendixes A through D.  Together,  these criteria and
methodologies specify minimum numerical limits  on pollutants in ambient Great
Lakes waters to protect human health,  aquatic life,  and wildlife.

          Antidearadation policy; The minimum antidegradation policy,
including an antidegradation standard,  implementation procedures,
demonstration provisions, and decision provisions are contained in appendix E.

          Implementation procedures:  Minimum implementation procedures are
contained in appendix F to part 132.

B.    Definitions

      1.    Proposal.   Section 132.2  of the proposed Guidance contained
definitions of 71 terms that would apply to part  132.  EPA proposed the
definitions as a partial list of terms which need to be defined for consistent
interpretation of the Guidance.  Proposed § 132.4(a) provided that Great Lakes
States and Tribes were to adopt requirements applicable to waters of the Great
Lakes System for the purposes of sections  118,  303,  and 402 of the CWA that
are consistent with these definitions,  or  be subject to EPA promulgation under
§ 132.5.  Other definitions, bearing on the individual portions of the
Guidance,  were contained in proposed appendixes A through F.  Generally, where
terms were applied in the proposed Guidance in  the same manner as in previous
National regulations,  such as those in 40  CFR 122.2, 130.2, and 131.3, the
proposal did not provide a duplicate  definition.  In some cases, however, the
proposal provided a duplicate definition to assist the reader.

      2.    Comments.   Several comments recommended  that EPA resolve
differences between definitions proposed in § 132.2  and those proposed in
various  appendixes to part 132.

      A number of comments were received on definitions that are discussed
elsewhere in this document.  These include: the definition of "bioaccumulative
chemical of concern,11  discussed further in section II.C.8 of this document;
and the  definition of "existing discharger," discussed further in section
VTII.I of this document.

      One comment recommended deleting the definition of "reasonable
potential."  EPA agrees.  Since the definition  of "reasonable potential" is
not necessary to implement the Guidance,  EPA has  deleted the definition.

-------
30     Water Quality Guidance for the Great Lakes System - Supplementary Information Document

      Technical and editorial comments were also received concerning  a  number
of definitions.

      EPA agrees that definitions should be consistent throughout  the
Guidance.  In some cases, as discussed below under the Final Guidance section,
definitions have been modified or deleted from § 132.2 to eliminate such
inconsistencies.

      EPA agrees with the majority of technical and editorial comments
received, and has made changes accordingly.  Some of these changes are
discussed further under the Final Guidance section below.  EPA's responses  to
individual technical and editorial comments on definitions in §  132.2 are
included in the docket for this rulemaking.

      3.    Final Guidance.  Section 132.2 of the final Guidance includes 56
definitions, reflecting deletion of 17 proposed definitions, and addition of
two new definitions.  Thirty six definitions were clarified or modified, and
18 were unchanged.

      The following definitions have been deleted from § 132.2 because  they
are defined adequately in one or more of the appendixes: "acceptable  daily
exposure (ADE)," "acute toxic unit (TU.) , "  "biomagnification,"  "chronic toxic
unit  (TU0) ,"  "depuration,"  "dilution  fraction,"  "linear multi-stage model,"
11 octanol/water partition coefficient  (KQW) , " "relative source contribution
(RSC)," and "slope factor."

      The definition of "allowable dilution flow  (Q^)" has been deleted from
§ 132.2.  In response to comments, changes have been made to the final
procedure 3* of appendix F that made this definition unnecessary.   This  change
is discussed further in section VIII.C of this document.

      The definition of "compliance evaluation level  (CEL)" has been  deleted
because the term is no longer used.  This change is discussed further in
section VIII.H of this document.

      The definition of "existing uses" has been deleted to because it
duplicates the definition in 40 CFR 131.3.

      EPA deleted the proposed definition of "noncarcinogen" because  it should
be clear that a noncarcinogen is any substance which is not a carcinogen.
Furthermore, EPA found that the proposed definition could have been confusing
when  compared to the definition of "carcinogen."

      The definition of "reasonable potential" has been deleted  from  §  132.2
as discussed in Comments above.

      The definitions of "steady-state BAF/BCF" and "superlipophilic
chemicals" were deleted, since revisions to the methodology for  development of
BAFs made the terms unnecessary.  These changes are discussed further in
section IV of this document.

      The definition of "biota-sediment accumulation factor  (BSAF)" has been
added, and is discussed in section IV of this document.

      The definition of "new Great Lakes discharger" has been added,  and the
definition of "existing discharger" has been modified and redefined as
"existing Great Lakes discharger."  Together, the two definitions  categorize
all discharges to waters of the Great Lakes System depending on  whether the
construction of buildings, structures, facilities, or  installations generating
the discharge commenced before  ("existing") or after  ("new") March 23,  1997.
Under procedure 3 of appendix F, mixing zones for BCCs are not available to
new Great Lakes dischargers after March 23, 1997; for  existing Great  Lakes
dischargers, mixing zones  for BCCs are generally not available after  March  23,
2007  except under specified circumstances.  These provisions are discussed

-------
                           Section II: Regulatory Requirements                        31
more fully in sectipn VTII.C of this document.  Under procedure 9 of
appendix F, NPDES permits may allow compliance schedules under certain
circumstances to existing Great Lakes dischargers but not to new Great Lakes
dischargers.  These provisions are discussed more fully in section VIII.J of
this document.

      .The definition of "bioaccumulative chemical of concern  (BCC)" has been
modified, and is discussed in section II.C of this document.

      The definition of "carcinogen" has been clarified to add a reference to
a discussion of the classification of carcinogens in section II.A of appendix
C to the final Guidance.

      The definition of "chronic toxicity" has been clarified to add reference
to "delayed adverse effects" to reflect EPA's concern that pollutants such as
2,3,7,8-tetrachlorodibenzo-p-dioxin (2378-TCDD) have been observed to cause
delayed adverse effects days, weeks, or even months after exposure.  See, for
example, "Interim Report on Data and Methods for Assessment of 2378-TCDD Risks
to Aquatic Life and Associated Wildlife," EPA, March 1993  (EPA 600/R-93/055),
which is available in the docket for this rulemaking.

      The definition of "Great Lakes States and Great Lakes Tribes" has been
modified to clarify that an Indian Tribe in the Great Lakes basin would be a
"Great Lakes Tribe" subject to the provisions of the final Guidance only if
EPA has approved water quality standards for the Tribe, or if EPA has
authorized the Tribe to administer an NPDES program.  This clarification was
necessary because the Guidance provisions for water quality standards would be
meaningless unless at least a basic set of standards are in place, including
designated uses, narrative criteria, and legal authority references.
Similarly, the Guidance provisions for NPDES permits would be meaningless
unless the Tribe has been authorized to administer an NPDES program.
Provisions applicable to Tribes are discussed further in section II.D.3 of
this document.

      The definition of "Tier II value" has been modified, and is discussed in
sections II.C.2 and VT of this document.

      The definition of "wet weather point source" has been modified to
incorporate the existing definition of point source contained in § 122.2, for
the convenience of the reader.  Furthermore, in response to comments and to
clarify possible ambiguities in the proposal, EPA has modified the definition
in the final Guidance to clarify that the definition includes only certain
types of discharges, to specify the types of excluded discharges more
explicitly, and to delete an unnecessary definition of combined sewer
overflow.  The exclusion for wet weather point sources is discussed in section
II.C.7.

      The following definitions were modified to make technical clarifications
and corrections: "acute toxicity," "acute-chronic ratio (ACR),"
"bioaccumulation," "bioaccumulation factor  (BAF)," "bioconcentration,"
"bioconcentration factor (BCF)," "chronic toxicity," "criterion continuous
concentration (CCC)," "criterion maximum concentration (CMC)," "EC50,"
"endangered or threatened species," "final acute value (FAV)," "final chronic
value (FCV)," "final plant value (FPV)," "Great Lakes System," "human cancer
value (HCV)," "human noncancer value  (HNV)," "LC50," "minimum level  (ML)," "no
observed adverse effect level (NOAEL)," "no observed effect concentration
(NOEC)," "quantification level," "quantitative structure activity relationship
(QSAR)," "risk associated dose (RAD)," "species mean acute value  (SMAV),"
"species mean chronic value  (SMCV)," "stream design flow," "uncertainty factor
(UF)," "uptake," and "wasteload allocation  (WLA)."

      The following- definitions are the same as proposed,  or the same with
editorial corrections: "adverse effect," "connecting channels of the Great
Lakes," "detection level," "Federal Indian Reservation," "genus mean acute

-------
32     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

value (GMAV)," "genus mean chronic value  (GMCV)," "Great Lakes," "human cancer
criterion (HCC)," "human noncancer criterion (HNC)," "Indian Tribe," "load
allocation  (LA)," "loading capacity," "lowest observed adverse effect level
(LOAEL)," "open waters of the Great Lakes," "threshold effect," "Tier I
criteria,"  "total maximum daily load (TMDL)," and "tributaries of the Great
Lakes System."

C.    Adoption-and Application of Criteria. Methodologies. Policies, and
      Procedures

1.    Adoption of Tier I Criteria and Methodologies

      a.     Proposal.  The proposed Guidance included a two-tiered approach,
consisting of methodologies to develop water quality criteria  (Tier I), and
methodologies to calculate water quality values (Tier II) with fewer data than
the full minimum data required for a Tier I criterion calculation.  The
purpose of Tier II methodologies is to provide a uniform approach for
evaluating and controlling pollutants when  there are insufficient data to
develop Tier I criteria.  In the absence of State- or Tribal-adopted Tier I
criteria for a pollutant, the Tier I and Tier II methodologies provide
mechanisms with which to interpret and ensure that the States' and Tribes'
narrative criteria prohibiting the discharge of toxic pollutants in toxic
amounts are reflected in water quality-based effluent limits.  The preamble to
the proposed Guidance described the two-tiered approach in more detail (58 FR
20835; April 16,  1993).

      Proposed § 132.4(a) provided that Great Lakes States and Tribes were to
adopt Tier  I methodologies for developing numeric water quality criteria to
protect aquatic life, human health, and wildlife consistent with those
specified in appendixes A through D, or be  subject to EPA promulgation under
§ 132.5.

      Upon State or Tribal adoption or EPA promulgation of the Guidance,
proposed §  132.4(b) specified that Great Lakes States or Tribes would have
used the Tier I methodologies in appendixes A, C, and D and the methodology in
appendix B when adopting or revising numeric water quality criteria.  In
addition, if a Great Lakes State or Tribe had not adopted a numeric water
quality criterion for a pollutant in its water quality standards, but enough
data existed to meet Tier I minimum data requirements, proposed § 132.4(c)
provided for use of the Tier I methodologies for any development of numerical
criteria to implement narrative criteria.   Such implementation would include
development of water quality-based effluent limits, where appropriate.

      The proposed Guidance also included numeric criteria in Tables 1 through
4 of part 132 that were derived using the Tier I methodologies.  Proposed
§ 132.3 would have required Great Lakes States and Tribes to adopt these
specific numeric criteria or more stringent criteria into their water quality
standards for the Great Lakes System or be  subject to EPA promulgation under
§ 132.5.

      b.    Comments.  EPA received a number of comments on the requirements
for adopting additional Tier I criteria beyond those listed in Tables 1
through 4 of part 132, and for revising the criteria in  the Tables.  Many
commenters  supported the proposed approach  because it would ensure  increased
consistency among the water quality standards of Great Lakes States and
Tribes.  Other commenters urged EPA to provide more flexibility to  States  and
Tribes in adopting criteria consistent with the Guidance, including the
flexibility to modify the criteria in Tables 1 through 4 to reflect new
scientific  findings and data.  Many commenters argued that the criteria
methodologies and the criteria in Tables  1  through 4 should be provided  as
non-binding guidance.

      A large number of comments  also suggested that EPA should be  responsible
for developing any new  or revised Tier I  criteria as additional data  become

-------
                           Section II: Regulatory Requirements                        33
available and are evaluated.  These commenters believe that it is more
efficient for EPA to develop criteria than for States or dischargers who may
not have the expertise or resources to develop criteria independently.  Some
of these commenters also suggested that EPA should ensure full public review
of such new or revised criteria.  Some commenters also urged EPA to amend
Tables 1 through 4 in future EPA rulemakings.  Additionally, a few commenters
were concerned that the Guidance as proposed may not provide sufficient
protection to endangered or threatened species from pollutants that may be
found to have particularly potent effects.
      EPA agrees that it is important for States and Tribes to have the
flexibility to modify criteria and values, including the criteria in Tables 1
through 4, in appropriate circumstances when new scientific findings and data
become available.  EPA has determined that the proposed minimum requirements,
together with changes to the procedure for site-specific criteria
modifications in the final Guidance, as well as EPA's planned approach to
assist States and Tribes in implementing the Guidance, provide adequate
flexibility for incorporating new information without sacrificing the improved
consistency in Great Lakes water quality standards envisioned in the Great
Lakes Critical Programs Act ("CPA").

      Site-specific criteria modifications provide a degree of flexibility to
incorporate new scientific findings and data as they may affect a specific
waterbody.  EPA has retained the provisions allowing site-specific
modifications to criteria developed under the Guidance, including criteria in
Tables 1 through 4, and has expanded the flexibility of the provisions to
allow less restrictive site-specific modifications for aquatic life, human
health and wildlife criteria under certain conditions using the final
procedure 1 of appendix F.  These provisions are discussed further in section
VTll.A of this document.

      Additionally,'there are several steps within the criteria methodologies
where flexibility is available to reflect new findings and data.  For example,
although EPA states a preference for using the linearized multistage model
when deriving human health criteria, the final Guidance allows the use of
different models if the data support their use.  In addition, the final
Guidance recognizes that the EPA methodology for conducting cancer risk
assessments is currently under review and that any changes adopted by EPA can
be incorporated by the States and Tribes.  Also, the section VIII.A of
appendix B provides that BAFs derived in accordance with the BAF methodology
in the final Guidance should be modified if changes are justified by available
data.

      EPA's planned approach for implementing the final Guidance will further
facilitate flexibility to incorporate new scientific findings and data.  EPA
intends to expand the future assistance it provides to Great Lakes States and
Tribes over that envisioned at the time of the proposal.  EPA's planned
implementation is as follows:

          EPA Region 5, in cooperation with EPA Regions 2 and 3 and
Headquarters offices,  and the Great Lakes States and Tribes, will establish a
Great Lakes Initiative (GLI)  Clearinghouse to assist States and Tribes in
developing numeric Tier I water quality criteria and Tier II water quality
values.  Further information about the GLI Clearinghouse is available from the
address listed at the beginning of this document.  As additional toxicological
data and exposure data become available or additional Tier I numeric criteria
and Tier II values are calculated by EPA, States, or Tribes, Region 5 will
ensure that this information is disseminated to the Great Lakes States and
Tribes.

            EPA Region 5 will work with the States and Tribes, EPA Regions 2
and 3,  EPA Headquarters offices, and EPA research laboratories to review new
toxicological and exposure data.  The review will include consideration of
data quality and appropriateness for use with the Guidance methodologies.  For
pollutants of especially high interest and/or concern, Region 5 and the other

-------
34	Water Quality Guidance for the Great Lakes System - Supplementary Information Document

EPA offices identified above intend from time-to-time to use the Clearinghouse
information to develop GLI criteria guidance documents similar to those
supporting the proposed and final Guidance.  EPA will then publish a notice in
the Federal Register announcing the availability of such documents and
inviting public comment on them.  After reviewing the comments, EPA will
finalize the GLI criteria guidance documents and make them available as
guidance to Great Lakes States and Tribes.  The GLI criteria guidance
documents would represent EPA's best current information about effects  of the
pollutants in the Great Lakes System.  The GLI criteria guidance documents
could address either or both Tier I criteria and Tier II values.

            Regions 2, 3 and 5, through their review and approval of State
water quality standards under section 303 of the CWA, will work with the
States to ensure that all Great Lakes States and Tribes maintain a minimum
consistent level of protection for aquatic life, human health, and wildlife
throughout the Great Lakes System consistent with the methodologies in  the
final Guidance.  EPA will also make a special effort to encourage Great Lakes
States and Great Lakes Tribes to adopt criteria based on any newly calculated
GLI criteria guidance documents in the next triennial review of water quality
standards under section 303.  If such efforts are not successful for all
States, EPA could evaluate use of section 303(c)(4), if appropriate, to
determine that a particular State or Tribe needed new or revised water  quality
standards reflecting the new criteria,  and promulgate Federal criteria  if
necessary for that State or Tribe to ensure minimum consistent criteria are
present in all States and Tribes.

      EPA believes the above plans will provide the flexibility to incorporate
new data into water quality criteria development.  For example, if new  data
become available that would result in significant changes to criteria in
Tables 1 through 4, EPA may use the above process to develop one or more
revised GLI criteria guidance documents.  EPA would then work with the  States
and Tribes in their adoption of the revised criteria.  If the revised criteria
are more stringent than the corresponding criteria in Tables 1 through  4 of
part 132, States and Tribes would be able to adopt them without further EPA
rulemaking.  If the'revised criteria are less stringent than the corresponding
criteria in Tables 1 through 4 of part 132, EPA would consider initiating a
rulemaking action to delete or revise criteria in the Tables if necessary to
allow or facilitate State and Tribal adoption of the less stringent criteria.

      It is also possible that the new information could be so substantial
that it makes some criteria in Tables 1 through 4 scientifically indefensible.
In this situation, States or Tribes could utilize the provisions in § 132.4(h)
to adopt new criteria even if they were less stringent than criteria in Tables
1 through 4 without further EPA rulemaking.  EPA expects that this situation
would occur rarely,.if at all.

      The above plans will also result in an efficient use of resources to
develop additional criteria and values where they are most needed.
Furthermore, this approach provides more flexibility than the alternative of
EPA conducting a rulemaking every time any additional data could support a
modification, to a criterion in Tables 1 through 4.  Finally, the final
Guidance retains the existing State and Tribal responsibility to adopt  water
quality standards as necessary under the CWA without waiting for EPA to
develop detailed criteria guidance.  At the same time, however, the
involvement of EPA in facilitating the review of data, developing GLI criteria
guidance documents as appropriate, in conjunction with continued oversight of
the section 303 water quality standards programs will contribute to more
consistent criteria in the future throughout the Great Lakes System.

      EPA considered but rejected the suggestion of some commenters that the
criteria methodologies and the numeric criteria in Tables 1 through 4 of part
132 should be provided as guidance but not  specified as minimum requirements
for State and Tribal adoption.  First, EPA  does not believe that the suggested
approach would satisfy the requirements of  section  118(c)(2).  That section

-------
                           Section II: Regulatory Requirements                        35
not only directs EP£ to provide guidance to the Great Lakes States on
"minimum" water quality standards, antidegradation policies, and
implementation procedures for the Great Lakes System, but also requires the
States to adopt provisions consistent with the guidance or be subject to EPA
promulgation.  EPA believes that whether States and Tribes adopt standards,
policies, and procedures consistent with the final Guidance, or whether EPA
promulgates them, Congress intended that the final Guidance would establish
minimum requirements for the Great Lakes System.  This issue is discussed
further in section II.D.2 of this document.

      Second, the suggested approach could result in significant variations
between State and Tribal programs submitted under section 118(c)(2)(C).  As
discussed in the preamble to the proposed Guidance, the establishment of more
uniform control of water pollution in the Great Lakes System was one of the
most important goals of the CPA, and the 1986 Great Lakes Toxic Substances
Control Agreement signed by the eight Great Lakes Governors (58 PR 20820-23,
and 20838-39).  Progress toward this goal will not be achieved unless the
Guidance specifies the "minimum" requirement for the Great Lakes System.

      Similarly, EPA considered but rejected an alternative approach that
would retain the provisions for States and Tribes to adopt criteria consistent
with the criteria in Tables 1 through 4, but would provide the methodologies
in appendixes A through D as guidance but not specified as minimum
requirements for State and Tribal adoption.  EPA believes the criteria in
Tables 1 through 4 are necessary to protect human health, wildlife, and
aquatic life in the Great Lakes System, but is concerned that these criteria
alone would not be sufficient.  The pollutants in Tables 1 through 4 are good
examples of pollutants in the Great Lakes System, but they are clearly not the
only criteria necessary to protect human health, wildlife and aquatic life in
the Great Lakes System.  For example, in sections I and II of this document
EPA identifies BCCs as a category of pollutants of high concern in the System,
yet because of the limited time and resources available to develop the
proposed and final Guidance, Tables 1 through 4 do not contain criteria for
all 22 currently-identified BCCs.  Also, in some cases a pollutant that
appears in only one of the four Tables would be likely to have a lower
criterion in another of the Tables if the other criteria were developed.  For
these reasons, EPA believes that the final Guidance must also continue to
provide for State and Tribal adoption of criteria methodologies consistent
with appendixes A though D in order to increase the consistency between water
quality criteria and permit limits throughout the Great Lakes System and to
satisfy EPA's obligation to specify minimum requirements necessary to protect
human health, wildlife, and aquatic life.

      EPA also considered but rejected an alternative approach under which
States and Tribes could use the Guidance's criteria methodologies to adopt and
submit for EPA review criteria less stringent than those in Tables 1 through 4
solely on the basis of new data or information.  Such an approach would have
two major problems.  First, it would be difficult to develop an operational
definition for "new" data that would define all factors necessary for
concluding that data and information were sufficiently new or significant to
justify EPA approval of a submitted criterion less stringent than in Tables 1
through 4.  States, Tribes and EPA would also face an additional
administrative burden of determining whether particular data sets met the
definition.

      Second, although this approach would enable new data to be incorporated
rapidly,  EPA is concerned that it could result in significant departures from
the increased uniformity supported by EPA and the States and envisioned in the
CPA.  For example, one State could use a new toxicological study to develop
and adopt less stringent criteria before EPA and other experts participating
in the GLI Clearinghouse described above could review the quality and
appropriateness of the study.  EPA believes such peer review is desirable for
all pollutants, but especially for pollutants such as those in Tables 1
through 4 that have benefitted from extensive EPA, State, and public review,

-------
36     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

and include pollutants of widespread concern and interest in the Great Lakes
System.  These numeric criteria and methodologies were initially developed by
technical experts from EPA, the Great Lakes States, the U.S. Fish and Wildlife
Service, and other participants in the Great Lakes Water Quality Initiative,
who are familiar with the data form and the unique characteristics of the
Great Lakes System.  During the development of the Guidance, this cooperative
effort to review and process toxicological data also resulted in a high degree
of consensus for the numeric criteria in Tables 1 through 4.  If EPA and the
other participants had unlimited time and resources, it would be desirable to
utilize the same process for all pollutants requiring numeric criteria.

      EPA believes that the implementation plans discussed above provide a
reasonable and systematic way to incorporate the effects of new data into
water quality criteria development without the disadvantages of the
alternative approach.  Additionally, as discussed further below, the Guidance
provides adequate flexibility to the States and Tribes to modify the criteria
in Tables 1 through 4.  This flexibility is provided through § 132.4 (i), which
preserves State and Tribal authority to adopt more stringent provisions;
through § 132.4(h) which provides for more or less stringent modifications
based on scientific indefensibility and is discussed in section II.C.6 of this
document; and through § 132.3 which provides for more or less stringent site-
specific modifications of the criteria in the circumstances specified in
procedure 1 of appendix F.

      EPA also does not believe that it is necessary or appropriate for EPA to
be solely responsible for developing new or revised Tier I criteria, or to
make State or Tribal adoption of new or revised criteria contingent on EPA
amendment of Tables 1 through 4 of part 132 in future rulemakings.  The CWA
has always placed the primary responsibility for developing and adopting
criteria on the States and Tribes approved to administer water quality
standards programs.  If the final Guidance were to alter this relationship and
make EPA responsible in the future for developing and/or conducting rulemaking
for all new or revised criteria. State and Tribal adoption of criteria that
might otherwise be necessary to protect aquatic life, human health, or
wildlife and for which data were available could be significantly delayed.
EPA believes that the provision for States to adopt the methodologies found in
appendixes A through D, together with the information exchange and peer review
that will occur through the operation of the GLI Clearinghouse, will provide
the States and Tribes with an adequate framework for ensuring consistent and
timely interpretation of narrative standards.  At the same time, EPA
recognizes that there are some situations, especially with toxic pollutants of
high concern and interest, where additional EPA involvement in criteria
development is desirable.  'Furthermore, EPA's planned involvement in
coordinating and disseminating information will maximize efficient use of
limited State, Tribal, and Federal resources.  For this reason, EPA is
prepared to participate actively in operating the GLI Clearinghouse described
above, and is committed to working with States and Tribes to develop, review,
analyze, and disseminate data, and to develop GLI criteria guidance documents
as necessary in accordance with available resources.

      EPA also agrees with comments that EPA could help facilitate public
review of the adoption of new or revised criteria, and implementation of the
criteria in NPDES permits.  Dnder 40 CFR 131.20, States and Tribes are to
provide an opportunity for public participation in adoption or revision of
water quality standards.  Under 40 CFR 123,25, States implementing the NPDES
program are to provide an opportunity for public participation during
development of discharge permits.  EPA believes the GLI Clearinghouse will
assist States and Tribes in  implementing these responsibilities by providing
an additional opportunity for the public to review data that may be used in
the development of future water quality standards and NPDES permit limits.
Furthermore, EPA intends to  provide an opportunity for public review and
comment before finalizing any future GLI criteria guidance  documents.

-------
                           Section II: Regulatory Requirements                        37
      Finally, EPA agrees with commenters that it is important to incorporate
into regulatory programs any new toxicblogical data which shows that a
threatened or endangered species--or a surrogate for such species--is
particularly sensitive to a pollutant.  EPA believes the implementation plans
described above will ensure that endangered or threatened species are
adequately protected.  For example, when relevant new data become available,
the data will be placed in the GLI Clearinghouse where they will be available
to States, Tribes, and other interested persons.  EPA will place special
emphasis in the Clearinghouse on data relevant to protection of endangered or
threatened species, and will alert States and Tribes to data that indicate
unusual sensitivity of these species or their surrogates in the Great Lakes
basin.  EPA and the U.S. Fish and Wildlife Service will also work with States
and Tribes to identify sites where special protection may be recommended.
States and Tribes must use such data wherever appropriate in their water
programs, and EPA will evaluate these actions during the triennial reviews of
water quality standards under section 303(c) of the CWA.  Before approving
State and Tribal water quality standards, EPA will ensure that they are not
likely to jeopardize the continued existence of any listed endangered or
threatened species, or result in the destruction or adverse modification of
such species' critical habitat.  EPA could not approve State or Tribal
adoption of a criterion that does not adequately protect endangered or
threatened species.  EPA's implementation of the Endangered Species Act is
discussed further in section II.G of this document.

      c.    Final Guidance.  For the reasons above, the proposed minimum
requirements regarding Tier I criteria and methodologies generally have been
retained in the final Guidance.

      In response to issues and comments discussed elsewhere in this document,
the minimum required application of the Tier I methodology for development of
wildlife criteria has been limited to BCCs.  This change is reflected in §
132.4(a)(5), § 132.4(d)(4), and appendix D to part 132, and is discussed
further in section VI of this document.  The final Guidance retains the
proposed provisions to apply the Tier I methodologies for human health and
aquatic life to all pollutants except those in Table 5 of part 132.

2.    Adoption and Application of Tier II Methodologies

      a.    Proposal.  The proposed Guidance included a mechanism to interpret
and ensure that existing narrative prohibitions against the discharge of toxic
substances in toxic amounts are reflected in water quality-based effluent
limitations in a consistent way throughout the Great Lakes System, even where
data are limited.  If a State or Tribe has not adopted a numeric water quality
criterion for a pollutant and insufficient data exist to meet Tier I minimum
data requirements, proposed § 132.4(c) would provide for application of Tier
II methodologies to develop Tier II values to implement the narrative
criteria.   Additionally, if sufficient data to calculate a Tier II value for a
pollutant on Table 6 of part 132 do not exist, procedure 5 of the proposed
implementation procedures (appendix F to part 132) provided that the
permitting authority,  under specified circumstances, would generate or require
the permittee to generate the data necessary to derive Tier II values.

      The above approach is consistent with existing EPA national regulations
at 40 CFR 122.44(d)(vi) which require permit authorities to establish effluent
limits for individual pollutants or indicator parameters under specified
circumstances when the pollutant is present in the effluent at a concentration
that causes, has the reasonable potential to cause, or contributes to an
excursion above any water quality standard, including numeric and narrative
criteria for water quality.  This approach is discussed further in section
VIII.E of this document.

      b.    Comments/  EPA received many comments objecting to the inclusion
of Tier II methodologies in the Guidance.  Among the concerns raised were that
the Tier II methodologies are not scientifically valid for use in determining

-------
38     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

water quality-based effluent limits, that it is inappropriate to encourage
dischargers to generate data to be used in development of Tier II values, that
such data generation is too costly and time-consuming, and that the Tier II
methodologies are overly restrictive.  A large number of comments singled out
the Tier II methodology for aquatic life as a particular concern because it
allows use of fewer data than the established national aquatic criteria
methodology, because values are adjusted to be more restrictive when fewer
data points are available, and because of possible overlap with the
implementation procedure for whole effluent toxicity.  Many comments singled
out the Tier II methodology for wildlife because of concerns about the
appropriateness of the scientific approach.  Some commenters suggested  the use
of biological tests such as whole effluent toxicity tests and bioconcentration
testing in place of' the Tier II methodologies for numerical pollutant-specific
values.  Some commenters suggested that anti-backsliding and antidegradation
requirements would prevent future adjustments in Tier II numbers when more
data become available.  Some commenters suggested that EPA should have  the
responsibility to generate Tier II values.

      EPA also received a number of comments supporting the proposed use of
Tier II methodologies in the Guidance.  These comments indicated that it is
critical for the waters of the Great Lakes System that toxic substances not
remain unregulated simply because adequate data to calculate a Tier I
criterion are not available.  These commenters also believe that the proposed
approach properly places the burden of proof on dischargers to demonstrate
that their discharges will not damage the aquatic ecosystem or human health.

      With regard to the aquatic life methodology, EPA carefully reviewed the
concerns of commenters about potential overlap between Tier II requirements
for aquatic life and the appendix F requirements for whole effluent toxicity
testing.  The final Guidance clarifies and elaborates on the option of  using
indicator parameter limits under 40 CFR 122.44(d)(1)(vi)(c).  A full
discussion of the use of the indicator parameter option and its role in the
final Guidance appears in section VIII.E of this document.  As discussed in
that section of the*document, the State or Tribe is still required to adopt
the Tier II methodology for aquatic life.  As described in procedure 5  of the
final Guidance, States and Tribes will then be required under this Guidance to
implement the Tier II values through water quality-based effluent limitations
(WQBEL) based on such values when, under procedure 5, such limits are
determined to be required.  When deriving limits to meet Tier II values.
States and Tribes have the option of using an indicator parameter limit,
including use of a WET limit under appropriate conditions, in lieu of a Tier
II-based limit.  If an indicator parameter is used, the State or Tribe  must
ensure that the indicator parameter will attain the  "applicable water quality
standard"  (as described in 40 CFR 122.44(d)(1)(vi)(C).  The "applicable water
quality standard" in this instance would be the State's or Tribe's narrative
water quality standard that protects aquatic life, as interpreted using the
Tier II methodology.

      EPA does not agree, however, with comments that questioned the
scientific validity of the proposed aquatic life Tier II methodology.   EPA has
reviewed commenters1 concerns, and concluded that the Tier II approach  is
scientifically sound and necessary for the Great Lakes System, and is
appropriate for development of water quality-based effluent limits.
Furthermore, EPA does not agree with comments that the methodology is overly
restrictive.  For these reasons,  the proposed aquatic life Tier II methodology
is retained in the final Guidance.  Further discussion of this issue is found
in section III of this document.

      EPA also does not agree with comments  that  the proposed human health
Tier II methodology is not  scientifically valid.  A  full discussion of  this
issue is found in section V of this document.

      EPA does agree with  some of the concerns expressed about the
applicability of the wildlife methodology for developing wildlife  criteria  or

-------
                           Section II: Regulatory Requirements                        39
values beyond the BCCs.  As a result, EPA has modified the final Guidance  to
no longer require Great Lakes States arid Tribes to adopt provisions mandating
the use of the proposed wildlife Tier II methodology when developing water
quality-based effluent limits, nor to require the use of the wildlife Tier I
methodology for pollutants other than BCCs.  These issues are discussed
further in section VI of this document.

      EPA also agrees with comments that it is appropriate for dischargers to
share the cost of developing data on pollutants for which there are no Tier I
criteria and/or Tier II values, and does not agree with comments that it is
inappropriate.  EPA recognizes that the ultimate statutory responsibility  for
developing, adopting, and approving water quality standards rests with States,
Tribes, and EPA.  The CWA, however, also makes dischargers ultimately
responsible for the content of their discharges, and gives broad authority to
the Administrator and the States for data gathering and reporting concerning
such discharges  (CWA sections 308, 402).  As a matter of public policy, EPA
believes it is appropriate to provide incentives to reduce the presence of
toxic pollutants in discharges.  Procedure 5.C of appendix F of the final
Guidance provides such an incentive.  Upon adoption or promulgation, it would
require NPDES permit authorities to generate toxicological data sufficient to
calculate Tier II values in certain circumstances, but would retain their
current ability to require permittees to generate such data.

      At the same time, EPA does not want to impose an undue burden on
dischargers, and has reviewed carefully the comments of those concerned about
the cost and time required to generate Tier II data.  EPA has concluded that
because of the amount of existing data already available for the GLI
Clearinghouse, the potential burden to generate required Tier II data in
specified circumstances will be relatively insignificant.  EPA's analysis  is
as follows.

      Upon adoption or promulgation, the Guidance would provide that
dischargers could be required, under conditions and with some exceptions
specified in procedure 5.C of appendix F, to generate data to derive Tier  II
values if sufficient data do not exist for the 138 pollutants selected by  the
States and EPA for the initial focus of the Initiative  (listed in Table 6  of
part 132).  This means that a permitting authority, when faced with the need
to develop an effluent limit for a pollutant for which there is no Tier I
criterion or Tier II value, would need to generate or require the permittee to
generate data to develop a Tier II value for that pollutant.  The conditions
and exceptions for this procedure are described further in section VIII.E  of
this document.  For human health, EPA has already developed Tier I criteria
for 18 of the 138 pollutants, and has tentatively determined that there is
currently enough toxicological information available to calculate at least 101
more Tier I criteria, leaving only 14 pollutants for which there is not enough
data to calculate at least a Tier II human health value, and 5 pollutants
which are currently under review by EPA.  For aquatic life, EPA has developed
Tier I criteria for 15 of the 138 pollutants, and in addition has tentatively
determined that there is currently enough information available to calculate
at least 98 more Tier I criteria or Tier II values.  This means that the
maximum potential burden for dischargers to generate required data would be
for 14 human health values and 25 aquatic life values.  For several of the
remaining 25 pollutants, however, it is possible that the pollutant is
insoluble in water at levels that are acutely toxic to daphnids.  In these
situations, under the scientific defensibility exclusion of § 132.4(g) there
would be no need to conduct the toxicity tests.

      This maximum potential burden may never be realized by dischargers for
several reasons.  First, in operating the GLI Clearinghouse, EPA and the
States may identify additional data on the remaining pollutants.  Second,  EPA
may decide to generate the necessary data for one or more of the remaining
pollutants.  Third, one or more States may decide to generate the data, and
not require dischargers to generate the data.  Finally, the pollutants may not
be in an individual discharger's effluent, and therefore would not trigger the

-------
40     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

need to generate data under procedure 5 of appendix F.  Because of the above
mitigating factors, EPA believes the burden to generate required Tier II data
is likely to be small.

      EPA is also aware that because the Tier II methodologies generally yield
more conservative numbers than Tier I,  to reflect the greater uncertainty
related to the absence of complete data sets, an incentive is created for
dischargers to generate additional toxicological data to enable generation of
a new Tier II value or a Tier I criterion.  To the extent that dischargers
choose to conduct studies to generate new data, this is an optional activity
and not a requirement of the final Guidance.  Additionally, the cost of such
testing could be offset by any reduction in treatment costs associated with
less stringent permit limits based on Tier I criteria.  EPA has therefore not
estimated the potential cost if any of this optional activity.

      Additionally, in situations where dischargers are required to generate
data to derive Tier I criteria or Tier II values, the final Guidance continues
to specify that permit authorities may grant a reasonable period of time in
which to provide additional studies necessary to develop a Tier I criterion or
to modify the Tier II value.  These provisions are discussed further in
section VTII.H of this document.

      EPA also agrees in part with commenters who recommended allowing the use
of established biological tests in place of the Tier II methodologies for
numerical pollutant,-specific values.  As described above, when deriving limits
to meet Tier II values, States and Tribes have the option of using an
indicator parameter limit, including use of a WET limit under appropriate
conditions, in lieu of a Tier II-based limit.  EPA does not agree, however,
with suggestions by commenters that the entire Tier II approach be replaced by
reliance on established biological tests.  Although whole effluent toxicity
testing and permit limits are established procedures that EPA believes are
effective in controlling toxicity to aquatic organisms in many circumstances,
such testing is not designed to address toxic effects on human health.
Furthermore, the whole effluent toxicity approach has some limitations for
aquatic life protection as discussed in the section III of this document.
Similarly, other ecological testing suggested by commenters, such as rapid
bioassessment and biological criteria evaluations, do not address human health
concerns.  Other suggestions, such as effluent and chemical-specific
bioconcentration testing, and fish tissue residue studies of receiving waters,
have been adopted into the Guidance as part of the methodology to develop
bioaccumulation factors in section IV of appendix B, and as part of procedure
8.F of appendix F.  They are measurement methods, however, not regulatory
approaches, and therefore should not be used as separate regulatory controls
to replace the Tier II approach as suggested by some commenters.  Furthermore,
they do not directly measure end points of concern in protecting aquatic life,
and would likely nof be acceptable indicators under existing
§ 122.44(d)(1)(vi).

      Finally, EPA does not agree with the concern of many commenters that the
CWA's anti-backsliding provisions will prevent the future adjustment of water
quality-based effluent limits based on Tier II values.  Application of anti-
backsliding provisions of the CWA are discussed in section II.C.3 of this
document.

      c.    Final Guidance.  For the reasons above, EPA has made specific
modifications and exceptions to the proposed minimum requirements for States
and Tribes to adopt-and use Tier II methodologies.  These changes include:

            The application of the Tier II methodology for aquatic life has
been modified to clarify that States and Tribes have the flexibility to use
either Tier II values or indicator parameters  including whole effluent
toxicity, where appropriate and justified,  consistent with the current
national NPDES permit program.  This change is reflected in changes to

-------
                           Section II: Regulatory Requirements                        41
procedure 5 of appendix F to part 132, and is discussed further  in VIII.E  of
this document.

            EPA has eliminated the proposed minimum requirement  to use  a Tier
II methodology for deriving wildlife values in the development of water
quality-based effluent limits.  This change is reflected by removing  the Tier
II wildlife methodology from appendix D to part 132, and retitling the
appendix; by removing requirements to adopt and ufie the methodology in
§§ 132.4(a)(5), 132.4(c), and 132.4(d)(4); and removing wildlife values from
the definition of Tier II value in § 132.2.  These changes are discussed
further  in section VI of this document.  Removing these requirements  from  part
132, however, does not relieve permit authorities from their responsibilities
under 40 CPR 122.44(d)(1) to ensure that narrative water quality standards are
implemented to protect wildlife.

3.    Application of Anti-backsliding Provisions of the Clean Water Act

      a.    Proposal.  The preamble to the proposed Guidance  (58 PR 20837)
discussed why in most cases the anti-backsliding provisions of the CWA  will
not prevent adjustments to either Tier I criteria or Tier II values.   ("Anti-
backsliding requirements" refers to the need to make certain showings to
justify making an effluent limitation less stringent.)  First, because  anti-
backsliding requirements do not apply to changes made in an effluent
limitation prior to its compliance date, they would not apply where permit
limits based on Tier II values are revised as a result of studies performed
under a compliance schedule pursuant to proposed procedure 9 of  appendix F.
Second, even where anti-backsliding requirements do apply  (e.g., where
effluent limitations based on Tier I criteria or Tier II values  change  after
the compliance date), dischargers may be able to meet the requirements  of
section 303(d)(4) and therefore be allowed to have less stringent effluent
limitations.

      b.    Comments

      Comment;  Comments indicated that EPA should exempt Tier II based limits
from anti-backsliding requirements regardless of the time period involved,
even after the limits become effective.

      Response:  The statute does not provide the flexibility to exempt Tier
II limits from anti-backsliding requirements after the limits become
effective.  However, as discussed below, even where anti-backsliding
requirements do apply, they do not prevent all changes for effluent
limitations.

      Comment:  Some comments asked for the statutory justification for
indicating that anti-backsliding requirements do not apply until the
compliance date or if a limitation is appealed.

      Response:  EPA has reviewed the statutory language contained in section
402(o)(1), which contains the general prohibition against backsliding for
WQBELs.  The statute refers to effluent limitations that have been
"established."  Restrictions on backsliding do not apply to challenged  permit
limits which have been stayed pending final agency action as these limitations
have not been "established."  For example, where a limit is challenged  in  an
evidentiary hearing or administrative appeal, the limit may be made more or
less stringent than the initially proposed limit, without the change  being
subject to the anti-backsliding requirements.  EPA has also determined  that
the anti-backsliding requirements do not apply to limits with a  delayed
compliance date, until the date of compliance, as the limitation again  is  not
yet "established" until that date.  EPA has codified in procedure 9.C of
appendix F of the final Guidance its interpretation that a limitation is not
established for the purposes of section 402(o) until its effective date.

-------
42     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      None of the Great Lakes States has indicated to EPA that  they have anti-
backsliding requirements that are more stringent than the EPA interpretation.

      Comment:  Commenters stated that it will be very  difficult  to satisfy
the anti-backsliding requirements.

      Response:  There is flexibility provided by EPA's interpretation of
sections 402(o) and 303(d)(4) to allow relaxation of effluent limitations in
many of the circumstances described in the comments.  Specifically,  EPA does
not interpret section 402(o) to require compliance with both sections
402(o)(2) and 303(d)(4), but only with one or the other (as well  as 402(o)(3))
in order to establish less stringent limitations.

      Comment:  Commenters stated that EPA should issue a regulation
interpreting the anti-backsliding requirements.

      Response:  EPA is not issuing a regulation defining national  anti-
backsliding policy in this rule, since such a regulation would  have much
broader applicability than the Great Lakes basin.  However, EPA is  repeating
and clarifying its interpretation of the anti-backsliding requirements of
sections 402(o) and 303(d)(4) in this document.  See the Final  Guidance
section below for a detailed explanation of EPA's interpretation.

      Comment:  Commenters stated that backsliding should not be  allowed from
current effluent limitations and existing conditions.

      Response:  The statute or existing regulations do not prohibit a change
from existing limitations, so long as the exceptions are met and  there is
compliance with the requirements contained in section 402(o)(3).

      Comment:  A number of comments dealt with the interaction of  anti-
backsliding and antidegradation requirements.

      Response:  Antidegradation requirements do apply  to an analysis  of the
anti-backsliding requirements, as provided under section 402(o)(3).  However,
this is consistent with the general requirement under the CWA that  permits
comply with water quality standards including antidegradation.   See the NPDES
regulations (40 CFR 122.44(b) and  (d)) on establishment of WQBELs to ensure
compliance with water quality standards, including antidegradation  policies.
Specific issues concerning compliance with antidegradation in the Great Lakes
are addressed in section VTI of this document.

      Comment:  Some Commenters expressed concern that  the antibacksliding
requirements of section 303(d)(4)(A) of the CWA would make it difficult to
revise limits based on Tier II values if studies were completed after  the
compliance date for the limit in question.  For example, if future  studies
resulted in a Tier II value being changed to a less stringent Tier  II  value or
Tier I criterion, after the effective date of an initial Tier II-based limit,
the concern is that antibacksliding would prohibit the  relaxation of the Tier
II-based limit.

      Response:  Section 303(d)(4)(A) requires that the limit which is to be
relaxed have been based on a TMDL or other wasteload allocation established
under section 303(d).  Since any effluent limit based on a Tier II  value will
also be based on procedure 3 of appendix F  (i.e., on an approved TMDL, WLA in
the absence of a TMDL, or assessment and remediation plan),  such effluent
limitations will satisfy the requirement that they be based  on  a TMDL  or other
wasteload allocation established under section 303(d).  EPA  therefore  believes
that permittees will be able to satisfy the requirements of  section
303(d) (4) (A) .

      c.    Final Guidance.  EPA has retained the proposed provisions
concerning the application of CWA anti-backsliding requirements to  the
implementation of the final Guidance.  Final procedure  9.C of appendix F

-------
                           Section II: Regulatory Requirements                        43
states explicitly that anti-backsliding requirements contained  in  section
402(o) do not apply to changes made in an effluent  limitation prior  to its
compliance date.  In addition, there is adequate flexibility contained in
EPA's interpretation of the anti-backsliding requirements of the CWA to allow
adjustments to either Tief I criteria or Tier II values  in many situations.

      The approach in the final Guidance regarding  the applicability of anti-
backsliding requirements will provide the greatest  degree of uniformity among
the States and Tribes in terms of when Tier II values would become a part of  a
final permit.  In addition, the approach of the final Guidance  will  likely
provide the greatest degree of environmental protection  in the  short term
because there will be a shorter time for completion of the studies and
issuance of a permit with final effluent limitations.  This approach also is
consistent with the language of the CWA itself.

      The following discussion clarifies and expands EPA's interpretation of
sections 402(o) and 303(d)(4), and explains why the States and  Tribes have
considerable flexibility to revise WQBELs.

      Section 402(o) of the CWA, added by the Water Quality Act of 1987 (WQA),
for the first time establishes express statutory language prohibiting the
backsliding of permit limits.  Section 402(o) consists of three main parts.
First, section 402(o)(1)  prohibits  (subject to exceptions in sections
303(d)(4) and/or 402(o)(2)) backsliding of two types of  permit  limits:  (1)
technology-based effluent limitations based on best professional judgment
(BPJ) being revised to reflect subsequently promulgated  effluent guidelines
which are less stringent, and  (2) water quality-based effluent  limitations
established on the basis of sections 301(b)(1)(C) or 303(d) or  (e).   Second,
section 402(o)(2) outlines six specific exceptions  to the prohibition
contained in section 402(o)(1).  Third, section 402(o)(3) outlines a baseline
requirement that must be met before any limits can  be relaxed,  namely that the
new limit must ensure compliance with applicable effluent guidelines and water
quality standards.

      EPA's pre-WQA anti-backsliding regulations were revised on January 4,
1989  (54 FR 246), to reflect the prohibition imposed by  section 402(o)  for the
first situation: revision of existing BPJ-based permit limitations to reflect
subsequently issued effluent guidelines  (40 CFR 122.44(1} (2)) .  EPA's current
anti-backsliding regulations have not been revised  to reflect the  1987 WQA
prohibition on the backsliding of the second situation:  relaxation of effluent
limitations established on the basis of sections 301(b)(1)(C) or 303(d)  or
(e).  However, EPA believes these provisions must be implemented based upon
the CWA in the meantime.

      All other types of backsliding--for example,  backsliding  from  effluent
guideline derived limits, from new source performance standards, from existing
BPJ limits to new BPJ limits, or from water quality-related standards or
conditions (except for effluent limitations)--remain unaffected by the 1987
WQA amendments and EPA's existing regulations at 40 CFR  122.44(1)(1)  will
continue to govern them.   This is because section 402(o) only prohibits the
backsliding of "effluent limits," not other standards or conditions  such as
monitoring frequency or changes in species or protocol for whole effluent
toxicity (WET) testing.  In addition, the requirements contained in  CWA
section 402(o) do not lend themselves to application to  changes of such
standards and conditions.  The relaxation of all other types of standards or
conditions contained in a permit are, however, subject to EPA's existing
backsliding regulations at 40 CFR 122.41(1) (1).  Under these regulations, a
permittee must meet a cause for modification in order to allow  relaxation.

      As indicated in the preamble to the proposed  Guidance  (58 FR 20837), EPA
has consistently interpreted section 402(o) of the  CWA to allow relaxation of
WQBELs if either the requirements of section 402(o)(2) or section  303(d)(4)
are met.  These are independent exceptions to the prohibition against
relaxation of water*quality-based permit limitations.  In dealing  with anti-

-------
44     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

backsliding issues under the final Guidance, section 303(d)(4) will,  in most
cases, provide the flexibility necessary for permitting authorities  to  issue
permits reflecting adjustments to Tier I criteria or II values.

      Section 402(o)(1) provides that backsliding from WQBELs  is prohibited
except in compliance with section 303(d)(4).  Section 303(d)(4) has  two parts
that must be considered, along with an identification requirement: paragraph
(A) which applies to "non-attainment waters" and paragraph  (B) which applies
to "attainment waters."

      Section 303(d) (4) (A) allows establishment of a less stringent  WQBEL when
the receiving water has been identified under section 303(d) (1) (A) and  where
applicable water quality standards are not being met  (i.e.,  a  "non-attainment
water"), if the permittee meets two conditions.  First, a permittee  may seek a
less stringent effluent limitation under section 303(d) (4) (A)  only if the
existing permit limitation was based on a total maximum daily  load  (TMDL)  or
other wasteload allocation  (WLA) established under section  303.  Second,
relaxation of a WQBEL is only allowed if attainment of water quality standards
is ensured, or if the designated use which is not being attained is  removed in
accordance with 40 CFR part 131.

      Section 303(d)(4)(B) applies to waters where the water quality equals or
exceeds levels necessary to protect the designated use, or  to  otherwise meet
applicable water quality standards  (i.e., an "attainment water").  Under
section 303(d)(4)(B), permit limitations based on a section 303 TMDL/WLA,  on
any water quality standards established under section 303,  or  on any other
permit standard may be relaxed only where this is consistent with a  State's
antidegradation policy  (see 40 CFR 131.12).

      Section 402(o)(2) also outlines exceptions to the general prohibition
against backsliding from WQBELs.  These exceptions are independent of the
section 303(d)(4) exception discussed above and are also applicable  to  the
backsliding of BPJ limits to reflect subsequently promulgated  less stringent
guidelines.

      Regardless of whether any of the backsliding exceptions  are applicable
and met, section 402(o)(3) acts as a floor and restricts the extent  to  which
WQBELs  (and BPJ limits) may be relaxed.  Specifically, section 402(o)(3)
prohibits the relaxation of such permit limitations below applicable
technology-based effluent limitation guidelines in effect at the time the
permit is renewed, reissued or modified.  In addition, it prohibits  the
relaxation of limits if such relaxation would result in a violation  of
applicable water quality standards, which include antidegradation
requirements.

      EPA is providing four examples of the application of  anti-backsliding
requirements.   (EPA has not provided the analysis of these  examples  under
section 402(o)(2), because section 303(d)(4) in almost all  cases is  more
flexible.)

            Examole  1
            Scenario; A publicly owned  treatment  works  (POTW)  seeks to relax
      its WQBEL for pollutant X.  The current permit  limitation is based on
      the TMDL and WLA for the POTW  developed in  accordance  with 40 CFR 130.7.
      The POTW is in compliance with its  existing limitation.   The applicable
      water quality standards for pollutant X is  attained.   The POTW has
      developed new models with new  river flow  information,  which indicate
      that the water quality standards  for pollutant  X  would be maintained
      with a relaxed permit limitation.   The permitting authority can revise
      the permittee's WLA to allow a larger discharge of pollutant X because
      another discharger to the TMDL ceased the discharge of pollutant X.  May
      the effluent limitation for pollutant X be  relaxed?

-------
                     Section II: Regulatory Requirements                        45
      Answer: Section 303(d)(4) may justify the requested relaxed permit
limitation.  Section 303(d)(4)(B) is the applicable provision because
the water quality standards for pollutant X is currently attained.
Under section 303(d)(4)(B), a permit limitation can be relaxed  if
antidegradation requirements are met.  Finally, the permitting  authority
may only allow backsliding if the relaxed limitation would not  result  in
violation of any effluent limitations guideline or other water  quality
standards.

      Example 2

      Scenario; On June 30, 1991, the State issued a NPDES permit to an
industrial permittee which for the first time included a WQBEL  for
pollutant Y.  The limitation for pollutant Y is a delayed effective date
limitation which is effective on June 30, 1994.  The WQBEL is derived
from the State's existing water quality criteria.  The State conducted
additional water quality studies on pollutant Y during its triennial
review in 1993 and relaxed the water quality criterion for pollutant Y.
On January 30, 1994, the permittee seeks to modify its permit to relax
the effluent limitation for pollutant Y, based upon the new State water
quality criterion.  Will the anti-backsliding provisions of the CWA and
NPDES regulations prevent relaxation of the permit limitation?

      Answer: No.  In this case, the permittee seeks to revise  an
effluent limitation which is not yet effective.  The anti-backsliding
provisions of the CWA and NPDES regulations do not apply to a delayed
effective date limitation until it is effective.  Prior to relaxing the
permit limitation, however, the permitting authority will need  to ensure
that the action is consistent with antidegradation provisions.

      Example 3

      Scenario; The State has a narrative water quality criterion of "no
toxics in toxic amounts."  On the basis of WET testing data or  other
information, the State finds reasonable potential to exceed the
narrative water quality criterion and imposes a WET limitation  under 40
CFR 122.44(d)(1)(v).  The permittee determines that pollutant Z is the
cause of the WET in its discharge.  The permittee can demonstrate
through sufficient data  (including WET testing data) that an effluent
limitation for pollutant Z will assure compliance with the narrative
water quality standards as well as the State's numeric criteria for
pollutant Z as required by 40 CFR 122.44(d)(1)(v).  May the State modify
the permit to delete the WET limitation and to add the limitation for
pollutant Z?

      Answer: Section 303(d)(4) may justify this action.  The applicable
provision of section 303(d)(4) is section 303(d)(4)(B) because  the
narrative water quality standards is currently attained.   (The  permittee
is currently complying with the existing WET limitation to attain and
maintain the State's narrative water quality standards.)  Under section
303(d)(4)(B) , the permittee may backslide so long as antidegradation
requirements will be met, and the relaxed limitation will not cause a
violation of any effluent limitations guidelines and water quality
standards applicable to the discharge.  In this case, this appears
likely because the discharger can demonstrate that the new limitation
for pollutant Z will assure compliance with applicable narrative as well
as numeric water quality standards.

      Example-4

      Scenario: An industrial permittee seeks to revise its WQBEL of
1000 mg/L for a pollutant to 6000 mg/L, its actual discharge level.  The
permittee has installed and properly operated and maintained its
treatment facilities, but has been unable to achieve the effluent

-------
46     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      limitation of 1000 mg/L.  The current permit limitation is based upon a
      TMDL and WLA for the permittee, which were developed in accordance with
      40 CFR 130.7.  The water quality standards for the pollutant is not
      being attained.  New modeling information shows that the water quality
      standards for the pollutant will be attained with a permit limitation of
      4000 mg/L.  The permitting authority is able to revise the discharger's
      WLA to allow a permit limitation of 4000 mg/L.  May the permit
      limitation be revised from 1000 mg/L to 6000 mg/L?

            Answer: No.  However, under sections 303(d)(4), the permit
      limitation may potentially be relaxed to 4000 mg/L.  The water quality
      standards for the pollutant is not currently being attained.  Therefore,
      the applicable 303(d)(4) provision is 303(d)(4)(A).  In this case, the
      permitting authority may allow backsliding to 4000 mg/L under section
      303(d)(4)(A) because the existing effluent limitation is based upon a
      TMDL/WLA and the data shows that attainment of the water quality
      standards is assured with a permit limitation of 4000 mg/L  (but not with
      a limitation of 6000 mg/L).  The permitting authority may also revise
      the discharger's WLA to allow a discharge limitation of 4000 mg/L.
      However, before backsliding to a limitation of 4000 mg/L will be allowed
      in this case, the permitting authority must also find that the relaxed
      limitation will not result in violation of applicable water quality
      standards  (including antidegradation requirements) and effluent
      limitations guidelines for the discharge.

4.    Basin-wide Application of Criteria and Values

      a.    Proposal.  With exceptions discussed below, the proposed Guidance
generally provided for State and Tribal application of the criteria, values
and methodologies in the Guidance to all waters of the Great Lakes System
regardless of current use designations.  This approach was selected in order
to provide the integrated Great Lakes ecosystem a consistent approach to
pollution control across the entire basin  (see 58 PR 20838-40).

      The proposal contained four exceptions to this approach: First, Great
Lakes States or Tribes could apply more stringent numeric criteria or values
to any waters of the Great Lakes System within their borders.  Second, Great
Lakes States or Tribes could develop less stringent site-specific
modifications to the criteria and values for aquatic life for specific waters
of the Great Lakes System in certain limited circumstances.  Third, Great
Lakes States and Tribes would be required to adopt different types of human
health criteria depending in part on the designated uses of the waters:
"drinking" criteria and values would apply  to open waters and connecting
channels of the Great Lakes, and to other waters designated for use as public
water supplies;  "nondrinking" criteria and  values would  apply to all other
waters.  Fourth, the Guidance provided general exceptions for 16 pollutants
listed in Table 5 of the proposed Guidance; for discharges from wet weather
point sources; and for situations where the criteria methodologies are not
scientifically defensible.

      b.    Comments.  A few commenters asserted  that  the application of human
health and/or wildlife criteria/values regardless of current use  designations
throughout the Great Lakes System seems overly restrictive.  In particular,
these commenters are concerned  that this practice is not justified for non-
bioaccumulative  substances because these pollutants may never reach the  lakes,
actual drinking water supplies,  or appropriate wildlife  habitats  when
discharged into upstream waters.  Other commenters  stated that Congress  did
not intend to take away the Great Lakes States' ability to develop use
designations and to  develop water quality  standards protective of those  uses.
Other commenters stated that  the approach  fails to  recognize the  ecological
diversity of the Great Lakes  ecosystem.

-------
                           Section II: Regulatory Requirements                        47
      A few comments supported the proposed approach which allows application
of site-specific criteria as a more appropriate mechanism for developing
criteria and values than through a use-designation mechanism.

      EPA carefully considered the concerns expressed about the proposed
approach, but continues to believe the proposed approach is appropriate.
First, section 118(c) of the CWA requires the Guidance to specify minimum
numerical limits on pollutants to protect human health, aquatic life, and
wildlife in the Great Lakes System.  EPA believes it is a reasonable
interpretation of this requirement to develop criteria that are generally
applicable to the entire Great Lakes System in order to improve consistency of
water quality criteria while providing sufficient flexibility to address site-
specific circumstances.   (See 58 FR 20837-40.)  This interpretation is also
reasonable in light of the short statutory deadline established by Congress to
complete the Guidance.

      Second, the approach was not developed because of an assumption that
pollutants move freely throughout the ecosystem, although both bioaccumulative
and non-bioaccumulative pollutants can become widely dispersed.  Rather, as
explained in the preamble to the proposed Guidance, EPA believes that the
Great Lakes are an integrated ecosystem necessitating a more consistent
approach to pollution control across the entire basin.

      Third, following the current National program of differing criteria
based on differing use designations could seriously hinder--and perhaps
prevent--the attainment of the goals of the CPA.  Contrary to the views of the
commenter above, EPA believes Congress did intend to restrict some of the
current flexibility in the national program in order to achieve more uniform
protection of the Great Lakes System through enactment of the special
requirements in section 118(c) for this ecosystem.

      Fourth, EPA believes that uniform minimum water quality standards can
help simplify implementation and avoid costly duplication of research and
standard-setting by EPA and the Great Lakes States and Tribes.

      Fifth, as a practical matter, designated uses currently do not exhibit
wide variation across the basin.  For example, use designations for most
waters within the Great Lakes System currently include protection of aquatic
life and recreational uses.  No comments were received in response to EPA's
request for comments on any waters within the Great Lakes System that are not
currently designated to protect these uses.

      Finally, to the extent that there may be unique local situations not
amenable to strict application of basin-wide criteria, the final Guidance
contains several areas of flexibility, including: the scientific defensibility
exclusion in § 132.4(h), discussed further in section IX.C.6 of this document;
site-specific criteria modifications available through procedure 1 of appendix
F, discussed further in section VIII.A of this document; and variances
available through procedure 2 of appendix F, discussed further in section
VIII.B of this document.

      EPA agrees with the comment, however, that the Guidance should recognize
the ecological diversity of the Great Lakes ecosystem.  It is reasonable to
provide differential criteria or flexibility in implementing the criteria to
reflect variations due to local physical, chemical, and biological factors.
The proposed Guidance provided some different criteria--for example, different
aquatic life criteria depending on water hardness, and different human health
criteria for open waters of the Great Lakes--to reflect common variations.  It
also provided flexibility of implementation, including site-specific
modifications to criteria to address more specific or localized differences.

      The final Guidance also includes additional flexibility in applying
criteria and values to further recognize ecological diversity within the Great
Lakes basin.  Procedure 1 of appendix F now allows site-specific modifications

-------
48     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

of human health and wildlife criteria and values that can be either more
stringent or less stringent to reflect site-specific information on
bioaccumulation factors.  This change,  together with the flexibility already
provided in the proposal allowing both more stringent and less stringent site-
specific modifications for aquatic life criteria/values, should provide
sufficient flexibility to reflect site-specific conditions and the ecological
diversity of the Great Lakes basin.  Changes to the procedure for site-
specific modifications are discussed further in section VIII.A of this
document.  In addition, changes were made in the definition of high quality
waters in the antidegradation policy to exclude certain waters from an
antidegradation review depending on their ecological, recreational, or
aesthetic significance.  Changes to the antidegradation policy are discussed
further in section VII of this document.

      c.    Final Guidance.  With one exception, the proposed provisions of
§ 132.4(d) specifying the applicability of  criteria and values have not been
changed in the final Guidance.  The exception is that § 132.4(d)(4) has been
modified to eliminate the requirement for States and Tribes to adopt
provisions consistent with the Tier II wildlife methodology, as discussed
above and in section VI of this document.

5.    Pollutants Subject to Federal. State, and Tribal Requirements

      a.    Proposal.  The proposal left to the discretion of the States and
Tribes whether to adopt provisions requiring the use of the Guidance's
criteria development methodologies or implementation procedures 1, 2, 3, 4, 5,
7, 8, and 9 in appendix F for a pollutant if it was listed in Table 5 of the
proposed Guidance.  Proposed Table 5 listed 16 pollutants selected by the
Great Lakes States and EPA during the Great Lakes Initiative process:
alkalinity, ammonia, bacteria, biochemical  oxygen demand, chlorine, color,
dissolved oxygen, dissolved solids, hydrogen sulfide, pH, phosphorus,
salinity, sulfide, temperature, total and suspended solids, and turbidity.
These pollutants would continue to be subject to existing water programs,  such
as State or Tribal programs implementing 40 CFR part 131 for the development
and adoption of water quality standards and criteria, 40 CFR part 122 for
development of NPDES permits, and other appropriate requirements and guidance
under the CWA or State or Tribal law.  They would also be subject to the
antidegradation policy in appendix E.  The  proposal did not need to exempt the
Table 5 pollutants from procedure 6, since  procedure 6 applies to whole
effluent toxicity not to individual pollutants.

      As discussed more fully in the preamble to the proposed Guidance  (58 FR
20842-43), the Initiative Committees believed that regulatory authorities
should retain the flexibility to address these pollutants in their existing
water quality programs.

      b.    Comments.  Many commenters supported the proposed listing of
pollutants in Table 5 that would only be subject to existing Federal, State,
and Tribal requirements.  In particular, several commenters supported the
proposed listing of chlorine in Table 5, pointing out that States need
continuing flexibility  in their programs to deal with the special uses of
chlorine as a disinfectant in water and wastewater treatment.  Other
commenters supported the proposed listing of ammonia, citing its role as a
nutrient as well as its toxicity.

      Other commenters urged EPA  to delete  ammonia and  chlorine from Table 5.
These commenters believe that ammonia and chlorine, although regulated by
states  in the basin, should be controlled more  consistently because of their
potential adverse effects on aquatic biota. Commenters  also argued that the
Guidance methodologies  and procedures can technically be applied to ammonia
and  chlorine, and that  the States are not consistently  regulating  these
pollutants within the basin.

-------
                           Section II: Regulatory Requirements                        49
      Various individual commenters suggested: deleting hydrogen sulfide and
sulfide from Table 5; deleting salinity from Table 5; adding non-toxic
pollutants to Table 5; and adding common inorganic constituents such as
chloride, sulfate, sodium, and calcium to Table 5.

      EPA agrees with comments that Table 5 should be retained in the final
Guidance, because for the pollutants in the final Table 5 it would be
scientifically and technically inappropriate to require use of some or all of
the Guidance's methodologies and procedures.  EPA agrees with comments that
the proposed Table 5 should be modified, however, to remove two pollutants--
hydrogen sulfide and sulfide.  The reasons for listing or removing pollutants
in Table 5, together with EPA's response to comments concerning specific
pollutants, are as follows.

                Alkalinity.  The Guidance methodologies and implementation
      procedures are not scientifically and technically appropriate for
      developing and applying water quality criteria and values for
      alkalinity:  The Guidance methodologies for numeric criteria and values
      and the corresponding implementation procedures are designed to be used
      with specific chemicals.  Alkalinity is not a specific chemical, but the
      combined effect of several substances such as carbonates, bicarbonates,
      hydroxides, borates, silicates, and phosphates.  Furthermore, the
      Guidance methodologies for numeric criteria and values and the
      corresponding implementation procedures are designed to be used with
      chemicals that exert an adverse effect as the concentration of the
      chemical becomes too high, while alkalinity exerts adverse effects on
      aquatic life if it is too low, and human health--irritation to swimmers
      by altering the pH of the lacrimal fluid around the eye--if it is too
      high.  Additionally, some components of alkalinity such as carbonate and
      bicarbonate can have a beneficial effect on water quality by complexing
      some toxic heavy metals and reducing their toxicity to aquatic life.  It
      has also been noted that some waterfowl habitats are more productive
      with higher alkalinities.  The criteria development methodologies and
      implementation procedures in the final Guidance do not address these
      types of issues or end points.  The final Guidance also does not address
      other end points of concern commonly identified for low alkalinity,
      including adverse impacts on industrial uses such as food and beverage
      production.  Therefore, EPA believes it is not appropriate to apply the
      Guidance methodologies and implementation procedures in the case of
      alkalinity.

                Ammonia.  E.PA considered carefully the concerns of some
      commenters that ammonia be removed from Table 5 and made subject to all
      provisions of the final Guidance.  EPA shares concerns of some
      commenters that there may be inconsistencies among State water programs
      addressing ammonia.  Nevertheless, EPA found that there would be
      significant problems in applying the aquatic life criteria methodology
      and corresponding implementation procedures for ammonia, and that even
      if these could be overcome, there would likely be no significant
      improvements in consistency of permit limits for this pollutant
      throughout the Great Lakes basin.

            For reasons described in section II.C.4 of this document, the
      aquatic life methodology in the final Guidance was developed to provide
      generally a single set of criteria (or a single equation, in the case of
      pollutants that are adjusted for a water quality characteristic) to
      protect aquatic life throughout the basin regardless of the specific
      species present in different waterbodies.  For ammonia, however, EPA's
      best current information is that a single set of water quality criteria
      to protect aquatic life is not appropriate.  In January 1985 EPA issued
      a water quality criteria document for ammonia,  "Ambient Water Quality
      Criteria for Ammonia - 1984."  The criteria document was supplemented by
      additional guidance to EPA Regional water quality standards coordinators
      in July 1992 in a memorandum, "Revised Tables for Determining Average

-------
50	Water Quality Guidance for the Great Lakes System ~ Supplementary Information Document

      Freshwater Ammonia Concentrations."  These documents are available in
      the docket for this rulemaking.  In these documents, two sets of
      criteria are presented: one set to protect aquatic life when salmonids
      or other sensitive coldwater species are present, and one set to protect
      aquatic life when salmonids or other sensitive coldwater species are
      absent.  Both sets include separate chronic and acute criteria for
      temperatures ranging from 0° C to 30° C.  The two sets of chronic
      criteria differ from each other by factors of approximately 1.4 at
      temperatures above 15° C, and are the same at temperatures 15° C or
      below.  The two sets of acute criteria differ from each other by factors
      of approximately 1.4 at temperatures above 20° C, and are the same at
      temperatures 20° C or below.

            Furthermore,  EPA has determined that even if ammonia were removed
      from Table 5, there would likely be no significant change in consistency
      among States and Tribes in implementing the criteria.  The EPA 1984
      ammonia criteria document provides general recommendations on
      implementation which emphasizes consideration of site-specific factors.
      The document recommends including site-specific factors when conducting
      wasteload allocation modeling in those situations, including
      consideration of effluent variability, and the selection of different
      design flows for steady-state wasteload allocation modeling depending on
      whether the systems are stressed or unstressed.  Accordingly, States and
      Tribes would likely develop site-specific criteria modifications and/or
      site-specific adaptations of wasteload allocation models for the
      majority of waters throughout the basin.

            Because the implementation procedures in the final Guidance are
      designed to be applicable to a wide range of pollutants, they do not
      provide details that would assist States and Tribes in making site-
      specific modifications and adaptations in the specific case of ammonia.
      For example, appendix F of the final Guidance: provides no direction
      concerning what site-specific information should be developed for
      wasteload allocation modeling in the above situation or how the
      information should be used; does not include direction on how to
      evaluate whether systems are stressed or unstressed; and does not
      describe how design flows should be selected in stressed or unstressed
      systems.  Therefore, in the absence of such direction, States and Tribes
      would simply be implementing current laws and national program guidance.

            EPA considered the possibility of developing more detailed
      guidance concerning the site-specific modifications and adaptations
      recommended for ammonia.  EPA also considered developing site-specific
      numeric criteria and/or site-specific wasteload allocation modeling
      parameters for ammonia for all waters in the entire basin.  Neither
      approach would be possible or administratively feasible within the time
      EPA had available to complete the final Guidance.

            Another problem raised by commenters is that ammonia not only
      produces toxic effects in aquatic life but also is a nutrient that with
      other forms of nitrogen can contribute to accelerated eutrophication of
      lakes and other waters.  EPA agrees that ammonia can contribute to
      eutrophication problems, although if it is controlled to prevent toxic
      effects, its contribution to eutrophication will be less.  Nevertheless
      ammonia needs, to be considered because of its interaction with other
      forms of nitrogen when performing total maximum daily loads and
      wasteload allocations to address eutrophication problems.  The Guidance
      methodologies and procedures do not reflect eutrophication as an end
      point.

            For these reasons, EPA decided to retain ammonia as a pollutant
      listed in Table 5 to part 132.  Nevertheless, because of inconsistencies
      among State programs in the Great Lakes basin in addressing ammonia, EPA

-------
                     Section H: Regulatory Requirements                        51
will take additional steps to review such programs.  These  steps are
discussed below under the Final Guidance section.

    •  EPA notes that under § 132.4 (e) (2), discharges that contain
pollutants on Table 5 such as ammonia are still subject to  the whole
effluent toxicity provisions contained in procedure 6 of appendix F of
the final Guidance.  These provisions are discussed further under the
Final Guidance section below and in section VTII.F of this  document.

          Bacteria.  The methodologies in appendixes B and  C and the
implementation procedures in appendix F are not scientifically and
technically appropriate for developing and applying criteria and values
for bacteria.  First, the concept of bioaccumulation in appendix B
cannot be applied in the case of bacteria.  Second, the human health
methodology in appendix C is based on toxicological analysis of dose-
response relationships, while the criteria for bacteria are based on
upper limits for densities of indicator bacteria in waters  that have
been associated with acceptable health risks for swimmers.  Finally, the
implementation procedures for individual pollutants in appendix F are
generally designed for chemicals whose effects are related  to their
concentration in the water column, whereas the effects of bacteria are
generally related to the density of bacterial colonies measured after
several days of culturing in the laboratory.  Neither steady state nor
dynamic water quality modeling used in appendix F has been  found to be
applicable to-developing total maximum daily loads, water quality-based
effluent limits, or loading limits for bacteria.  Therefore, EPA
believes it is not appropriate to apply the Guidance methodologies and
implementation procedures in the case of bacteria.

          Biochemical oxygen demand.  The Guidance methodologies and
implementation procedures are not scientifically and technically
appropriate for developing and applying water quality criteria and
values for biochemical oxygen demand  (BOD).  First, the Guidance
methodologies for numeric criteria and values and the corresponding
implementation procedures are designed to be used with specific
chemicals or related congeners.  BOD is not a specific chemical, but the
combined effect of many chemicals which lead to the depletion of oxygen
as they are degraded by aquatic organisms.  Second, BOD exerts no
measurable toxic effect itself on aquatic life, human health, or
wildlife, but rather is a precursor together with other factors of an
adverse effect on aquatic life, the lowering of dissolved oxygen.
Third, the water quality modeling needed to develop total maximum daily
loads and water quality-based effluent limits for BOD requires different
data inputs and different mathematical calculations than the modeling
generally required in appendix F for pollutants that exert  a measurable
toxic effect.- Therefore, EPA believes it is not appropriate to apply
the Guidance methodologies and implementation procedures in the case of
BOD.

          Chlorine.  EPA does not agree with comments that  chlorine
should be removed from Table 5.  Key portions of the Guidance
implementation procedures are not scientifically and technically
appropriate for applying water quality criteria and values  for chlorine.
First, chlorine exerts acute toxicity effects, but is chemically highly
reactive and degrades in receiving waters much more rapidly than most
other pollutants.  The half life of total residual chlorine can range
from 1 to 3 hours, which is far shorter than most other pollutants.
Procedures 3.C.4 and 3.D.3 of appendix F provide that wasteload
allocations to protect aquatic life from acute effects shall not exceed
the Final Acute Value (FAV) for the pollutant in order to provide an
adequate margin of safety, as a default assumption in the absence of
site-specific data.  The FAV cap may be exceeded; however,  if a site-
specific mixing zone demonstration is conducted and approved pursuant to
procedure 3 of appendix F.  EPA is concerned that because of chlorine's

-------
52     Water Quality Guidance for the Great Lakes System - Supplementary Information Document

      unique characteristics, the above procedures, together with related
      elements of procedure 5,  have the potential to result in overly
      stringent controls on discharges of chlorine in the Great Lakes, basin.
      Because the Initiative Committees excluded chlorine when developing the
      draft Guidance, State and EPA technical staff did not consider
      chlorine's unique characteristics when developing procedures 3.C.4 and
      3-D.3.  Because of these concerns, EPA does not believe it would be
      appropriate scientifically and technically to apply the Guidance
      implementation procedures without modifications to take into account
      chlorine's unique properties.  EPA considered the possibility of
      reviewing procedures 3.C.4 and 3.D.3 to evaluate and address these
      concerns, but concluded this approach would not be possible or
      administratively feasible within the time EPA had available to complete
      the final Guidance.

            Second,' in the Great Lakes basin there are a number of dischargers
      who practice deliberate,  controlled, repetitive, intermittent
      chlorination in order to control undesirable organisms in their
      production or treatment processes.  Under these scenarios, the total
      time chlorine is discharged is often limited to several hours, and the
      concentration of chlorine discharged over that time period is quite
      variable, such that the peak concentration may be experienced for only a
      limited period of time during those few hours.  The Guidance
      implementation procedures may not be scientifically and technically .
      appropriate for applying water quality criteria in waters affected by
      such discharges.  Furthermore, the restrictions in procedures 3.C.4 and
      3.D.4 concerning wasteload allocations based on acute aquatic life
      criteria were not designed for such situations, and if applied as
      specified in the final Guidance could result in wasteload allocations
      that may not be scientifically and technically appropriate.  Therefore,
      EPA believes it would be inappropriate to require States and Tribes to
      apply the aquatic life methodology and corresponding implementation
      procedures to chlorine.  EPA considered the possibility of developing
      more appropriate implementation procedures for chlorine under these
      circumstances.   EPA concluded that the effort to develop such procedures
      would not be possible or administratively feasible within the time EPA
      had available-to complete the final Guidance.

            For these reasons,  EPA decided to retain chlorine as a pollutant
      listed in Table 5 to part 132.  Nevertheless, because of inconsistencies
      among State programs in the Great Lakes basin in addressing chlorine,
      EPA will take steps to review such programs.  These steps are discussed
      below under the Final Guidance section.

            EPA notes that under § 132.4(e)(2), discharges that contain
      pollutants on Table 5 such as chlorine are still subject to the whole
      effluent toxicity provisions contained in procedure 6 of appendix F of
      the final Guidance.  These provisions are discussed further under the
      Final Guidance section below and in section VTII.F of this document.

            --  Color.  The Guidance methodologies and implementation
      procedures are not scientifically and technically appropriate for
      developing and applying water quality criteria and values for color.
      First, the Guidance methodologies for numeric criteria and values and
      the corresponding implementation procedures are designed to be used with
      specific chemicals or related congeners.  Color is not a specific
      chemical, but the result of degradation processes in the natural
      environment. -There is no agreement as to the chemical composition of
      color, and in fact the composition may vary chemically from place to
      place.  Second, the aquatic life methodology is designed to protect
      aquatic life from adverse toxicological effects, while the effects of
      color in water on aquatic life principally are to reduce light
      penetration and thereby generally reduce photosynthesis by phytoplankton
      and to restrict the zone for aquatic vascular plant growth.  The

-------
                     Section II: Regulatory Requirements                        53
criteria development methodologies in the final Guidance do not address
these end points.  Third, the water quality modeling needed to develop
total maximum daily loads and water quality-based effluent limits for
color would require different data inputs and different mathematical
calculations from the water quality models generally required to
implement the total maximum daily load procedures in the final Guidance.
EPA is not aware of any water quality modeling that is generally
available to support development of total maximum daily loads for color.
Therefore, EPA believes it is not appropriate to apply the Guidance
methodologies and implementation procedures in the case of color.

          Dissolved oxygen.  The Guidance methodologies and
implementation procedures are not scientifically and technically
appropriate for developing and applying water quality criteria and
values for dissolved oxygen.  First, the Guidance methodologies for
numeric criteria and values and the corresponding implementation
procedures are designed to be used with chemicals that exert an adverse
effect as the concentration of the chemical becomes too high, while
dissolved oxygen exerts adverse effects on aquatic life if it is either
too low or too high.  The Guidance methodology to develop aquatic life
criteria and values, and implementation procedures such as those for
total maximum daily loads and loading limits would require significant
revision to accommodate such a pollutant.  Second, the Guidance
implementation procedures are generally designed to develop water
quality-based effluent limits for the same pollutant that has reasonable
potential to exceed water quality standards, while to achieve a
dissolved oxygen level that is not too low requires water quality-based
effluent limits on different pollutants, primarily biochemical oxygen
demand.  The Guidance's implementation procedures would require
significant revision to accommodate these situations.  Third, to achieve
a dissolved oxygen level that is not too high generally requires
addressing eutrophication problems, which are not addressed by the
Guidance, or ensuring that a dam or other hydrologic modification does
not induce oxygen supersaturation through turbulent mixing, which may or
may not involve development of water quality-based effluent limits for
dissolved oxygen.  The Guidance's implementation procedures would
require significant revision to accommodate these situations.
Therefore, EPA believes it is not appropriate to apply the Guidance
methodologies and implementation procedures in the case of dissolved
oxygen.

      EPA considered the possibility of developing more appropriate
criteria development methodologies and implementation procedures for
dissolved oxygen.  EPA concluded that the effort to develop such
methodologies and procedures would not be possible or administratively
feasible within the time EPA had available to complete the final
guidance.

          Dissolved solids and salinity.  The Guidance methodologies and
implementation procedures are not scientifically and technically
appropriate for developing and applying water quality criteria and
values for dissolved solids and salinity.  First, the Guidance
methodologies for numeric criteria and values and the corresponding
implementation procedures are designed to be used with specific
chemicals or related congeners.  Dissolved solids and salinity are not
specific chemicals, but the combined effect of several unrelated
substances.  Second, the methodologies and implementation procedures are
designed to be used with chemicals that exert an adverse effect as the
concentration of the chemical becomes too high.  Although dissolved
solids and salinity can produce adverse effects at high levels, they can
also produce more subtle effects at lower concentrations that affect the
ecological balance of an ecosystem.  For example, increased dissolved
solids and salinity can favor aquatic species that are not native to an
ecosystem and can disrupt the community structure.  The Guidance

-------
54     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      methodologies and implementation procedures do not address these types
      of end points.  Third, the establishment of water quality criteria for
      dissolved solids and salinity often involves site-specific
      considerations that .are not addressed by the final Guidance.  Therefore,
      EPA believes it is not appropriate to apply the Guidance methodologies
      and implementation procedures in the case of dissolved solids and
      salinity.

                Hydrogen sulfide and sulfide.   EPA agrees.with comments that
      the Guidance methodologies and procedures are scientifically appropriate
      for developing and implementing criteria and values for hydrogen sulfide
      and sulfide, and has therefore removed hydrogen sulfide and sulfide from
      Table 5 in the final Guidance.  These pollutants were originally listed
      by the Initiative Committees because they were believed to have only
      organoleptic effects.  In response to comments, EPA reviewed the
      available scientific information, including EPA's Quality Criteria for
      Water, July 1976, and found that the pollutants have adverse effects
      that can be addressed by the Guidance methodologies and procedures.  EPA
      is therefore removing them from Table 5.

            --  pH.  The Guidance methodologies and implementation procedures
      are not scientifically and technically appropriate for developing and
      applying water quality criteria and values for pH.  First, the Guidance
      methodologies for numeric criteria and values and the corresponding
      implementation procedures are designed to be used with chemicals that
      exert an adverse effect as the concentration of the chemical becomes too
      high, while pH exerts adverse effects on aquatic life if it is either
      too low or top high.  The Guidance methodology to develop aquatic life
      criteria and values, and implementation procedures such as those for
      total maximum daily loads and loading limits do not accommodate such a
      pollutant.  Second, the implementation procedures in the final Guidance,
      including the procedures for total maximum daily loads, reasonable
      potential, and loading limits, are generally designed to develop water
      quality-based effluent limits for the same pollutant that has reasonable
      potential to exceed water quality standards.  Achieving a pH level that
      is not too low or too high, however, might require water quality-based
      effluent limits on different pollutants, such as specific acids, bases,
      or buffering compounds that affect the overall pH of the effluent and
      receiving water.  The Guidance's implementation procedures do not
      accommodate these situations.  Third, the implementation procedures for
      total maximum daily loads are designed for pollutants that are expressed
      as an ordinary concentration, while pH is expressed as the negative
      logarithm of the hydrogen ion concentration.  Therefore, EPA believes it
      is not appropriate to apply the Guidance methodologies and
      implementation procedures in the case of pH.

            EPA considered the possibility of developing more appropriate
      criteria development methodologies and implementation procedures for pH.
      EPA concluded that the effort to develop such methodologies and
      procedures would not be possible or administratively feasible within the
      time EPA had available to complete the final Guidance.

            By listing pH among the pollutants in Table 5, EPA does not  intend
      to exclude pH as a water quality factor in calculating criteria and
      values for other pollutants.  Where appropriate, pH should continue to
      be used as such a factor.

                 Phosphorus.  The Guidance methodologies and implementation
      procedures are not scientifically and technically appropriate for
      developing and applying water quality criteria and values for
      phosphorus.  The Guidance methodologies for numeric criteria and values
      and the  corresponding  implementation procedures  are designed to be used
      with  chemicals that  exert an adverse effect, while  the primary
      environmental concern with phosphorus is its role as a nutrient  in

-------
                           Section II: Regulatory Requirements                        55
      accelerating eutrophication of lakes and other waterbodies.  The
      Guidance methodologies and procedures do not reflect eutrophication as
      an end point.  Therefore, EPA believes it is not appropriate to apply
      the Guidance methodologies and implementation procedures in the case of
      phosphorus.

                Temperature.  The Guidance methodologies and implementation
      procedures are not scientifically and technically appropriate for
      developing and applying water quality criteria and values for
      temperature.  The Guidance methodologies for numeric criteria and values
      and the corresponding implementation procedures are designed to be used
      with chemicals that exert an adverse effect as the concentration of the
      chemical becomes too high, while both high and low temperatures can have
      both beneficial and adverse effects, depending on aquatic species, stage
      of life cycle, and other chemical and physical factors.  Furthermore,
      adverse effects from temperature often arise from abrupt spatial and
      temporal differences in temperature, rather than from temperature
      extremes.  The Guidance methodologies and implementation procedures do
      not take these types of factors into account.  Therefore, EPA believes
      it is not appropriate to apply the Guidance methodologies and
      implementation procedures in the case of temperature.

            By listing temperature among the pollutants in Table 5, EPA does
      not intend to exclude temperature as a water quality factor in
      determining criteria and values for other pollutants.  Where
      appropriate, temperature should continue to be used as such a factor.

            --  Total and suspended solids, and turbidity.  The Guidance
      methodologies and implementation procedures are not scientifically and
      technically appropriate for developing and applying water quality
      criteria and values for total and suspended solids, and turbidity.
      First, the Guidance methodologies for numeric criteria and values and
      the corresponding implementation procedures are designed to be used with
      specific chemicals or related congeners.  Total and suspended solids,
      and turbidity are not specific chemicals, but the combined effect of
      several unrelated substances.  Second, the aquatic life methodology and
      corresponding implementation procedures are designed to be used with
      chemicals that exert a direct adverse effect on aquatic life, while
      these pollutants exert their adverse effects through both direct and
      indirect means.  For example, turbidity not only directly clogs gills,
      mats eyes, and impairs respiration, but also indirectly affects aquatic
      life by impairing visibility necessary for finding food, altering water
      temperature, and reducing primary productivity that serves as the base
      of the aquatic food chain.  Total and suspended solids produce many of
      the same effects as turbidity, and also affect reproduction of some
      aquatic life by blanketing spawning areas and smothering eggs in stream
      beds.  The Guidance's aquatic life methodology and implementation
      procedures do not reflect the indirect effects described above.  Third,
      the Guidance implementation procedures are generally designed to develop
      water quality-based effluent limits for the same pollutant that has
      reasonable potential to exceed water quality standards, while to achieve
      a turbidity level that protects aquatic life will require water quality-
      based effluent limits on different pollutants, such as total or
      suspended solids.  Therefore, EPA believes it is not appropriate to
      apply the Guidance methodologies and implementation procedures in the
      case of total and suspended solids and turbidity.

      EPA does not agree with the comments to add additional pollutants to
Table 5.  Some comments suggested adding "non-toxic" pollutants, including
common inorganic constituents such as chloride, sulfate, sodium, and calcium.
EPA has decided not to add such pollutants.  First, it would likely be
difficult to develop an operational definition of "non-toxic."  For example,
each of the inorganic constituents suggested by commenters has some degree of
adverse effects.  Furthermore, the effect of a pollutant is a function of both

-------
56     Water Quality Guidance for the Great Lakes System - Supplementary Information Document

toxicity and exposure, not toxicity alone.  Listing a pollutant in Table 5
solely on the basis of low toxicity, however defined, would not be
appropriate, because even pollutants with relatively low toxicity could have
adverse effects in the environment if present in high concentrations.  Second,
EPA believes that the final Guidance's criteria development methodologies and
implementation procedures are designed to develop appropriate criteria, total
maximum daily loads, and water quality-based effluent limits over a wide -range
of toxicities, including pollutants with lower toxicities.  For example, under
procedure 5 of appendix F, a discharge containing relatively small amounts of
a pollutant with low toxicity might not have reasonable potential to exceed
water quality standards, and therefore would not require water quality-based
effluent limits.  If the pollutants did have reasonable potential to exceed
water quality standards, the Guidance methodologies and implementation
procedures would be used to develop water quality-based effluent limits.  If a
pollutant identified in the future has unique "non-toxic" features making one
or more of the Guidance methodologies or procedures scientifically
indefensible in a particular situation, then § 132.4(h) of the final Guidance
could be used generally to  exempt that pollutant from selected provisions of
the Guidance.  Therefore, for all these reasons, it is neither appropriate nor
necessary to add "non-toxic" pollutants to Table 5.

      The scientific defensibility exclusion discussed in section II.C.6 of
this document is designed to achieve a purpose similar to although more
limited than Table 5.  EPA believes both provisions are useful and appropriate
in different circumstances.  The scientific defensibility exclusion is
available for a pollutant for which the State or Tribe demonstrates that a
methodology or procedure in this part is not scientifically defensible.  It
enables Great Lakes States and Tribes to apply an alternative methodology or
procedure acceptable under 40 CFR part 131 when developing water quality
criteria or implementing narrative criteria, or to apply an alternative
implementation procedure that is consistent with all applicable Federal,
State, and Tribal laws.  This provision would be used to provide exclusions
for reasons which are currently unidentified, or not broadly applicable.  The
Table 5 exclusions, in contrast, are useful for the 14 pollutants where valid
scientific and technical reasons for not regulating them under such provisions
are already available and broadly applicable as discussed above.  The use of
the Table 5 exclusion in § 132.4(g) promotes administrative efficiency and
conserves resources of States, Tribes, and dischargers by not repeating the
analysis of scientific defensibility for each occurrence or discharge of these
pollutants.

      c.    Final Guidance.  The final Guidance retains the proposed Table 5
of part 132.  Table 5 has been renamed "Pollutants Subject to Federal, State,
and Tribal Requirements" to provide a more accurate description.

      The final Guidance also retains the proposed provisions for excluding
these pollutants from certain specified provisions of part 132, but not from
all requirements in Federal, State, or Tribal water quality programs.
Sections 132.4(b), 132.4(c), and 132.4(h){l) of the final Guidance provide
that States and Tribes do not need to apply the methodologies for development
of criteria and values in appendixes A through D for pollutants in Table 5,
but instead must apply any methodologies and procedures acceptable under 40
CFR part 131 when developing water quality criteria or implementing narrative
criteria for these pollutants.  Sections 4(e)(2) and 132.4(h)(2) of the final
Guidance provide that States and Tribes do not need to apply implementation
procedures 1, 2, 3, 4, 5, 7, 8, and 9 of appendix F of part 132 for pollutants
in Table 5, but any alternative procedures used instead must be consistent
with all applicable Federal, State, and Tribal laws.

      EPA recognizes that some of the methodologies or implementation
procedures of the final Guidance  could technically be applied in establishing
controls on the discharge of some or all of  the pollutants listed in Table 5.
For example, procedure 2  (Variances from Water Quality Standards) could be
applied in determining whether to grant a variance from water quality

-------
                           Section II: Regulatory Requirements                        57
standards to a point source discharger of any pollutant.  Great Lakes  States
and Tribes may apply such methodologies or implementation procedures in
establishing water quality criteria or controls on the discharge of any
pollutant in Table 5 of the proposed Guidance.

      In applying alternative methodologies and procedures in the above
situations, States and Tribes are encouraged to apply technical guidance that
EPA has issued where appropriate to assist in developing and implementing
consistent water quality-based controls for these pollutants.

      Section 132.4(f) provides no exclusion for Table 5 pollutants from the
antidegradation provisions of the Guidance.  The antidegradation policy in
appendix E, however, provides for different requirements for BCCs than for
non-BCCs.  The antidegradation policy is discussed in section VII of this
document.

      For reasons discussed above, hydrogen sulfide and sulfide have been
removed from Table 5.  The other 14 pollutants listed in the proposal  have
been retained.  No new pollutants have been added.

      As discussed in the Comments section above, EPA is concerned about the
possibility of inconsistencies among State programs in the Great Lakes basin
in addressing ammonia and chlorine.  For example, there are differences in
numeric criteria adopted by different States for these two pollutants,
especially for ammonia.  There has not been a systematic evaluation, however,
of the implementation procedures used in applying the criteria in the  Great
Lakes States, nor has there been an evaluation of resulting water quality-
based effluent limits in NPDES permits.  In part this is because of the
complexity of comparing implementation procedures and effluent limits  from
State to State where States have flexibility in implementation of water
quality programs, including the authority to be more stringent than the CWA
and implementing regulations.

      In order to determine whether significant inconsistency exists in the
level of protection^ of aquatic life from adverse effects of ammonia and
chlorine, and to take corrective action if warranted, EPA will work with the
States in reviewing water quality standards and implementation of those
standards for ammonia and chlorine in the Great Lakes basin as part of EPA's
responsibilities under section 303(c) of the CWA.  Under section 303(c),
States and Tribes must review and revise their water quality standards every
three years, and EPA must review and approve or disapprove such standards.
For the next two triennial review cycles, EPA will give special attention to
working with the States with respect to standards and implementation
procedures affecting ammonia and chlorine in the Great Lakes basin to  ensure
that any inconsistencies are addressed.  The review will include the following
steps:

         EPA will coordinate with the States and Tribes in reviewing whether
water quality criteria for ammonia and chlorine have been adopted in all
appropriate waters.  Under 40 CFR part 131, States and Tribes must adopt
criteria necessary to protect designated uses.

         EPA will place existing toxicological data on ammonia and chlorine in
the GLI Clearinghouse, including the EPA criteria documents and supplementary
information issued in 1984 and 1992 for ammonia, and in 1985 for chlorine,
under section 304(a)  of the CWA.  It will include any new data or additional
valid data that is hot be reflected in the above documents.

         To assist the States in ensuring that all States adopt adequate and
consistent criteria,  EPA will review numeric criteria adopted by the States
and Tribes for ammonia and chlorine to determine whether the criteria  meet all
applicable requirements and are based on consideration of current available
scientific information, including data in the GLI Clearinghouse.  The  review

-------
58     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

will include consideration of any site-specific modification procedures used
by the States and Tribes.

         EPA will also assist the States by reviewing implementation
procedures in use by States and Tribes to develop total maximum daily loads,
wasteload allocations, and water quality-based effluent limits for these two
pollutants as-part of EPA's ongoing review of TMDLs and NPDES programs and
permits.  To assist in this review, State and Tribal procedures will be
compared with existing national policies and guidance.

      -  EPA will use the results of the review in fulfilling its
responsibilities to review the triennial submissions of State and Tribal water
quality standards.  If a State or Tribe does not adopt and submit appropriate
criteria for these pollutants, EPA will take appropriate steps including
disapproval of the water quality standards, and promulgation of Federal
standards for affected waters to ensure consistent and adequate standards for
all the Great Lakes System.

      The above review steps differ from triennial reviews normally conducted
by EPA in five ways.  First, they specify two pollutants for in-depth
evaluation.  That is, EPA plans to evaluate more detailed supporting
information concerning State and Tribal standards for these pollutants.
Second, EPA will give special attention to identifying and evaluating
differences among State and Tribal programs for these pollutants.  Third, EPA
will coordinate reviews of procedures used in developing total maximum daily
loads and water quality-based effluent limits with the triennial reviews, even
though they are not a part of the review and approval process of 40 CFR part
131.  Fourth, EPA will assist State and Tribal development of water quality
standards by providing information through the GLI Clearinghouse.  Fifth, the
review will be specific to one drainage basin, the Great Lakes basin.  Because
of differences in the timing of different States' and Tribes' review cycles,
and because of the complexities of the interstate comparisons, EPA expects the
full review of ammonia and chlorine standards to require up to two triennial
review cycles.  EPA will make every effort to complete the review no later
than 2003.

      As part of the consultation with the U.S. Fish and Wildlife Service
 (FWS) under section 7 of the Endangered Species Act concerning the Guidance,
the FWS raised concerns that inconsistencies in standards and how they are
applied for ammonia and chlorine may  be adversely affecting endangered or
threatened species in the Great Lakes basin.  In implementing the above
review, EPA will consult with the FWS concerning EPA' s approval of State and
Tribal water quality standards under  section 303(c) of the CWA.  In addition,
EPA will invite and encourage the States and Tribes to participate actively in
the consultations.  EPA and the FWS believe that the review described above,
together with the involvement of  FWS  in consultations with EPA and the States
and Tribes, will serve on an expedited basis to minimize inconsistencies in
controls for ammonia and chlorine.

6.    Scientific Defensibilitv Exclusion

      a.    Proposal.  The proposed Guidance at  § 132.4(g) provides that the
Great Lakes States and Tribes need not apply the proposed criteria
methodologies and implementation  procedures to any pollutant for which the
regulatory authority demonstrates that one  or more procedures in the Guidance
are not scientifically  defensible.  The reason for this exclusion  is that
there may be pollutants  identified in the  future for which  some of  the
methodologies or procedures  in the final Guidance may not be technically
appropriate.

      EPA specifically  invited comment on  whether the final  Guidance  should
specify minimum  requirements  for  use  of this exclusion, demonstration
elements, or procedures for  EPA review of  these  submissions.

-------
                           Section II: Regulatory Requirements                        59
      b.    Comments.  Some commenters recommended removal of the language  in
§ 132.4(g) allowing pollutants to be excluded from Guidance procedures and
State adoption based simply on a State demonstration, preferring that EPA
formally add such pollutants to Table 5.

      Several commenters suggested that no minimum requirement should be
specified for an exclusion demonstration, and that the Guidance should strive
to maintain the flexibility needed for consideration of all site-specific
possibilities.

      EPA has determined that the scientific defensibility exclusion proposed
in § 132.4(g) is necessary and appropriate to include in the final Guidance.
Based on long experience in implementing the CWA, EPA has found that no matter
how carefully a regulatory requirement is planned, there is no way to
anticipate all possibilities.  In the water quality standards program in
particular, new scientific information inevitably arises that cannot always be
accommodated within existing program structures.  Eliminating the exclusion
would likely require future rulemaking by EPA, States, and Tribes to adjust to
new scientific information.  Because rulemaking is often a long process,
eliminating the exclusion would make it very difficult to adapt quickly to new
information when necessary.

      EPA believes that the exclusion will be implemented in a way that will
maintain a reasonable consistency in State and Tribal programs.  As explained
in the preamble to the proposed Guidance (58 PR 20843), EPA Regional Offices
will work with the States and review State demonstrations during water quality
standards submissions, TMDL approvals, and NPDES program implementation.
Through this process, the Regional Offices and States will ensure that the
scientific defensibility exclusions, if approved, will be consistent with the
Guidance, other EPA regulations, and current EPA policy and guidance.

      EPA agrees with comments that implementation of the exclusion should
proceed without further detailed guidance on minimum requirements,
demonstration elements, or review procedures.  Since the nature of the
exclusions cannot be predicted in detail, such guidance would need to be
highly speculative and likely could not anticipate all the circumstances.  EPA
anticipates that the States and EPA Regions will be able to address most
situations in a reasonable way using professional judgment.

      c.    Final Guidance.  For the reasons above, EPA has retained the
exclusion for scientific defensibility in the final Guidance.

      During its review, EPA discovered that there may have been ambiguity
concerning the scope of the exclusion.  EPA intended that the exclusion be
limited to each specific element of the Guidance that was demonstrated to be
inappropriate if applied to a specific situation, and not to all other
elements.  For example, EPA intended that a pollutant for which a State
demonstrated that the Guidance's methodology for development of aquatic life
criteria could not be applied should still be subject to other provisions of
the Guidance adopted into State or Tribal law, such as other criteria
methodologies, site-specific modification procedures, implementation
procedures,  and antidegradation policies.  Accordingly, in order to improve
the clarity of the scientific defensibility exclusion, EPA has created a new
§ 132.4(h) with clarified wording, and deleted the corresponding text from
§ 132.4(g).   The new wording, similar to the proposal, is intended to allow
States or Tribes to use an alternative methodology or procedure that
corresponds to the methodology or procedure found to be scientifically
indefensible.

7.    Wet Weather Exclusion

      a.    Proposal.  The proposed Guidance allowed, where appropriate, but
did not require,  the Great Lakes States and Tribes to adopt provisions
consistent with any of the proposed implementation procedures for establishing

-------
60     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

controls on wet weather point source discharges.  Proposed § 132.4(e)(l)
provided that "any procedures applied in lieu of these implementation
procedures shall be consistent with all applicable Federal, State, and  Tribal
requirements."  Accordingly, even though permitting authorities would not be
required to apply State of Tribal requirements consistent with the part 132
implementation procedures in establishing controls on wet-weather point source
discharges, all permits would still be required to contain any limitations and
conditions necessary to ensure compliance with the CWA and implementing
Federal and State regulations.  The proposed wet weather exclusion did  not
apply to the development of water quality criteria and values, since water
quality criteria and values apply to waters of the Great Lakes System
regardless of the source of the pollutant.

      EPA proposed this exclusion from the Guidance implementation procedures
because they do not address the significant differences that can exist  between
wet weather point source discharges and dry-weather point source discharges.
The preamble to the proposed Guidance discussed these differences, including
the significant variability that exists during and immediately following wet
weather events in rates, durations, and composition of wet weather flows  (58
FR 20840-42).

      b.    Comments.  Several comments were received supporting retaining the
wet weather exclusion in the final Guidance, including comments that
application of stringent provisions in the Guidance would in essence prohibit
all combined sewer overflows.

      A few comments, while not opposing the exclusion on a temporary basis,
urged EPA to develop a mechanism for addressing pollution during wet weather
point source discharges, since commenters believe that wet weather discharges
contribute significant loadings of toxic pollutants to the Great Lakes  System
and must be stemmed.

      No comments on the definition of "wet weather point source" itself were
received, although some comments expressed concerns about whether specific
types of discharges were or were not excluded from the Guidance implementation
procedures.

      EPA agrees with the comments that the exclusion should be retained, for
the reasons given in the preamble to the proposal  (58 FR 20840-42).  EPA also
agrees that mechanisms are needed for addressing pollution during wet-weather
events.  Accordingly, procedure 3.B.8 of appendix F has been clarified  to
provide that States and Tribes must consider pollution resulting from wet
weather events, where appropriate, when developing TMDLs.  States and Tribes
retain flexibility, however, in determining how to account for such discharges
and are free to choose the specific procedures they deem most appropriate.

      EPA recognizes that the proposed definition of "wet weather point
source" should be clarified.  As a result, the technical modifications  to the
definition have been made to clarify which types of discharges are included in
the definition, to clarify which types of discharges are excluded, and  to
delete an unnecessary definition of combined sewer overflow.  The definition
has also been modified to use terminology from the existing definition  of
point source contained in § 122.2, for the convenience of the reader.

      c.    Final Guidance.  Section  132.4(e) of the final Guidance retains
the proposed wet weather exclusion.

      The exclusion applies to "wet weather point sources" as defined in  §
132.2.  For the reasons above, the definition in the final Guidance has been
modified to read as follows:

      Wet weather point source means  any discernible, confined and discrete
      conveyance from which pollutants are, or may be, discharged  as  the
      result of a wet weather event.  Discharges from wet weather point

-------
                           Section II: Regulatory Requirements                        61
      sources shall" include only: discharges of storm water from  a municipal
      separate storm sewer as defined at 40 CFR 122.26(b)(8);  storm water
      discharge associated with industrial activity as defined at 40  CFR
      122;26(b)(14); discharges of storm water and sanitary wastewaters
       (domestic, commercial, and industrial) from a combined sewer overflow;
      or any other  stormwater discharge for which a permit is  required under
      s.ection 402 (p) of the CWA.  A storm water discharge associated  with'
      industrial activity which is mixed with process wastewater  shall not be
      considered a  wet weather point source.

      A combined sewer overflow (CSO) is the discharge from a  combined sewer
system  (CSS) at a point prior to the headworks of a publicly owned treatment
works  (POTW) treatment plant.  A CSS is a wastewater collection system owned
by a State or municipality  (as defined by section 502(4) of the CWA)  which
conveys sanitary wastewaters  (domestic, commercial and industrial wastewaters)
and storm water through a single-pipe system to a POTW treatment  plant  (as
defined in 40 CFR 403.3(p).  CSOs consist of mixtures of domestic sewage,
industrial and commercial wastewaters, and storm water runoff.  CSOs  are point
sources subject to  NPDES permit requirements including both technology-based
and water quality-based requirements of the CWA.  The NPDES permit
requirements for CSOs are explained in EPA1s Combined Sewer Overflow  Control
Policy  (59 FR 18688', April 19, 1994).

      EPA would like to clarify that although Great Lakes States  and  Tribes
are not required to apply Guidance procedures in establishing  controls on the
discharge of pollutants by wet weather point sources  (except for  procedure
3.B.8 of appendix F which provides that TMDLs must consider discharges from
wet weather events, where appropriate), they may nevertheless  choose  to  do so.
Furthermore, the use of such procedures as variances and compliance schedules
are available under existing State programs pursuant to parts  122 and 131.

8.    Bioaccumulative Chemicals of Concern

      a.    Proposal.  The proposed Guidance identified a class of highly
bioaccumulative pollutants, termed BCCs, for special attention.   The  Great
Lakes Initiative Steering Committee believed that every reasonable effort
should be made to reduce loadings of all BCCs, because these pollutants  tend
to persist throughout the Great Lakes ecosystem and have a propensity to
bioaccumulate in the food chain, and have been associated with serious and
systemwide impacts.

      The BCCs were defined in general as those chemicals which bioaccumulate
in aquatic organisms by a human health bioaccumulation factor  (BAF) greater
than 1000, in accordance with the BAF methodology proposed in  appendix B to
part 132.  BCCs would include, but not be limited to the pollutants identified
as BCCs in Table 6.

      In the proposed Guidance, a pollutant found to be a BCC  would be subject
to the following special provisions upon State or Tribal adoption or  EPA
promulgation:

         Under the methodology for deriving non-cancer human health criteria,
the proposed Guidance assumed a relative source contribution (RSC) from
surface water pathways  (water and fish) of 80 percent for BCCs, to at least
partially account for exposures through other pathways.  For non-BCCs, the RSC
assumption was 100 percent.  Accordingly, the proposed methodology for
deriving the human health non-cancer criteria would be 20 percent more
conservative for BCCs than for non-BCCs.

         Under the antidegradation procedures of the proposed  Guidance,  any
action by a discharger that results in an increase in the baseline rate  of
mass loading of a BCC would be considered a significant lowering  of water
quality, and thereby trigger an antidegradation review.

-------
62     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      -  The proposed Guidance provided, in general, that mixing zones for
existing discharges'of BCCs would be eliminated within 10 years after
publication of the final Guidance.  For new sources of BCCs, mixing zones
would not be allowed beginning on the date of publication of the final
Guidance.

         The proposed Guidance provided that for BCCs, permit authorities must
generate or cause the discharger to generate the data necessary to derive Tier
II values to protect aquatic life in certain circumstances that would not
apply to non-BCCs.

         The proposed Guidance would establish a specific process to regulate
the discharge of any pollutant for which the water quality based effluent
limit was below a level that could be analytically quantified.  If the
pollutant were a BCC, the discharger would also have to determine if the BCC
were bioconcentrating or bioaccumulating in fish exposed to the effluent.  If
such monitoring revealed unacceptable accumulation in fish tissue, additional
actions would be required of the discharger.

      b.    Comments.  EPA received a number of comments suggesting that
establishing special provisions for BCCs is not warranted because
bioaccumulation of pollutants is already taken into account in the proposed
methodologies for developing criteria fully protective of the Great Lakes
basin.  EPA also received comments supporting the BCC provisions,  stating that
such provisions are necessary because of the long retention times of the Great
Lakes, and because of the widespread adverse effects to human health and
wildlife from these types of pollutants.

      A number of comments recommended that if special provisions were to be
established for BCCs, the BCCs should be limited to substances that have been
shown through environmental monitoring, including sampling of fish tissues, to
be present at concentrations of concern.  Other comments stated that such an
approach would be overly simplistic, and would not provide a preventive
approach for these pollutants.

      A large number of comments questioned various aspects of the definition
of BCCs.  Many were concerned that the definition did not include the concepts
of toxicity and persistence.  Of these, some comments recommended deleting
pollutants from the BCC list that are not strongly toxic or do not persist for
long periods even if they are highly bioaccumulative, while others supported
adding pollutants that are persistent even if they are not highly
bioaccumulative.  Many comments were received concerning the proposed BAF
cutoff level of 1000 for defining BCCs: many stated that the proposed level
was arbitrarily selected, or based on inadequate scientific analysis; some
felt the proposed cutoff was too low; some felt it was too high; and some
supported the proposed level.  Some comments recommended that only field-
measured BAFs, not predicted BAFs, should be used in determining BCCs; others
supported the proposed use of both field-measured and predicted BAFs.  Some
comments recommended considering the sediment route of exposure in
establishing BCCs.

      Some comments recommended that additional chemicals should not be
subject to the special provisions for BCCs until after formal public comment
in the Federal Register including review of the BAF.

      After careful consideration of the comments, EPA continues to believe
that the special provisions for BCCs are warranted.  EPA's continued support
of the special emphasis on BCCs parallels the position of the Great Lakes
States as initially expressed by State  representatives on the Initiative
Committees.  EPA believes that these special provisions for BCCs are a
reasonable approach to address the issue of persistent bioaccumulative
pollutants in the Great Lakes System, for the following reasons.  First,
persistence of toxic pollutants is a major  concern  in an aquatic system like
the Great Lakes, for the reasons discussed  in the preamble to the proposed

-------
                           Section II: Regulatory Requirements                        63
Guidance  (58 FR 20820-23, 58 FR 20844-45), and in section I.B of this
document.  Persistence is especially problematic for chemicals that are highly
bioaccumulative, because the most important exposure pathway for these
chemicals for humans and wildlife in the Great Lakes System is consumption of
fish and other aquatic organisms.  Persistent bioaccumulative chemicals will
result in high exposures to humans and wildlife for a long time to come.

      Second, proposed procedures 3A and 3B in appendix F for development of
total maximum daily loads  (TMDLs) envisioned  predominant use of a simple,
steady-state mass balance approach for both bioaccumulative and non-
bioaccumulative pollutants.  Although the final Guidance would allow use of
other approaches for developing TMDLs, EPA expects that the steady-state mass
balance approach will be the approach used in a majority of waters within the
Great Lakes System because of its ease of implementation.  The steady-state
mass balance approach is a method used to approximate the mass of pollutants
within a water body.  This approach assumes that the input of mass into the
system (e.g., through point and nonpoint source loadings, atmospheric
deposition, groundwater seepage) equals the loss of mass from the system plus
any losses due to transformation of mass within the system.  In other words,
the approach assumes that no mass accumulates in the system.  This method
provides a first approximation of allowable loading allocations.

      For persistent bioaccumulative pollutants, however, approximation based
on a steady-state mass balance approach will likely not be accurate.  As
discussed in section I.A of the preamble to the proposed Guidance  (58 FR
20822), there are significant interactive physical, chemical, and biological
processes that affect the long-term behavior of persistent bioaccumulative
pollutants in the Great Lakes System, resulting in fairly common occurrences
where such pollutants do accumulate in the system.  Additionally, although the
phased approach to TMDLs discussed in section VTII.C of this document
recommends subsequent monitoring to identify any shortcomings in the chosen
control approach, this approach may present a significant risk of allowing
persistent bioaccumulative pollutants to concentrate in the ecosystem above
ambient criteria levels before the control approach can be evaluated and
revised as necessary.  EPA believes the costs of future remediation actions to
address BCCs would be significantly more expensive than efforts to control the
BCCs before they enter the environment.  Accordingly, additional controls
intended to prevent concentrations of persistent bioaccumulative pollutants
from increasing to the level of criteria concentrations in Great Lakes waters
are reasonable.

      In the proposal, EPA requested comment on issues concerning the details
of the proposed special provisions for BCCs.  After analyzing those issues and
the comments received, EPA has modified several of the provisions in ways that
may in some cases reduce costs for the regulated community without
significantly increasing the risk from BCCs.  EPA believes that with these
modifications the provisions for BCCs will continue to address the concerns of
the Initiative Committees for controlling the discharges of BCCs.  These
modifications include:

          Modifying the methodology for deriving non-cancer human health
criteria to assume a relative source contribution  (RSC) from surface water
pathways of 80 percent for all pollutants, not just BCCs.  Therefore, the 80
percent RSC still applies to BCCs, but is no longer a "special" provision for
BCCs.  The reasons for this change are discussed in section V.C.5 of this
document.

          Changing the antidegradation provisions to replace numeric existing
effluent quality-based ("EEQ")  limits as a means of implementing
antidegradation for BCCs with a narrative description of the types of
activities that will trigger an antidegradation review, and to provide greater
flexibility in the implementation, demonstration and decision components.
These modifications are discussed further in section VILA of this document.

-------
64     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

          Providing an opportunity for dischargers to retain a limited mixing
zone for a BCC under certain limited circumstances if the discharger is
complying with all applicable technology-based and water quality-based
requirements and has reduced its discharge of the BCC to the maximum extent
possible, but is unable to meet water quality standards in the absence of a
mixing zone.  This modification is discussed further in section VIII.C.4 of
this document.

          Deleting the provision for dischargers to determine if BCCs were
bioconcentrating or bioaccumulating in fish exposed to the effluent in
situations where the water quality-based effluent limit was below a level that
could be analytically quantified.  This modification is discussed further in
section VIII.H.4 of'this document.

      EPA could not accept the concept put forth by commenters that pollutants
should not be regulated as BCCs until they are shown to be present at
concentrations of concern in the Great Lakes System.  As discussed above, EPA
is concerned about preventing concentrations of BCCs from increasing to the
level of criteria concentrations in Great Lakes waters.  The regulatory
approach, suggested by some commenters, that would not trigger preventive
action until some measurable concentration resulting in adverse conditions is
reached in the environment would not be effective in addressing this concern,
particularly because of the difficulties of measuring these pollutants at
levels of concern in the environment.  As discussed further in sections VII.B
and VIII.C.4 of this document, the special provisions for BCCs in the final
Guidance will take full effect over the next twelve years (two years for
State/Tribal adoption or promulgation,  plus ten year phase-in period).  A
program requiring systematic environmental monitoring followed by a regulatory
process to designate BCCs could significantly delay implementation of these
provisions and allow build-up of new persistent, bioaccumulative pollutants in
the Great Lakes System.  The risks to the Great Lakes ecosystem of such a
delay are too great to warrant such an approach.

      EPA agrees with comments that toxicity and persistence should be
included in the definition of BCC.  As discussed in the preamble to the
proposed Guidance (58 FR 20807, 58 FR 20820), toxic pollutants that are
persistent and bioaccumulate are of particular concern in the Great Lakes
System.  With regard to toxicity, EPA has amended the definition of BCC in the
final Guidance to provide that a chemical must also have "the potential to
cause adverse effects" in order to be a BCC.  Under this revised definition,
if data become available showing that a chemical that otherwise meets the BCC
definition does not have the potential to cause an adverse effect, State or
Tribal authorities would not have to apply any adopted or promulgated
provisions for BCCs to that chemical.  EPA expects that very few pollutants,
if any, would be excluded as BCCs in this way, since most substances have
potential adverse effects at some level of concentration.  Nevertheless, if
scientifically valid experimental evidence is provided which demonstrates that
a chemical has no potential for producing adverse effects, then a State or
Tribe could find that the special provisions for BCCs need not be applied for
that pollutant.

      EPA has also made changes to the final definition of BCCs with regard to
persistence.  As discussed in the preamble to the proposed Guidance  (58 FR
20821), the proposed definition of BCC was based principally on the concept of
bioaccumulation.  EPA and the Initiative Committees had considered including
persistence in the definition, but found that data were not systematically
available concerning persistence.  That is,  systematic data were not found
concerning the cumulative effect of relevant fate and effect processes for the
full range of specific pollutants in the Great Lakes System under field
conditions, or under laboratory conditions which had been field correlated and
verified.  Upon review  of comments received, and after reevaluation  of
available data, EPA believes that it is possible, though unlikely, that very
specific data might become available showing that a toxic, bioaccumulative
pollutant is very short-lived in  the aquatic environment.  For example,  the

-------
                           Section II: Regulatory Requirements                        65
pollutant may be highly volatile and escape to the atmosphere.  In this
situation, it would not be necessary to have a full range of data on the
cumulative effect of relevant fate and effect pathways for comparison with
other pollutants.  Rather, the specific data could be used to determine that
the pollutant is not persistent.  For this reason, EPA has amended the
definition of BCC in the final Guidance to include the qualification that
"chemicals with half-lives of less than eight weeks in the water column,
sediment, and biota are not BCCs."  The definition conforms with the Great
Lakes Water Quality Agreement, which defines persistent toxic substances as
any substance with a half life in water greater than eight weeks.

      EPA does not agree with comments that highly persistent pollutants
should be subject to the special provisions developed for BCCs even if they
are not highly bioaccumulative.  First, as discussed above, the special
provisions for BCCs in the final Guidance are designed to ensure that exposure
to humans and wildlife from BCCs is minimized.  The most likely fate for many
persistent but not highly bioaccumulative pollutants is to be deposited in
sediments, where they will likely remain for long periods.  The potential for
exposure to humans and wildlife while possible is diminished because the
contaminants do not bioaccumulate and in many cases are buried in the
sediments.

      Second, the threat these pollutants pose to benthic and other aquatic
organisms that come in direct contact with the sediment is being addressed
through other approaches.  Benthic organisms are represented in the
methodology for development of criteria to protect aquatic life.  In addition,
the potential toxicity to benthic organisms from desorption of pollutants from
sediment is addressed in existing State programs on a case-by-case basis
through implementation of narrative criteria.  EPA has also developed a
methodology which, when finalized, will be available to assist States and
Tribes in addressing the potential toxicity to benthic organisms more
systematically.  In January 1994 EPA published a notice announcing the
availability of proposed national sediment quality criteria for the protection
of benthic organisms, guidelines for deriving these criteria on a site-
specific basis, and the technical basis for deriving the criteria (59 FR 2652,
January 18, 1994).  EPA is analyzing the comments received in response to the
notice, and will be developing final sediment quality criteria based on the
analysis.  When the.methodology is finalized, it will be available for EPA,
States, and Tribes to develop sediment quality criteria for the protection of
benthic organisms.  This approach is scientifically more appropriate for the
control of persistent but not highly bioaccumulative pollutants than the
special provisions developed for BCCs in the final Guidance.  The BCC
provisions were designed to reduce loadings, not to specifically achieve
protective levels of contaminants in sediments.

      Third, it is reasonable to limit application of the special BCC
provisions to highly bioaccumulative pollutants.  The special provisions for
BCCs and the methodology for defining these pollutants were developed by the
senior water program managers in the eight Great Lakes States and three EPA
Regional Offices.  These managers selected this approach based on their many
years of regulating pollutants, including direct experience in the Great Lakes
basin.

      EPA agrees with comments that recommended considering the sediment route
of exposure in predicting BAFs, and partially agrees with comments that
recommended that only field-measured BAFs, not predicted BAFs, should be used
in determining BCCs.  As a result, EPA has added the field-measured biota-
sediment accumulation factor  (BSAF), which considers the sediment route of
exposure, to the hierarchy of methods for deriving BAFs; has selected a new
BAF model that better accounts for chemical uptake through sediments; and has
modified the Tier I minimum data requirements for human health and wildlife
criteria to specify minimum bioaccumulation data.  The definition of BCCs has
also been revised to be consistent with the above changes.  The final Guidance
now specifies that the minimum BAF information needed to define an organic

-------
66     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

chemical as a BCC is either a field-measured BAF or a BAF derived using the
BSAF methodology, and that the minimum BAF information needed to define an
inorganic chemical, including an organometal, as a BCC is either a field-
measured BAF or a laboratory-measured BCF.  The reasons for these changes as
well as other changes affecting the derivation of BAFs are discussed in
section IV.B.2 of this document.

      In response to a number of comments on the BAF cutoff level for defining
a BCC, EPA has reviewed all of the information and policy .considerations in
selecting the cutoff level.  As a result, EPA has made the risk management
decision to retain the proposed BAF cutoff level of 1000 for defining BCCs.

      EPA does not agree with the comments that the selection of a BAF  cutoff
level of 1000 for defining BCCs is arbitrary.  EPA believes that this comment
may have resulted from a confusion about the nature of risk management
decisions.  As EPA explained in the preamble to the proposal (58 FR 20844),
the selection of a BAF cutoff level is a risk management decision that
involves weighing information and policy considerations, rather than a  risk
assessment assumption that results solely from a scientific analysis.   It is
not possible, therefore, to specify a mathematical formula or systematic
algorithm employing environmental data to select a cutoff level.

      EPA weighed a wide range of information and policy considerations in
this decision.  These include the following considerations:

      --  The cutoff level for a BAF should include pollutants that are
"highly bioaccumulative" based on exercise of reasonable best professional
scientific judgment.  EPA believes that a BAF of over two or three orders of
magnitude  (100 or 1000) would meet this definition.

          The cutoff level should be high enough that it includes most
pollutants for which the fish consumption pathway is the most important route
of exposure for humans and wildlife.  To select a lower level would cause a
scientist using reasonable best professional judgment to question whether the
pollutant was "highly bioaccumulative."

          The cutoff level should be low enough to include those persistent,
bioaccumulative pollutants already found to be causing significant
contamination, including 2,3,7,8-TCDD (dioxin), mercury, and PCBs.
                   •
          The cutoff level should be low enough to provide adequate assurance
that other chemicals that could potentially contaminate the food web of the
Great Lakes ecosystem in the future are subject to the special provisions for
BCCs.

      --  The cutoff level should not be set so low that the regulatory and
administrative structure in place in the Great Lakes States and Tribes  for
control of discharges to waters would be overwhelmed.

          The cutoff level should be developed after consideration of the
scientific, policy,'administrative, and technical input from stakeholders in
the Great Lakes basin--State regulators, regulated community, and public
interest groups.

      EPA has determined that the cutoff level of 1000 initially selected by
the GLI Steering Committee meets all of the above considerations.  As
explained in the preamble to the proposal, a pollutant with a BAF greater than
1000 was believed by the Steering Committee to have a high potential to be
found in aquatic organisms of the Great Lakes System and therefore to have the
potential to cause a significant risk to the health of the aquatic life and
consumers of the aquatic life such as wildlife and humans inhabiting the Great
Lakes basin.  The Steering Committee made its recommendation on  the basis of
information available to them as managers of water quality programs.

-------
                           Section II: Regulatory Requirements                        67
      EPA recognizes that other numbers meeting the above considerations could
have been selected as a cutoff.  For example, several commenters suggested
lowering the cutoff to 250.  Others suggested a cutoff of 100, and others
suggested including all pollutants in Table 6 regardless of BAF.  None of
these commenters, however,' provided reasons other than increasing the number
of pollutants to be treated as BCCs as high as possible to avoid as much risk
as possible.  In other words, none of these commenters provided a rationale
that would prefer one "low" cutoff over any other.

      On the other hand, a few commenters suggested raising the cutoff to
10,000 or 100,000.  They appear to believe the cutoff should be as high as
possible without exceeding the BAF of the least bioaccumulative pollutant
currently known to cause problems in the basin.  EPA believes there are
currently known "problem" BCCs, such as lindane, with BAFs lower than 10,000.
EPA believes that it is reasonable and appropriate to retain the proposed
cutoff of 1000 not only to avoid excluding such pollutants, but also to
prevent adverse inputs from additional bioaccumulative chemicals in the
future.  Past discharges of highly bioaccumulative pollutants have resulted in
contamination of the Great Lakes System that is taking decades to subside.
EPA believes it is reasonable and appropriate to prevent this from happening
with other chemicals in the future.

      EPA does not agree with comments that EPA should solicit formal public
comment before States or Tribes treat any additional chemicals as BCCs in the
future.  EPA believes that the States and Tribes should have the ability to
designate additional chemicals for BCC controls based on information available
to them without waiting for EPA to act.  As discussed above, EPA will operate
the GLI Clearinghouse as a means to share pollutant information, including
BAFs, as quickly as possible.  If new information becomes available showing an
organic chemical to have a field-measured BAF of over 1000, for example, this
information would be reviewed by EPA and other Clearinghouse participants and
placed in the Clearinghouse, where States and Tribes would be alerted.  States
and Tribes would be able to apply the special BCC provisions to the pollutant
after following their applicable State or Tribal public review procedures for
revisions to water quality standards or for permit development.  For example,
the State or Tribe could include a description of the special BCC provisions
in the public notice for a NPDES permit.  EPA believes this would be a more
efficient approach than relying in all cases on EPA to sponsor a public review
and comment process, which has often taken several years for similar types of
actions.

      EPA has some concern that inconsistencies could arise among States and
Tribes concerning future identification of BCCs under the above approach.  EPA
believes operation of the Clearinghouse will minimize this possibility.
Nevertheless, if serious inconsistencies arise, EPA may from time to time
publish available BAF data for a pollutant and solicit public comments.  EPA
could then issue final technical assistance and recommendations concerning the
pollutant to assist State and Tribal revisions to water programs.

      c.    Final Guidance.  EPA has revised the definition of a BCC in
§ 132.2 of the final Guidance to exclude chemicals that do not have the
potential to cause adverse effects.  EPA has also modified the proposed
definition to exclude chemicals with half-lives of less than eight weeks in
the water column, sediment, and biota.  Finally, the definition specifies that
the human health BAF for non-metals must be derived from a field-measured BAF
or a field-measured BSAF.  EPA has retained all other features of the proposed
definition,  including the BAF cutoff level of 1000.  The revised definition
reads:

      BCC is any chemical that has the potential to cause adverse effects
      which, upon entering the surface waters, by itself or as its toxic
      transformation product, accumulates in aquatic organisms by a human
      health bioaccumulation factor greater than 1000, after considering
      metabolism and other physicochemical properties that might enhance or

-------
68     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      inhibit bioaccumulation, in accordance with the methodology in appendix
      B of this part.  Chemicals with half-lives of less than eight weeks in
      the water column, sediment, and biota are not BCCs.  The minimum BAF
      information needed to define an organic chemical as a BCC is either a
      field-measured BAF or a BAF derived using the BSAF methodology.  The
      minimum BAF information needed to define an inorganic chemical,
      including an -organometal,  as a BCC is either a field-measured BAF or a
      laboratory-measured BCF.  BCCs include, but are not limited to, the
      pollutants identified as BCCs in section A of Table ,6 of this part.

      EPA has applied the above changes in the definition to calculate BAFs
for individual pollutants listed in Table 6.  The results are included in the
Great Lakes Water Quality Initiative Bioaccumulation Factor Technical Support
Document, which is available in the docket for this rulemaking.  These
calculations have resulted in the deletion of six chemicals from the proposed
list of pollutants in Table 6A:  aldrin, 4-bromophenyl phenyl ether, endrin,
heptachlor, heptachlor epoxide,  and methoxychlor.  For these six chemicals
there were insufficient data to develop a field-measured BAF or a BAF
predicted from a field-measured BSAF.  BAFs based on laboratory-measured BCFs
are available for endrin, heptachlor, heptachlor epoxide, and methoxychlor;
BAFs based on a laboratory-predicted BCF are available for aldrin and 4-
bromophenyl phenyl ether.

      As a result, the final Guidance lists 22 BCCs in Table 6A to part 132.
They represent EPA's best scientific judgment at this time concerning which
pollutants on Table 6 meet the final definition of BCCs.  States and Tribes
may determine, however, that additional pollutants meet the definition or
should be subject to the special provisions for BCCs.  If so, they should take
appropriate regulatory action as discussed above to treat the chemicals as
BCCs in their water quality standards and NPDES permit programs, including the
special provisions for BCCs contained in the final Guidance.

      EPA has underway field testing programs to develop field-measured BAFs
and field-measured BSAFs for additional pollutants, including the six
chemicals deleted from the proposed list of BCCs.  It is possible that some of
this testing may result in identification of additional pollutants that meet
the definition of BCC.  When this testing is complete and BAFs have been
calculated, EPA will place the results in the GLI Clearinghouse where they
will be available for States to use in making their determinations.

9.    Potential Bioaccumulative Chemicals of Concern

      a.    Proposal.  The proposal listed  ten chemicals as potential BCCs in
section B of the proposed Table 6.  Although the ten chemicals had BAFs of
greater than 1000 based on predicted BCFs  (eight chemicals) and on laboratory-
measured BCFs  (two chemicals), other available information described in the
proposal to the preamble raised serious doubts that the actual BAFs for these
chemicals exceeded 1000.  For this reason,  EPA proposed that the special
provisions for BCCs would not apply to these ten pollutants.  EPA invited
comment on whether any or all of the potential BCCs should be listed as BCCs
and any additional data relevant to these determinations.

      b.    Comments.  Many  comments recommended that EPA delete the list  of
potential BCCs and not apply the special BCC provisions to  these chemicals.
These commenters asserted that the ten potential BCCs are readily metabolized
and do not bioaccumulate at  high rates.

      Some comments recommended  that several polynuclear aromatic hydrocarbons
 (PAHs) on the  list of potential BCCs should be treated  as BCCs, asserting  that
these chemicals  have relatively high BAFs when measured in  lower trophic level
species, and therefore pose  a risk to  wildlife or people eating those  species,
even though they are metabolized by higher  trophic  level species and therefore
have relatively  low' human health BAFs.

-------
                           Section II: Regulatory Requirements                        69
      EPA agrees that the potential BCCs should not be treated as BCCs, and
should be deleted from the final Guidance.  For the reasons discussed above  in
section II.C.8 of this document, EPA has modified the definition of BCC in the
final Guidance to use only the two methods of developing BAFs for non-metals
that take into account metabolism.  None of the ten potential BCCs have BAFs
available using these two methods, and therefore they are not listed as BCCs
in Table 6.  Furthermore, as discussed in the preamble to the proposed
Guidance, there are indications that the ten chemicals would likely not have
BAFs exceeding 1000 if metabolism were taken into account,.  If reliable data
were to become available in the future showing a BAF of over 1000 for any of
these ten chemicals or other chemicals based on a field-measured BAF or BSAF,
then States and Tribes would need to apply the special BCC provisions to these
chemicals.

      EPA does not agree that the PAHs with high BAFs at low trophic levels
should be treated as BCCs at this time.  The special provisions for BCCs are
designed to address pollutants that accumulate in the food web of the Great
Lakes ecosystem.  Pollutants that accumulate in lower trophic levels but which
are metabolized at higher levels are not as likely to affect the food web as a
whole as those that continue to accumulate at higher trophic levels.  The
definition of a BCC in the final Guidance therefore does not include any
special procedures that would classify such a pollutant as a BCC.  Such a
pollutant would, however, become a BCC if data were to become available in the
future resulting in a human health BAF of over 1000.

      EPA established the category of "potential BCC" in the proposed Guidance
primarily to obtain comment on whether such pollutants met the definition of
BCCs and should, accordingly, be subject to the special provisions for that
class of pollutants.  Since EPA has now completed its analysis of comments and
determined that potential BCCs do not appear to satisfy the final definition
of BCC, and since no special provisions in the Guidance apply to potential
BCCs that do not apply to other pollutants, there is no further purpose for
retaining the list of potential BCCs in the final Guidance.  Therefore, the
proposed list of potential BCCs has been deleted.

      c.    Final Guidance.  For the reasons discussed above, the proposed
list of potential BCCs in section B of Table 6 of part 132 has been deleted.
Section C of Table 6 has been redesignated as section B, Pollutants That Are
Not BCCs.

10.   Pollutants of Initial Focus

      a.    Proposal.  As described in the preamble to the proposed Guidance
(58 FR 20843-44), 138 pollutants were identified by the Great Lakes Initiative
Steering Committee and listed in Table 6 of the proposed Guidance as the
pollutants of initial focus in the Great Lakes Water Quality Initiative.  The
138 pollutants in proposed Table 6 included: (a) the 126 pollutants identified
by EPA as priority toxic pollutants in appendix A of 40 CFR 423; (b) selected
pollutants listed in the Great Lakes Water Quality Agreement of 1978 (as
amended by the Protocol signed November 18, 1987) ; (c) certain pollutants
categorized under the Lake Ontario Toxics Management Plan and the Niagara
River Toxics Management Plan; and (d) three pollutants included on a case-by-
case basis.

      The primary purpose of the Initiative Committees in specifying
pollutants in Table 6 was to provide an initial focus for criteria development
and the calculation'of bioaccumulation factors in the Great Lakes System.  The
pollutants included in Table 6 were not intended to be a comprehensive
inventory of all pollutants present, used,  manufactured, or stored in the
Great Lakes System, but were thought to represent pollutants which may be of
concern in the Great Lakes, and for which adequate effluent, ambient, and
toxicity data would be available to assess the impact of the various options
considered in developing the Guidance.

-------
70     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      The proposed Guidance provided an initial focus on the Table 6
pollutants in the following three ways:  First, the pollutants for which EPA
and the States applied the proposed criteria methodologies to derive numeric
water quality criteria--that is, the pollutants in Tables 1 through 4--were
selected from the list of pollutants in Table 6.  Second, EPA and the Great
Lakes States calculated bioaccumulation factors (BAFs) for the Table 6
pollutants to.assist EPA and the States in developing criteria and values to
protect human health and wildlife.  Third, the Table 6 list of pollutants is
one factor used in determining when States, Tribes, and/or permittees need to
generate data necessary to calculate Tier II values used in developing water
quality-based effluent limits.  Comments on the use of Table & in the data
generation provisions of procedure 5 of appendix F are discussed in section
VIII.E of this document.

      b.    Comments.  Some comments suggested additional pollutants to be
added to Table 6, including: pollutants listed in the Great Lakes Water
Quality Agreement, Annex 10, Appendix l (Hazardous Polluting Substances) or
Appendix 2  (Potentially Hazardous Polluting Substances) ,- high-use pesticides;
persistent pollutants that do not bioaccumulate; all drinking water
contaminants as listed in the 1986 amendments to the Safe Drinking Water Act;
pollutants categorized as ID or IE in the Categorization of Toxics in Lake
Ontario or in the Categorization of Toxic Substances in the Niagara River; all
the dioxins, furans and PCBs that are known to operate via the Ah receptor in
any animal species,- as well as a number of specific pollutants listed by
commenters.

      Some comments suggested that Table 6 should be limited to chemicals
which are causing demonstrated impacts to water quality in the Great Lakes
System.  Other comments suggest limiting Table 6 to pollutants which pose
water quality concerns unique to the Great Lake system, such as the list of
critical pollutants being developed by Lakewide Management Plans.  Other
comments recommended that in the interest of establishing reasonable
priorities for the Great Lakes States, EPA should eliminate non-BCCs from
Table 6.  Other comments recommended removing specific pollutants from Table
6, such as asbestos, fluoride, methylene chloride, phenol, phthalate esters,
and silver.

      Some comments expressed concern that the proposed Guidance did not
include clear procedures on how additional toxic pollutants that are
introduced or discovered in the Great Lakes Ecosystem will be added to Table
6, or be regulated prior to formal revision of the Guidance.

      EPA does not agree that additional pollutants should be added to Table
6.  EPA believes it would be counterproductive to expand the pollutants of
initial focus to a significantly larger set of pollutants.  Table 6 has
already been used successfully during the development of the proposed and
final Guidance to focus development of data needed to implement the final
Guidance.  As a result of EPA and State efforts, many of the data gaps that
existed for these pollutants at the start of the Great Lakes Water Quality
Initiative have been filled.  Applying limited EPA, State, Tribal, and
discharger resources at this time to address a broader list of pollutants in
the ways described under the Proposal section above would divert resources
away from other important actions, such as developing a richer data base
concerning effects of pollutants  in the proposed Table 6 on endangered or
threatened species.

      EPA also does*not agree with comments  that the list of 138 should be
reduced.  All of the 138 pollutants have been identified as either priority
pollutants under the CWA or as pollutants of specific concern in the Great
Lakes basin.  EPA also does not agree that Table 6 should be limited to BCCs.
BCCs are not the only types of pollutants currently or potentially affecting
the Great Lakes ecosystem.  For the same reasons, EPA does not agree with
comments that specific pollutants be removed from the  list.

-------
                           Section II: Regulatory Requirements                        71
      EPA recognizes that the final Guidance does not include procedures for
adding additional pollutants to Table 6.  EPA believes that the GLI
Clearinghouse can be used as a forum for determining additional needs for BAF
calculations, and for sharing of BAF results.  If it should become apparent in
the future that adding pollutants to Table 6 would assist in reducing
disparities between' data generation approaches of the States and Tribes under
procedure 5.C of appendix F, EPA would consider recommending that States and
Tribes expand the Table 6 lists that they have adopted into their programs to
be consistent with the final Guidance.  This could be done during the
triennial review of water quality standards programs under section 303 of the
CWA.

      c.    Final Guidance.  The final Guidance retains Table 6 as proposed.

      EPA would like to clarify that the methodologies and procedures in the
final Guidance generally apply to all pollutants, except for the pollutants in
Table 5.  The Table 6 list of pollutants is one factor used in determining
when States, Tribes, and/or permittees need to generate data necessary to
calculate Tier II values used in developing water quality-based effluent
limits.

D.    Procedures for Adoption and EPA Review

1.    Adoption Procedures

      a.    Proposal.  Proposed § 132.5(a) required the Great Lakes States and
Tribes to adopt and-submit for EPA review and approval the criteria,
methodologies, policies and procedures developed pursuant to part 132 by a
date no later than 18 months from the date of final publication of the part
132 requirements.  EPA proposed the 18-month deadline in order to allow the
full time available under the statute for EPA review and approval of
submissions and for States and Tribes to correct any identified deficiencies,
and still allow EPA to meet the section 118(c)(2)(C) requirement for review,
approval or disapproval and promulgation by EPA, if necessary, within two
years after the final publication of the Guidance.  Proposed § 132.5(d) also
provided a 30-day public comment period on the submissions.

      If a Great Lakes State or Tribe failed to submit criteria,
methodologies, policies, and procedures to EPA for review, or if EPA
disapproved portions of all of a State or Tribal submission because it was
inconsistent with part 132, proposed § 132.5 provided that the requirements of
part 132 would apply to discharges within the State or Federal Indian
Reservation upon EPA's publication of a final rule in the Federal Register
indicating the effective date of the part 132 requirements in the identified
jurisdictions.

      b.    Comments.  A few commenters recommended that the Guidance be
adopted as soon as possible because they believe it is a major step forward in
efforts to protect the Great Lakes and fulfill promises of the U.S.-Canada
Great Lakes Water Quality Agreement and the Great Lakes Governors' Toxic
Substances Control Agreement.

      Some commenters indicated that the two-year adoption period was too
short.  Other commenters recommended that EPA eliminate the 18-month deadline
for State submittals, or allow extensions to the deadline on a case-by-case
basis as long as the State and EPA agree on a schedule that will meet the
statutory deadline.  Several commenters stated that the public comment
requirements during EPA's review of State submittals may not be necessary, and
are inconsistent with the current national program.

      EPA agrees that States and Tribes should adopt provisions consistent
with the Guidance as soon as possible.  EPA has reviewed the steps States,
Tribes,  and EPA must take to develop, adopt,  approve, and if necessary
disapprove and promulgate the Guidance, however, and has concluded that it

-------
72     Water Quality Guidance for the Great Lakes System - Supplementary Information Document

would not be possible nor practical to complete these steps in less time than
the two years specified in section 118(c) (2) (C) of the CWA.  EPA has therefore
not shortened the deadline in the final § 132.5(a).  EPA encourages States and
Tribes to accelerate their adoption and submission processes as much as
possible, and as discussed further below, EPA Regional Offices will work with
States and Tribes to facilitate early adoption and approval where possible.

      EPA recognizes that the two-year deadline specified by the Congress may
be extremely ambitious.  EPA's experience in reviewing State water quality
standards adopted under section 303(c)  indicates that many of the Great Lakes
States have had difficulties adopting even routine water quality standards
revisions within three years as specified for triennial reviews in the CWA.
Nevertheless, EPA believes the States,  Tribes, and EPA Regional Offices can
work together to meet the deadline, especially since the majority of
provisions of the final Guidance were developed by the States themselves as
part of the Great Lakes Water Quality Initiative and many are currently in
effect in one or more of the States.  Furthermore, the two-year deadline is
established explicitly by section 118(c)(2)(C) of the CWA.  Therefore, the
final Guidance retains the two-year deadline for final approval or
promulgation.

      EPA agrees with comments, however, that flexibility may be appropriate
in some cases concerning the eighteen-month deadline for State and Tribal
submissions to EPA under § 132.5.  There may be situations where the EPA
Regional Office has worked closely with the State or Tribe during their
development and adoption of provisions consistent with the Guidance and
believes that the full six month period for EPA review of the submission will
be unnecessary.  For example, a State may decide to adopt the provisions of
the Guidance with only minor modifications, and may have early indications
that the public supports this approach.  In this situation, EPA believes that
an extension to the eighteen-month deadline would be reasonable because EPA
would have a high degree of assurance that it would still be able to review
and approve the submission within the two-year statutory deadline.  Therefore,
EPA has added § 132.5(c) that allows the EPA Regional Administrator to extend
the deadline for a State or Tribal submission beyond 18 months if the Regional
Administrator believes that the submission will be consistent with the
requirements of the.final Guidance and can be reviewed and approved within the
two-year deadline.  In these cases, EPA expects that the Regional
Administrator will need to reach early agreement with the State or Tribal
director on a joint schedule for specific steps that the State or Tribe and
EPA will take over the two-year period to meet the requirements of § 132.5.
To ensure success, the EPA Regional Offices will likely need to provide
technical guidance and assistance to the States or Tribes early in the
adoption process, and will need to begin analyzing drafts of the State or
Tribal submission even before it is formally  submitted to determine whether
provisions are consistent with the final Guidance.

      EPA also agrees in part with comments  that the proposed requirement  for
a 30-day comment period was unnecessary and  should be revised.  The proposed
provision for public comment was patterned  in general on existing minimum
public comment requirements for proposed modifications to State NPDES programs
under 40 CFR 123.62(b) and State water quality standards programs under part
131.  Section 123.62(b) requires a public comment period on State NPDES
submissions only if the proposed program modification is substantial.
Furthermore, section 303 of the CWA and part  131 do not require EPA to provide
public review and comment on EPA  approvals  of State water quality standards
submissions.  Citv of Albuquerque v. Browner.  38 ERG 2062  (DNM  1993) .  EPA
believes that it is.not necessary or appropriate to impose more extensive
public notice-and-comment requirements  for  submissions under part 132 than for
the submissions  in the underlying NPDES  and water  quality  standards programs
affected by part 132.  EPA also believes  that EPA  Regional Offices will be
able to determine whether a  specific provision in  the submission constitutes a
substantial modification to  the underlying  State program without great
difficulty  because of their  familiarity with the existing  State programs.

-------
                           Section II: Regulatory Requirements                        73
Therefore, EPA has modified proposed § 132.5(d), now redesignated  §  132.5(e),
to conform with the minimum public notice and comment requirements of parts
123 and 131.

      One commenter urged that phase-in periods for implementing certain
provisions  (e.g., the ban on mixing zones for BCCs) be dated from  the time of
final federal publication of the Guidance, not the date of State and Tribal
final adoptions.  EPA does not agree with this comment.  Section 118(c) (2) (C)
requires States to adopt provisions consistent with the final Guidance by no
later than two years from date of publication, or be subject to EPA
promulgation within that two-year period.  Because the provisions  will not be
effective until they are promulgated by a State, Tribe, or Federal agency, EPA
believes that it is reasonable to delay the computation of any phase-in period
until the statutory deadline for such promulgation.  Accordingly,  all phase-in
periods and similar provisions that distinguish between activities occurring
before or after a specified date  (e.g., the definition of "new Great Lakes
discharger") are calculated from the date two years after publication of this
final Guidance.

      c.    Final Guidance.  Section 132.5(a) of the final Guidance  retains
the general deadline of eighteen months after publication of the final
Guidance, or September 23, 1996, for States and Tribes to submit criteria,
methodologies, policies, and procedures developed pursuant to the  Guidance.

      The final Guidance has been modified to add new § 132.5(c) which allows
the Regional Administrator to extend the deadline for the submission required
in § 132.5(a) if the Regional Administrator believes that the submission will
be consistent with the requirements of the final Guidance and can  be reviewed
and approved no later than March 23, 1997.  As discussed under the Comments
section above, EPA expects that before granting an extension the Regional
Administrator will reach agreement with the State or Tribal director on a
joint schedule for specific steps to meet the requirements of § 132.5.  The
joint schedule should include opportunities for EPA to review drafts of State
or Tribal provisions in order to facilitate EPA review and approval.  In these
cases, EPA intends to provide technical guidance and assistance to the State
or Tribe throughout the adoption process, beginning as soon as possible.
Conforming provisions have been added to § 132.5(a), and proposed  §  132.5(c)
has been redesignated as § 132.5(d).

      The provisions for public notice and comment in proposed § 132.5(d) have
been redesignated as § 132.5 (e) and modified to provide for public notice and
at least 30 days' public comment only in situations where the State  submits
substantial modifications to its NPDES program.

      The provisions for EPA approval and disapproval in proposed  §  132.5(d)
have been moved to new § 132.5 (f) .

      In reviewing proposed § 132.5, EPA noted that it would be
administratively impossible for EPA to provide the 30-day comment  period on
State submissions and still issue a notice of approval of a State  or Tribe's
submission within 60 days of receipt of the submission as proposed in
§ 132.5(d)(1).  The proposed 60-day deadline was based on the similar deadline
in section 303(c) of the CWA for approving State water quality standards.  The
section 303(c) approach does not include a requirement for public  notice and
comment,  however.  Since the final Guidance includes a minimum of  30 days for
such a requirement in cases where a State submits a substantial modification
to its NPDES program, EPA has modified the final Guidance to add 30  days to
the deadline for EPA approval.  Final § 132.5(f)(1) now requires EPA to issue
a notice of approval of a State or Tribe's submission within 90 days of
receipt of the submission.

      Proposed § 132.5(e) has been redesignated as § 132.5(g).

-------
74     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      A new § 132.5(h) has been added concerning results of consultation under
the Endangered Species Act.  It is discussed further in section II.G of this
document.

      Proposed § 132.5 (f) has been redesignated as § 132.5 (i).

2.     Interpretation of "Consistent With"

      a.    Proposal.  Section 118(c)(2)(C) of the CWA requires Great Lakes
States to adopt water quality standards, antidegradation policies, and
implementation procedures consistent with the final Guidance, or EPA is
required to promulgate for the States.   Section 132.5(e) of the proposed
Guidance specified the conditions under a State or Tribe submission is
consistent with the requirements of part 132.  Generally, the proposed
Guidance provided that submitted criteria, methodologies, policies and
procedures would be consistent with part 132 if they are "equal to or more
restrictive than" the provisions in the final Guidance.  Proposed
§ 132.5(e)(3) clarified EPA's intention to evaluate the State and Tribal
submissions on a provision-by-provision basis by providing that if States or
Tribes adopt provisions more restrictive than the final Guidance, the more
restrictive provision may not be offset by relaxation of other specific
elements of the final Guidance.

      b.    Comments.  Many comments urged EPA to strictly interpret the
"consistent with" requirements, and to require States and Tribes to justify
deviations from the Guidance, in order to achieve the desired standardization
of programs basin-wide.  Several comments urged EPA to require explicitly that
all State and Tribal procedures and criteria be consistent with and no less
stringent than Guidance procedures and criteria.  A few comments said States
should be directed to retain existing numeric water quality criteria and
procedures where they are more stringent than the final Guidance, in order to
maintain continual progress toward zero discharge as outlined in the Great
Lakes Water Quality Agreement and the CWA.

      Some commenters interpreted the CPA as directing EPA to publish guidance
to the states, rather than a fixed rule, and therefore urge EPA to give the
States a measure of  latitude in implementing the Guidance.  Some commenters
argued that EPA should not reduce State flexibility currently available in
States' implementation of section 303(c) of  the CWA.

      Some commenters believed that State flexibility is needed to enable
cost-effective implementation of the Guidance, to maintain State primacy as
provided in the CWA,  to minimize the administrative burden on the States, to
reflect the innovative nature of the Guidance, and because States are more
familiar with the environmental conditions  in local areas  and are better able
to work effectively  with dischargers to address problems.  Some commenters
urged additional state flexibility to ensure that future improvements in
science are readily  incorporated into the Guidance.  Some  comments stated that
more flexibility is  needed to reflect ecological differences between areas in
the Great Lakes basin which  require  different regulatory responses.  Some
commenters recommended that  since many  real-world permitting situations
probably exist which preclude a straightforward application of the proposed
implementation procedures, EPA should to  the maximum extent possible structure
the final Guidance to function truly as  "guidance" for  the States.

      Some comments  recommended that where  the overall  outcome is  the same,
EPA should allow a state to  be less  stringent in one area  of the methodology
if it is more stringent  in another.  Some  commenters suggested that  EPA should
interpret  "consistent with"  as meaning  "substantially  equivalent  to"  so that
States  and Tribes may adopt  criteria and procedures tailored to  their own
circumstances so long as the overall impact of the program is to treat
substantially similar- sources  in a substantially equivalent  manner.

-------
                           Section II: Regulatory Requirements                        75
      EPA does not agree that the amendments to section 118(c)(2) of the CWA
directed EPA simply to publish non-binding guidance.  This section not only
directs EPA to provide guidance to the Great Lakes States on minimum water
quality standards, antidegradation policies, and implementation procedures for
the Great Lakes System, but also requires the States to adopt water quality
standards, antidegradation policies, and implementation procedures for waters
within.the Great Lakes System which are consistent with such guidance or'be
subject to EPA promulgation  (section 118(c)(2)(C)) .  EPA believes that whether
States and Tribes adopt minimum standards, policies, and procedures consistent
with the final Guidance, or whether EPA promulgates them, the Congress
intended that the final Guidance would establish minimum, and ultimately,
enforceable requirements for the Great Lakes System.

      This interpretation of section 118(c)is supported by the primary authors
of the Critical Programs Act.  In a June 9, 1994, letter to Governor Cuomo
discussing this issue, Senators Levin, Glenn, and Kohl emphasized the need to
provide enforceable requirements through a federal regulation in order to
improve consistency in State water quality programs in the Great Lakes System.

      ...While flexibility is built into the law, the Great Lakes Guidance is
      nevertheless intended to be an enforceable federal
      regulation....Guidelines without enforceability cannot achieve the
      overarching goal of the Great Lakes Initiative to ensure consistent
      Great Lakes water quality protections throughout the basin.

            In light of the differences among our state water quality
      programs, the Great Lakes Guidance will no doubt require adjustments in
      each of our states.  But our states knew that when they launched the
      Initiative and committed themselves to developing a single set of rules
      for everyone to live by in the Great Lakes Basin.  In fact, reducing
      state disparities was a driving force behind the Initiative....

      This interpretation of section 118(c) is also supported by the extremely
short time frame specified in the statute for EPA promulgation of the Guidance
in the absence of State adoption.  As discussed above, provisions consistent
with the final Guidance must ultimately be adopted by Great Lakes States and
Tribes or be promulgated by EPA within two years.  This statutory deadline for
State adoption and EPA approval or promulgation of the Guidance does not allow
sufficient time for EPA to take additional steps, such as publishing the
proposed and final Guidance first, followed by proposing and finalizing a
separate rule governing EPA approval and promulgation.  Rather, the rule that
was proposed and is' now being finalized contains both the substance of the
Guidance,  including the criteria in Tables 1 through 4 and the methodologies,
policies,  and procedures in appendixes A through F, as well as the approval
and promulgation procedures that EPA will use for review of State and Tribal
submissions.

      EPA also does not agree that the CPA intended to afford States the same
flexibility currently available under section 303(c) of the CWA.  Section
303(c)  requires that States adopt and implement water quality standards to
help restore and maintain the chemical, physical, and biological integrity of
the Nation's waters.   EPA has periodically provided guidance to States in
implementing these requirements, including water quality criteria published
pursuant to section 304(a) of the CWA.  States have the flexibility to use
section 304(a)  guidance or other scientifically defensible approaches in
adopting and implementing water quality standards.  As discussed in the
preamble to the proposed Guidance (58 FR 20835), however, the Initiative
Committees believed this level of flexibility over the years had resulted in
significant differences in State adopted water quality standards,
antidegradation policies, and implementation procedures, as well as
inconsistencies in regulatory approaches and individual permit decisions in
the Great Lakes basin.  The Congress was aware of these inconsistencies and of
the efforts of the Initiative Committees to develop the Guidance when it
enacted the CPA.  The requirements in section 118(c)(2) described above

-------
76     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

represent the direction of the Congress to adopt or promulgate minimum  water
quality standards, antidegradation policies, and implementation procedures for
the entire" Great Lakes System.  In order to have any meaning, implementation
of these specific requirements for the Great Lakes System will necessarily
result in some reduction in the flexibility of the Great Lakes States.

      EPA agrees with comments that the final Guidance should require State
and Tribal submissions to be at least as protective as the final Guidance in
order to achieve the increased consistency and minimum threshold levels of
control intended by the Congress.  For this reason, EPA has retained the
proposed § 132.5 (e) with clarifying amendments, now redesignated as §
132.5(g), under which submitted criteria, methodologies, policies and
procedures will be considered consistent with part 132 if they are "as
protective as" the provisions in the final Guidance.  EPA believes that
specifying "as protective as" will ensure that provisions adopted by the
States and Tribes will be equivalent to or more protective than the final
Guidance.  Further, EPA believes that this is a reasonable and'appropriate
mechanism for implementing EPA's duty to define the "minimum" requirements for
water programs in the basin.

      EPA also agrees with comments, however, that it is appropriate to
provide reasonable flexibility to States and Tribes, to the extent that this
can be done and still meet the requirements and purpose of the CWA.  In
overseeing States' implementation of the CWA, EPA has found that reasonable
flexibility is not only necessary to accommodate unforeseen circumstances, but
is also appropriate to enable innovation and progress as new approaches and
information become available.  To address the need for flexibility, EPA made
several changes to the proposed Guidance.

      First, EPA reviewed all sections of the proposed Guidance and all
comments to determine the appropriate level of flexibility.  Based on this
review, the final Guidance provides increased flexibility for State and Tribal
adoption and implementation of these provisions in many areas, including
antidegradation, TMDLs, intake credits, site-specific modifications,
variances, compliance schedules, elimination of mixing zones for BCCs,  and the
scientific defensibility exclusion.  The final Guidance also provides reduced
detail of provisions in many areas, and provisions for the exercise of  best
professional judgment by the Great Lakes States and Tribes when implementing
many individual provisions.  This increased flexibility is discussed further
in section I of this document.

      Second, EPA clarified how  "offsets" will be considered between elements
of State and Tribal programs.  See discussion below.

      Third, EPA added  §132.1(b)  in the final Guidance to clarify that
verbatim adoption of the Guidance is not required.  EPA believes that State
and Tribal programs do not need to be identical to the Guidance  to result in
equally protective criteria, methodologies, policies, and procedures.   States
and Tribes instead have flexibility to adopt alternative provisions as  long as
they are as protective  of human health, wildlife, and aquatic life in the
Great Lakes System as the Guidance.

      EPA believes that the flexibility available in the final Guidance will
satisfy many of the specific concerns of commenters.  For example, based  on
their familiarity with  environmental  conditions in local areas,  States  and
Tribes may use the site-specific  modification procedures to adjust criteria
and values in order-to  reflect ecological differences from site  to site.
States and Tribes also  have flexibility  to  adjust criteria, methodologies, or
procedures to reflect future improvements in scientific understanding.  For
example, States and Tribes may use the scientific defensibility  exclusion in
§132.4(h) to apply a new, alternative methodology or implementation procedure
 (see section II.C.6 of  this document), or they may adjust elements within a
Guidance provision as long as the provision as a whole  is consistent with the
final Guidance  (see discussion of "offsets" below in this section).

-------
                           Section II: Regulatory Requirements                        77
Additionally, as discussed in section II.C.I of this document, there are
several steps within the criteria methodologies where some flexibility  is
available to reflect new findings and data.

      EPA understands that the science of risk assessment utilized  in
environmental protection, including criteria methodologies, is rapidly
evolving and improving.  Therefore, to ensure that the scientific basis for
the methodologies in appendices A through D is always current and peer
reviewed, EPA will review the methodologies and revise them as appropriate
every three years.

      The final Guidance is not, by itself, enforceable.  Provisions
consistent with the Guidance will become enforceable only when adopted  by a
State or Tribe as part of its NPDES or water quality standards programs,
promulgated by EPA in the absence of State or Tribal action, or when included
in a NPDES permit.

      EPA does not agree with comments that States should be directed to
retain existing numeric water quality criteria that are more stringent  than
the final Guidance.  Although under § 132.4(i) States may choose to retain
more stringent criteria, the final Guidance does not require them to do so.
The CPA requires the Guidance criteria to protect human health, aquatic life,
and wildlife.  EPA and the Initiative Committees have designed the  criteria
methodologies to meet this requirement.  Therefore, while the development of
more stringent provisions may be necessary because of site-specific conditions
within their jurisdiction and is also available as an option for States and
Tribes under any circumstances, automatic retention of existing more stringent
criteria is not required by the CPA.  For this reason, the final Guidance
generally requires only that the States and Tribes adopt and use criteria and
values that are as protective as those produced by the Guidance methodologies.

      EPA also does not agree with recommendations that the decision whether
State or Tribal submissions are "consistent with" the final Guidance should be
based only on the overall outcome of applying all of the Guidance provisions.
Such an approach, which would allow "offsets" between separate provisions,
would be technically and administratively unworkable for the following
reasons.  Because of the differing nature of each methodology and procedure
and differences in the site-specific characteristics of each discharge  and
discharge location, it would be administratively difficult, if not  impossible,
to quantify and compare numerically the effects of allowing more stringent
changes in one methodology or procedure, for example, to "offset" less
stringent changes in different methodologies or procedures.  The only
potential way to make a valid comparison would be to apply the full set of
Guidance methodologies and procedures to specific cases, such as developing
water quality-based effluent limits for a range of NPDES discharges to a range
of receiving waters, and comparing those limits with limits derived using
alternative State or Tribal procedures for the same dischargers.  Based on its
experience, EPA does not believe such analyses can produce useful results
within a reasonable time or with available resources.  Additionally, EPA does
not believe that it could undertake this lengthy and complex analysis within
the short time period specified by Congress in section 118(c)(2)(C) for review
and approval or disapproval and promulgation of the Guidance  provisions in
the eight Great Lakes States.  Finally, allowing offsets between provisions
(e.g.,  between the human health criteria methodology and the aquatic life
criteria methodology, or between the TMDL procedure and the mixing  zones
provisions),  would not ensure appropriate levels of protection for  human
health, wildlife, and aquatic life in the Great Lakes System.  For  these
reasons, EPA has retained the prohibition against offsetting changes between
different provisions, but has made editorial changes to clarify that
variations or offsets of elements within a particular provision are acceptable
as long as the submitted provision is consistent with the Guidance.

      c.    Final Guidance.  The final Guidance generally retains the proposed
conditions in § 132.5 (e), redes'ignated as § 132.5 (g), under which EPA will

-------
78     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

determine that the criteria, methodologies, policies, and procedures  in a
State or Tribal submission are consistent with the requirements  of  the  final
Guidance.

      EPA corrected.the header for §  132.5(g)(2) to refer to  "pollutants other
than those listed in Tables 1, 2, 3,  4, and 5."  In the proposal, reference to
Table 5 was omitted by error.

      EPA modified § 132.5(g)(2)(ii)(A) to refer to "numeric  criteria adopted
by the State into State water quality standards and approved  by  EPA prior  to
March 23, 1997."  EPA added the reference to EPA approval, and changed  the
date to reflect the two years allowed for State adoption of the  final
Guidance.  EPA found that these technical changes were needed to clarify the
intent of this section.

      For the reasons discussed in the Comments section above, EPA  made
editorial changes to clarify the general prohibition in § 132.5(g)(3) against
offsets between provisions.  The final Guidance states that "adoption of a.
more protective element in one provision is not justification for adoption of
a less protective element in another  provision of this part."  States and
Tribes are not precluded, however, from offsetting elements within  a
particular provision as long as the adopted provision is consistent with the
final Guidance.  For example, a State could not use a more protective element
in the human health cancer methodology to offset a less protective  element in
either the human health non-cancer methodology or the aquatic life
methodology.  In contrast, a State could use a more protective element  within
the human health cancer methodology   (e.g., cancer risk level assumption)  to
offset a less stringent element within the same methodology (e.g.,  fish
consumption value) as long as the resulting methodology and criteria  were  as
protective as the final Guidance.

      When determining whether criteria adopted by the State  or  Tribe comply
with § 132.5(g)(1) , EPA will consider not only whether the State or Tribal
criteria are numerically as protective as the criteria in Tables 1  through 4,
but also whether all other provisions of the Guidance pertaining to the
criteria have been fully incorporated.  For example, EPA will consider  under
procedure 1 of appendix F whether more stringent criteria are needed  to
protect a threatened or endangered species.  Additionally, if a  State or Tribe
submits site-specific criteria less stringent than the criteria  in  Tables  1
through 4, then the State or Tribe should submit documentation as required by
§ 132.5(b)(4) showing how procedure 1 has been used to develop the  less
stringent site-specific modifications.  If a State or Tribe uses the
scientific defensibility exclusion in § 132.4(h) to develop less stringent
criteria, then the State or Tribe should likewise submit full documentation of
its demonstration that the corresponding Guidance criteria methodology  is
scientifically indefensible for the particular situation.  It is expected  that
§ 132.4(h) will only be invoked for data which become available  after this
publication of the final Guidance.  That is, § 132.4(h) will  generally  not
serve as the basis to reconsider data already available and considered  by  EPA
prior to this publication.

      Similarly, in reviewing criteria and values for pollutants other  than
those in Tables 1 through 5, EPA will consider whether all relevant
requirements of the Guidance have been fully incorporated into  State  or Tribal
programs.  For example, States and Tribes  should provide documentation  on
their use of § 132.5(g)(2)(i) or  § 132.5(g)(2)(ii) to develop numeric water
quality criteria; all data supporting any  site-specific modifications;  and
whether provisions regarding endangered or threatened species have  been
fulfilled.   Similarly,  if the State or Tribe has adopted § 132.5(g)(2)(ii),
they should  include documentation supporting adoption and implementation of
the required procedure  for developing water quality-based effluent  limits  and
total maximum daily loads.

-------
                           Section II: Regulatory Requirements                        79
      Finally, supporting documentation will also be necessary to assist EPA
in determining whether adopted methodologies, policies, and procedures comply
with § 132.5(g)(3).  That section describes how EPA will evaluate whether
State or Tribal methodologies, policies, and procedures are consistent with
the Guidance.  For example, the documentation might simply state that the
State or Tribe has adopted provisions verbatim from the Guidance, or with only
conforming changes.  If provisions were adopted with modifications, the
supporting documentation should identify the modifications, and should
demonstrate that the resulting methodology, policy, or procedure is expected
to produce results that are as protective as the corresponding Guidance
methodology, policy, or procedure.

      EPA would like to clarify that EPA intends to apply the general factors
in § 132.5(g) when approving or disapproving State or Tribal water quality
standards under § 131.21(b) and modifications to NPDES programs under part
123.  That is, future submissions or revisions to water quality standards and
NPDES programs for the Great Lakes System must continue to be consistent with
the Guidance even after EPA has approved or promulgated standards under
§ 132.5.

3.    Indian Tribes

      a.    Proposal.  The CPA requires Great Lakes States to adopt, or EPA to
promulgate, water quality standards, antidegradation policies, and
implementation procedures for waters within the Great Lakes System which are
consistent with the final Guidance.  In the proposal, EPA expanded the
requirement to "Great Lakes States and Great Lakes Tribes."  "Great Lakes
Tribes" are defined as any Indian Tribe whose reservation lies in whole or in
part within the drainage basin of the Great Lakes, and for which EPA has
approved water quality standards under section 303 of the CWA or which EPA has
authorized to administer an NPDES program under section 402 of the CWA.
"Indian Tribe" is further defined as any Indian Tribe, band, group, or
community recognized by the Secretary of the Interior and exercising
governmental authority over a Federal Indian reservation.  EPA believes that
inclusion of Great Lakes Tribes in this way is necessary and appropriate to be
consistent with section 518 of the CWA.  The reasons for EPA's proposal are
discussed further in the preamble to the proposed Guidance  (58 FR 20834) .

      b.    Comments.  Several comments supported the inclusion of Great Lakes
Tribes in the proposal, asserting that all land that is part of the Great
Lakes ecosystem should be subject to the Guidance.

      No comments were received rejecting the proposed approach.  Some
comments urged EPA to clarify that time frames for Tribes to adopt the final
Guidance should not start until the date a Tribe is authorized to administer
the program.  Another urged EPA to clarify that the proposed 18-month adoption
deadline "in no way limits any Tribe's rights or time limits to seek
qualification" for treatment as a State.

      EPA agrees that Great Lakes Tribes should be subject to the final
Guidance for reasons described in the preamble to the proposed Guidance (58 FR
20834).  EPA has therefore retained the proposed approach that generally
includes Great Lakes Tribes in the provisions of part 132 that otherwise apply
to Great Lakes States.

      EPA also agrees that any time deadlines for Tribes to adopt the final
Guidance should not start until the date a Tribe has a water quality standards
or NPDES program in place.  For water quality standards, this means the date
on which EPA has approved water quality standards for the Tribe, including
such basic elements as designated uses, narrative criteria, numeric criteria
for at least some pollutants necessary to protect designated uses, and legal
authorities.  Without-such basic elements, the water quality standards
provisions in the final Guidance could not be implemented.  For NPDES

-------
80     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

programs, this means the date on which EPA has authorized the Tribe to
administer an NPDES program under section 402 of the CWA.

      If any Tribe had received either water quality standards approval or
NPDES program authorization prior to this publication of the final Guidance
(that is, March 23, 1995), then the Tribe would be subject to all provisions
of the final Guidance, including the submission requirements contained  in
§ 132.5.  No Tribes-in the Great Lakes basin have yet received such approval
or authorization.

      If any Great Lakes Tribe seeks to receive such approval or authorization
after March 23, 1995, then the Tribes will be subject to all provisions of the
Guidance except the submission requirements in § 132.5.  For example, in order
for a Tribe to receive approval of its water quality standards, assuming it
has been approved to administer a water quality standards program under
section 303, the Tribe will need to adopt and submit to EPA water quality
standards that meet all requirements of part 131, including § 131.5(a)(5),
which requires water quality standards to be consistent with the Guidance in
part 132.  The requirements for submission and EPA review for such standards
are specified in part 131.  Additionally, as discussed further in section
II.D.2 of this document, EPA will apply the general factors in § 132.5(g) for
determining consistency with the final Guidance, as well as all other
applicable requirements of part 131, when approving or disapproving Tribal
water quality standards under § 131.21(b).

      EPA also agrees with comments that the provisions of part 132, including
the submission deadlines in § 132.5, in no way limit any Tribe's rights or
time limits to seek qualification to administer water quality standards or
NPDES programs.

      Upon request, EPA will provide technical guidance and assistance
concerning both basic water quality provisions, and provisions consistent with
the Guidance, to Tribes that have applied or wish to apply for approval to
implement water quality standards or NPDES permit programs.

      c.    Final Guidance.  The final Guidance retains provisions for  Great
Lakes Tribes to adopt water quality standards, antidegradation policies, and
implementation procedures for waters within the Great Lakes System which are
consistent with the final Guidance.  The definition of Great Lakes Tribe has
been modified to clarify that this includes only those Indian Tribes for which
EPA has approved water quality standards under section 303 of the CWA or which
EPA has authorized to administer an NPDES program under section 402 of  the
CWA.

E.    Amendments to NPDES and Water Quality Standards Program Regulations

      1.    Proposal.  The proposal included conforming amendments to the
NPDES and water quality standards regulations in 40 CFR parts 122, 123, and
131 to insert references to the provisions of part 132 applicable to Great
Lakes States and Tribes.  These amendments ensure that future actions under
NPDES and water quality standards programs in the Great Lakes basin will be
consistent with the Guidance.

      2.    Final Guidance.  No specific comments were received concerning
these amendments.  The  final Guidance  retains the proposed amendments to  §§
122.4, 123.25, 123.44,  123.62, 123.63, 131.1, 131.5, and 131.21.  The
amendments to  §§ 122.4, 123.44, and 123.63 were modified to clarify that  they
apply only after provisions consistent with the final Guidance have been
promulgated by the State, Tribe,  or EPA.  The amendment to § 123.63 was
further modified to  clarify that  the criteria for withdrawal of a Great Lakes
State or Tribal NPDES program include  failure to adequately incorporate the
NPDES implementation procedures in 40  CFR part  132 into individual permits.

-------
                           Section II: Regulatory Requirements                        81
      F.    Precedential Effects of Elements of the Guidance

      1.    Proposal.  The requirements in the proposed Guidance were
expressly applicable only to the waters of the Great Lakes System.  EPA
requested comments on whether EPA should issue National guidance or propose
any modifications to 40 CFR parts 122  - 124, 130 and 131 in the future to
correspond with specific elements of the proposed Guidance.

      2.    Comments.  Some comments favored broader application, recommending
that either the entire final Guidance  or all non-Great Lakes-specific aspects
of the Guidance be applied nationally  to help enhance national consistency.

      Many commenters opposed broader  application of the Guidance because they
believe the Guidance was developed for the unique characteristics of the Great
Lakes basin, or because they believe the scientific basis of the Guidance is
unproven or its implementation untried.  Some commenters recommended
developing a national regulation instead of the Great Lakes-specific Guidance,
either because they believe the Guidance will reduce the competitiveness of
the Great Lakes region, or because it  will be counterproductive to have two
different NPDES programs within Great  Lakes States having waters both inside
and outside of the basin.  Other commenters recommended having States outside
the basin adopt existing nationwide EPA guidance and policy to help alleviate
some of the concern that the GLI will  cause a competitive disadvantage for the
Great Lakes basin.

      There are many provisions of the Guidance that might be beneficially
applied in other jurisdictions to improve the national program and foster
consistency.  For example, EPA believes many of the concepts in the
methodology for development of bioaccumulation factors and its use in
developing criteria to protect human health and wildlife could be applied
elsewhere.  At the same time, EPA would not consider applying specific
Guidance elements outside the basin in situations where they are not
scientifically or technically defensible.  For example, the special provisions
for BCCs may not be appropriate in systems not having the long retention times
and other chemical, biological, and physical characteristics of the Great
Lakes System described in sections I.A and I.B of this document.

      During the normal course of developing and improving national water
quality programs, EPA will consider incorporating any elements of the final
Guidance that appear to be scientifically and technically appropriate.  Such
changes could be implemented, for example, as internal guidelines for EPA
staff in developing criteria guidance  under section 304(a), as guidance to
States and Tribes in adopting criteria or implementation procedures into their
programs, or as Federal rulemaking.  Any significant proposed changes
affecting national programs would be announced in advance in order to solicit
comment from the scientific and technical community, as well as the public at
large.  In addition, EPA will seek peer review in accordance with EPA policy,
which states that major scientifically and technically based work products
related to EPA decisions normally should be peer reviewed.  EPA will solicit
external peer review for those work products that are intended to support the
most important decisions or that have  special importance in their own right.

      EPA is currently considering broader application of concepts in the
Guidance methodologies for development of bioaccumulation factors and wildlife
criteria.  In April 1994, EPA staff presented some of these concepts to the
EPA Science Advisory Board for review.  In addition, EPA is in the process of
reviewing and revising the 1980 National Guidelines which would apply to
development of EPA National human health water quality criteria under section
304(a) of the CWA.  If EPA decides to proceed with any of these concepts
nationally,  it will announce its proposal in the Federal Register.

      EPA does not agree that the scientific basis for the final Guidance is
unproven.  The scientific basis for the final Guidance is discussed further in
sections I and III through VIII of this document.  EPA agrees, however, that

-------
82     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

the experience of EPA, Great Lakes States, and Great Lakes Tribes in
implementing the final Guidance will be useful in deciding how elements of the
Guidance can be appropriately applied in other jurisdictions.

      EPA does not agree that the Guidance would lead to two different NPDES
programs within a State.  States are already accustomed to developing
different permit conditions and limitations depending on water quality
conditions in different parts of the State.  The Guidance would simply ensure
that consistent approaches are used for waters of the Great Lakes System.

      EPA also does.not agree with comments recommending a national regulation
in place of the proposed Guidance.  First, the CPA required EPA to issue water
quality guidance for the Great Lakes System, not for the nation.  Therefore,
EPA did not generally provide opportunity for States, Tribes, or members of
the public and regulated community outside the Great Lakes basin to
participate in developing the Guidance.  Second, the short statutory and
judicial deadlines for producing and implementing the Guidance did not allow
time for considering broader application beyond the Great Lakes System or
obtaining appropriate public comment.  Third, EPA does not believe there will
be significant detrimental effects to the economy of the Great Lakes region
that would place the region at a competitive disadvantage to other parts of
the country.  As discussed further in section IX.1 of this document, a study
conducted for the Council of Great Lakes Governors showed that such effects
are expected to be minimal.

      EPA agrees that it would be useful to increase the consistency of water
quality programs applied nationally.  EPA will continue to work toward this
goal in implementing water quality standards and NPDES permit programs
throughout the country.

      3.    Final Guidance.  The final Guidance contains no mandatory
requirements for discharges outside the Great Lakes System.

      EPA would like to reemphasize that the provisions in the proposed and
final Guidance are expressly applicable only to the waters of the Great Lakes
System.  EPA has initiated no rulemaking action to extend any Guidance
provisions beyond'the Great Lakes System.  EPA's request for comments in the
proposal was soliciting views only on whether any future national guidance or
rulemaking affecting water programs beyond the Great Lakes System should
include any concepts contained in the final Guidance.

      States or Tribes with waters outside the Great Lakes System, in whole or
in part, are encouraged to implement any of the Guidance methodologies or
procedures that are scientifically and technically appropriate for their
situations.  This would include any Great Lakes States or Tribes that may
choose to apply some or all of the Guidance in portions of their jurisdictions
outside the Great Lakes System.   Some Great Lakes States indicated their
intention to do so  in their written comments on the proposed Guidance.

G.    Implementation of Endangered Species Act

      1.    Proposal.  The proposed Guidance contained no specific provisions
concerning the protection  of endangered or  threatened species under  the
Endangered Species Act  (ESA) .  EPA invited  comment on several possible
approaches that EPA was considering including in the Guidance to reflect EPA's
responsibilities under  the ESA  (58 FR 20848-20849).  Under section  7(a)(2)  of
the ESA, each Federal agency shall, in consultation with the U.S. Fish  and
Wildlife Service  (FWS), or the National Marine  Fisheries Service for species
under its jurisdiction, ensure that actions authorized, funded  or carried  out
by the Federal agency are  not likely to jeopardize the  continued existence of
any endangered or threatened species listed under  section 4  of  the  ESA,  or
result in the destruction  or adverse modification  of  such species'  critical
habitat  (i.e., are  not  likely to  "cause jeopardy"  to  "listed species").

-------
                           Section II: Regulatory Requirements                        83
 (There are no species tinder the jurisdiction of the National Marine Fisheries
Service in the Great Lakes basin.)

      2.    Consultation with the U.S. Fish and Wildlife Service.  Procedural
requirements for the consultation envisioned in section 7(a)(2) of the ESA  are
found in 50 CFR part 402.  These regulations provide for two types of
consultation: formal and informal.  Formal consultation procedures are
required in situations where a Federal action may affect listed species or
critical habitat.  The formal procedures include requirements for the Federal
agency to provide available information to the FWS, and for the FWS to develop
a Biological Opinion discussing the effects of the Federal action on the
listed species.  Informal consultation is an optional procedure that includes
discussions and correspondence between the FWS and the Federal agency to
determine whether an action is likely to adversely affect a listed species.
If the Federal agency finds, and the FWS concurs in writing, that the action
is not likely to adversely affect listed species, consultation is concluded
and formal consultation is not necessary.

      In January 1993 EPA and the FWS initiated informal consultation under
section 7 of the ESA concerning the issuance of the final Guidance.  The two
agencies identified several specific issues for detailed consultation.  In
September 1994, EPA initiated formal consultation with the FWS on two of the
issues: adequacy of the Guidance's aquatic life criteria methodologies and
related implementation procedures to protect endangered mussel species in the
Great Lakes basin, and the appropriateness of the Guidance's methodology to
develop wildlife criteria.  On February 21, 1995, the FWS provided EPA with a
written Biological Opinion.  The Opinion is available in the docket for this
rulemaking.  The results of the formal and informal consultation are as
follows:

      a.    Protection of Endangered Mussel Species

      EPA and the FWS were concerned about the effects of water quality
resulting from implementation of the final Guidance on three listed endangered
freshwater mussel species in the Great Lakes System: Pleurobema clava
 (Clubshell), Epioblasma torulosa ranqiana  (Northern riffleshell), and E.
obliouata peroblioua (White cat's paw pearly mussel).  The agencies were
concerned that since there is very limited aquatic toxicity information for
these species or their surrogates, it is difficult to assess whether the
aquatic life criteria methodology in the final Guidance produces criteria that
are stringent enough to protect the endangered species.

      EPA and the FWS undertook a formal consultation to address this issue.
EPA provided information about the proposed Guidance and its expected effects
on mussel species to the FWS, including the best scientific and commercial
data available.  The FWS then carried out a review and evaluation that
resulted in the February 21, 1995, Biological Opinion.

      The Biological Opinion concluded that the water quality resulting from
implementation of the final Guidance is not likely to jeopardize the continued
existence of the mussel species.  The FWS identified possible concerns,
however,  that the aquatic life methodologies in the final Guidance may not  be
stringent enough to avoid adverse effects to listed mussels in all cases.   The
FWS conducted risk assessment analyses using existing toxicity studies on
freshwater mussels.  Although such studies are limited in number, and were  not
used by EPA either because they were not available in time for inclusion or
because they did not meet EPA data requirements, the FWS believes that
freshwater mussel data would provide a better indication of the sensitivity of
the listed mussels than data based on less closely related test organisms.  In
situations such as this, where information on listed species is limited, the
policy of the FWS is to provide the "benefit of the doubt" to the species
concerned (51 FR 19952, June 3, 1986; H.R. GOnf. Rep. No. 697, 96th Cong.,  2d
Sess. at 12 (1979)).  The FWS assessment resulted in effects concentrations
lower than the Guidance aquatic life criteria for some contaminants.  Based on

-------
84     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

these results, the FWS concluded that the water quality resulting from
implementation of the aquatic life criteria and values in the final Guidance
may result in "incidental take" of an unquantified number of endangered
freshwater mussels in the Great Lakes basin.

      Section 9 of the ESA prohibits  "take" of listed species of fish or
wildlife--that is, .harassing, harming, pursuing, hunting, shooting, wounding,
killing, trapping, capturing or collecting such species, or attempting to
engage in any such conduct--unless the take is exempted from this provision as
discussed below.  FWS policies further define "harm" to include significant
habitat modification or degradation that results in death or injury to listed
species by significantly impairing behavioral patterns such as breeding,
feeding, or sheltering  (50 CFR 17.3).  Under the terms of section 7(b)(4) and
7(o) (2) of the ESA, taking that is "incidental" to and not intended as part of
the Federal agency'p action is not considered a prohibited taking provided
that such taking is in compliance with requirements in the Biological Opinion
specifying reasonable and prudent measures necessary and appropriate to
minimize the incidental take, and specifying mandatory terms and conditions to
be followed.

      To minimize the amount or extent of any incidental take resulting from
implementation of the final Guidance, the FWS consulted closely with EPA to
develop a coordinated approach.  The final Biological Opinion specified
reasonable and prudent measures that the FWS considers necessary or
appropriate to minimize such impact as the aquatic life criteria,
methodologies, and related implementation procedures are implemented in the
Great Lakes System as specified in the final Guidance.  These measures are as
follows:

      --  During the next two years, EPA will undertake a screening analysis
consisting of acute toxicity testing of freshwater mussels for a limited
number of pollutants.  During the same period, to enable successful completion
of the measures, FWS will develop and analyze life history information on the
endangered mussels.  The life history information is particularly important in
identifying surrogate species, developing test protocols, and culturing test
organisms, and further determining what other environmental parameters may be
contributing to the'decline of the species.

          EPA and FWS will review the results from the above efforts.  The
purpose of the review is to compare the sensitivity of the surrogate mussel
species with the sensitivity of other species used in developing acute
criteria under the final Guidance.  The review will also help determine the
types of testing needed to conduct a  screening analysis for chronic effects.

          During a subsequent period  of approximately four years, EPA and FWS
will conduct a screening analysis consisting of selected chronic toxicity
testing of freshwater mussels for a limited number of pollutants.  The scope
and design of this testing will depend heavily on the results of the efforts
of the first two years described above.

          Based on the combined results of the acute and chronic screening
analyses, EPA and FWS will assess whether there is evidence that the three
endangered mussels or their surrogate species are more sensitive to aquatic
pollutants than other species assessed as part of EPA and State criteria
development efforts under the final Guidance.  If so, then the agencies will
develop a joint approach for developing the necessary information to adjust
criteria as necessary to protect the  species.
                    *
          EPA and the FWS will develop a protocol that can be used  to monitor
any populations of  listed mussels downstream from new discharges permitted
under the final Guidance.  EPA will then seek through its oversight of State
water quality programs  to ensure  that monitoring of such populations takes
place in accordance with the protocol to identify whether any incidental  take
occurs  that is  likely to cause jeopardy to  the  species.

-------
                           Section II: Regulatory Requirements                        85
      EPA believes that the methodologies and procedures  in the  final Guidance
will be protective of endangered or threatened aquatic species in  the Great
Lakes System.  EPA acknowledges that it  is possible that  data collected  in the
future, however, could identify an unusual sensitivity of mussels  to a
particular pollutant that is not reflected in the methodologies  and
procedures.  Based on experience in implementing water quality criteria  under
the CWA, EPA believes that this possibility is small, and that the degree of
any adverse effects would likely be limited.  Nevertheless, EPA  agrees that it
would be desirable to obtain additional  data on freshwater mussels to confirm
its belief that mussels will not be adversely affected by water  quality
resulting from implementation of the final Guidance.  For this reason, EPA has
agreed to implement the reasonable and prudent measures described  above.

      b.    Wildlife Criteria Methodology

      EPA and the FWS were concerned about the effects of water  quality
resulting from implementation of the final Guidance on five wildlife species
listed as endangered or threatened in the Great Lakes basin: Canis lupus (Gray
wolf), Mvotis sodalis  (Indiana bat), Haliaeetus leucocephalus  (Bald eagle),
Palco pereorinus  (Peregrine falcon), and Charadrius melodus  (Piping plover).
Of these, the primary concern was the bald eagle because  of its  relatively
greater dependence on food sources found at higher levels of the aquatic food
chain.

      EPA and the FWS undertook a formal consultation to  address this issue.
EPA provided information to the FWS about the wildlife methodology and its
predicted effects on bald eagles.  The FWS then carried out a review and
evaluation that was included in the February 21, 1995, Biological  Opinion.

      The Biological Opinion concluded that the final Guidance is  not likely
to jeopardize the continued existence of the bald eagle,  peregrine falcon, or
piping plover.  The FWS identified possible concerns, however, that the
wildlife methodology in the final Guidance may not be stringent  enough to
protect endangered or threatened species in all cases.  In their risk
assessment, the FWS used assumptions and assessment approaches that differ
from those developed by the Initiative Committees, whose  approach  was reviewed
by the EPA Science Advisory Board and used in the proposed and final Guidance.
The model developed by the FWS is based  on the dose to the target  tissue (the
egg), rather than the dose received in the diet.  The FWS assessment also
considers the protection of individuals  as opposed to populations, and
emphasizes the need for toxic equivalency factors to account for additive or
synergistic effects of complex mixtures  of bioaccumulative compounds to
wildlife.  Based on the results of their assessment, which gives the "benefit
of the doubt" to the species in accordance with FWS policy as discussed  above,
the FWS concluded that the water quality resulting from implementation of the
final Guidance may result in "incidental take" of an unquantified  number of
bald eagles, peregrine falcons, and piping plovers in the Great  Lakes basin
due to detrimental effects resulting from chronic toxicity.

      To minimize the amount or extent of any incidental  take resulting  from
implementation of the final Guidance, the FWS consulted closely  with EPA to
develop a coordinated approach.  The Biological Opinion specified  reasonable
and prudent measures that the FWS considers necessary or  appropriate to
minimize such impact as the wildlife criteria, methodologies, and  related
implementation procedures are implemented in the Great Lakes basin as
specified in the final Guidance.  These measures are as follows:

          During thfe nine years following publication of  the final Guidance,
EPA will gather information concerning the impacts on endangered or threatened
species in the Great Lakes basin from the four chemicals  with numeric wildlife
criteria in the final Guidance.  If at any time new information  indicates that
water quality after full implementation of the wildlife components of the
Guidance may adversely affect endangered or threatened species,  EPA will
immediately reinitiate consultation.

-------
86     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

          EPA will establish and begin implementing a data-gathering plan, in
cooperation with FWS and the States or Tribes, to determine progress in
reducing the concentrations of the four chemicals in the Great Lakes aquatic
ecosystem, and to help improve understanding of the sensitivities of the
listed species to the chemicals if possible.

          EPA will during the next five years reevaluate and better define
loadings of BCCs to the Great Lakes System that might affect listed wildlife
species, including a report describing ongoing activities (e.g., Lakewide
Management Plans and Remedial Action Plans) that are designed to reduce
current levels of contamination in the basin.

          EPA and FWS will cooperatively plan and host one or more workshops
to explore the utility of additivity models in developing wildlife criteria,
to consider the scientific basis for use of toxicity equivalency factors
(TEFs) when establishing total maximum daily loads or water quality-based
effluent limits to implement wildlife criteria, and to discuss research and
data needs.  The -findings of these workshops will be used to develop
approaches for utilizing TEFs if appropriate in the development and
implementation of wildlife criteria.  These approaches will be presented to
the EPA SAB for review and comment.

      EPA believes that the wildlife criteria methodology in the final
Guidance, including the procedures for site-specific criteria modifications,
will be protective of endangered or threatened wildlife species in the Great
Lakes basin.  Nevertheless, EPA believes it will be useful to conduct the type
of review described above to confirm that water quality resulting from
implementation of the final Guidance is not likely to adversely affect
endangered or threatened wildlife species.  For these reasons, EPA has agreed
to implement the reasonable and prudent measures described above.

      c.    Other Issues

      As a result of informal consultation, and after considering comments
(discussed below), EPA decided to include certain additional provisions in the
final Guidance to ensure protection of endangered or threatened species
consistent with EPA's responsibilities under section 7.  These provisions are
discussed below.  They generally reflect the types of.provisions discussed in
the preamble to the proposed Guidance  (58 PR 20848-49) .

      EPA and the FWS consulted informally on specific regulatory provisions
to ensure that aquatic life and wildlife criteria will be developed or
modified on a site-specific basis to protect endangered or threatened species
in the Great Lakes System.  The two agencies agreed that procedure 1 of
appendix F should be modified to provide that Great Lakes States and Tribes
must develop more stringent site-specific modifications of criteria and values
to protect listed and proposed species, where such modifications are necessary
to ensure that water quality is not likely to cause jeopardy to the species.
Because efforts to protect and promote recovery of listed and proposed species
generally focus on management of specific ecosystems, it is appropriate that
States, Tribes, and EPA utilize the procedure for site-specific modifications
of criteria and values to address conditions in those ecosystems.

      In the above provisions, EPA has included proposed as well as listed
species.  EPA believes that prevention of water quality conditions that would
be likely to jeopardize the continued existence of any species  is inherent in
the CWA's requirement for EPA to provide guidance on minimum water quality
standards to protect aquatic life and wildlife in the Great Lakes basin  (CWA
section 118(c) (2)), as well as the fundamental principle that water quality
criteria must protect designated uses  (CWA  section 303(c) (2) (A); 40 CFR
131.11(a) (2)) .  Obviously, if impaired water quality will likely cause the
extinction of a species, such water quality would not meet the Act's
requirements.  In a sense, then, mandating that States adopt site-specific

-------
                           Section II: Regulatory Requirements                        87
criteria to avoid causing jeopardy to any species is simply an explicit
statement of a principle that EPA believes is inherent in the CWA.

      The two agencies also agreed that States and Tribes should be encouraged
to adopt more stringent site-specific modifications of criteria or values
where necessary to protect candidate "Cl" species being considered by the FWS
for listing under section 4 of the ESA.  The "Cl" category refers to species
for which the FWS has enough substantial information to support proposals to
list as endangered or threatened under section 4 of the ESA.  The new
provisions of procedure 1 of appendix F to part 132 are discussed under the
Final Guidance section below.

      EPA and the FWS also consulted informally on concerns about whether the
final Guidance should apply to ammonia and chlorine.  The two agencies agreed
that EPA will undertake a review of water quality standards and implementation
of those standards for ammonia and chlorine in the Great Lakes basin as part
of EPA's responsibilities under section 303(c)  of the CWA.  This review is
discussed further in section II.C.B.c of this document.

      The two agencies also consulted informally concerning the adequacy of
the Guidance methodologies for developing criteria and values to protect
wildlife from certain pollutants that are metabolized by higher trophic level
aquatic organisms but not by lower trophic aquatic organisms.  Such pollutants
include, for example, polynuclear aromatic hydrocarbons (PAHs) which, because
of metabolism by fishes, have relatively high BAFs when measured in lower
trophic level species, even though they have relatively low BAFs in higher
trophic level fish.  The methodology for development of BAFs for organic
chemicals in appendix B includes procedures for calculating field-measured
and/or predicted BAFs for all trophic levels, and allows,  on a case-specific
basis, wildlife BAFs to be weighted to reflect the proportion of plants,
invertebrates, and fish in the diet of the species to be protected.

      If a listed wildlife species in the Great Lakes basin were found to be
more highly exposed to a bioaccumulative pollutant because of its diet than
the representative species used in the wildlife criteria methodology, then
procedure l.A of appendix F would specify that a site-specific modification is
needed to protect such species.  The site-specific modification could utilize
the method described above to weight the BAFs based on the diet of the listed
species.

      The two agencies agreed that the final methodology for the development
of bioaccumulation factors in appendix B, and procedure 1 of appendix F for
site-specific modifications, provide adequate protection for listed wildlife
species that may consume highly contaminated prey at lower trophic levels.

      The two agencies also agreed that the final Guidance should clarify
EPA's responsibilities for protecting endangered or threatened species when
reviewing and approving State and Tribal submissions under § 132.5.  Section
132.5(h) has been added to the final Guidance to require States and Tribes to
include in their part 132 submissions any provisions that EPA determines,
based on EPA's authorities under the CWA and the results of consultation under
section 7 of the ESA, are necessary to ensure that water quality is not likely
to cause jeopardy to listed species.  This provision is discussed under the
Final Guidance section below.

      Finally, the two agencies discussed and have agreed upon the procedures
the two agencies will use to conduct section 7 consultations where necessary
on future EPA actions.  These actions include EPA approval or disapproval of
submissions by Great Lakes States and Tribes under § 132.5, and EPA approval
actions for State and Tribal water quality standards, criteria modifications,
and variances under 40 CFR part 131.  The procedures are summarized under the
Final Guidance section below.

-------
88     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      With the issuance of the FWS1 Biological Opinion, formal consultation on
the issues concerning freshwater mussels and wildlife criteria has been
completed.  For the remaining issues, as a result of the informal consultation
described 'above, the FWS concurs with EPA that the final Guidance is not
likely to adversely affect endangered or threatened species in the Great Lakes
basin ecosystem.  Therefore, consultation concerning EPA's publication of the
final Guidance has been completed.

      3.    Comments.   Some comments asserted that specific ESA provisions in
the final Guidance would be an unnecessary additional burden on the States,
since most States have a state program for the protection of endangered or
threatened species which deals directly with the FWS.  Several commenters
supported using site-specific modifications to provide protection for
endangered or threatened species.  Many of these commenters preferred that
such a provision be guidance.  Other comments asserted that full consultation
by EPA with the FWS on administration of all elements of the final Guidance is
necessary, and that consultation procedures should be clarified.

      EPA does not agree that regulatory provisions regarding ESA are
unnecessary in the final Guidance.  The ESA imposes specific responsibilities
on Federal agencies to protect endangered or threatened species in actions the
agencies carry out.  EPA believes that the final Guidance should articulate
the conditions that.EPA,  based on its consultation with the Service, believes
are necessary to include in State and Tribal submissions to provide the
protection for endangered or threatened species envisioned in section 7(a)(2)
of the ESA.  For example, procedure l.A of appendix F clarifies the role of
site-specific modifications to water quality criteria in protecting endangered
or threatened species, and § 132.5(h) of the final Guidance clarifies that EPA
will make a determination based on EPA's authorities under the CWA and the
results of section 7 consultation concerning the adequacy of State or Tribal
provisions for protecting listed species contained in submissions under part
132.  EPA believes these clarifications will be helpful in ensuring an orderly
approach to protect listed species.  Furthermore, although individual States
may have State-imposed responsibilities to protect endangered or threatened
species, these responsibilities may not coincide in detail with applicable
Federal standards.  For example, a State's endangered species program may
address a list of species that could be different from the species listed
under section 4 of the ESA.  The clarifications in part 132 concerning the
Federal requirements therefore will be helpful in reducing confusion on such
matters.

      EPA agrees for the reasons discussed under section II.G.2 above that
site-specific modifications to criteria and values should be required to
protect endangered or threatened species.  EPA also agrees that the final
Guidance should provide States and Tribes flexibility in achieving such
protection, and therefore provides a choice of methods for adopting site-
specific modifications.  This flexibility in adopting site-specific
modifications to protect endangered or threatened aquatic and wildlife species
is discussed further in section VIII.A of this document.

      4.    Final Guidance.  As a result of consultation with the FWS, and in
response to comments, EPA has made the following changes in the final
Guidance:

      EPA has added.§ 132.5(h) to the final Guidance which provides that Great
Lakes States and Tribes will need to include provisions that EPA determines,
based on the results of consultation under  section 7 of the ESA, are necessary
to ensure that water quality is not  likely  to jeopardize the continued
existence of any endangered or threatened species listed under  section 4 of
the ESA or result in the destruction or adverse modification of such species'
critical habitat.

      To carry out  the ESA, relevant EPA Regional Offices will  initiate
consultation with the FWS as soon as possible concerning EPA's  approval

-------
                           Section II: Regulatory Requirements                        89
actions under part 132 for each of the Great Lakes States and Tribes.  Such
consultations will be facilitated if States and Tribes have already consulted
with the FWS under State and Tribal arrangements with FWS where such
arrangements exist.  Initiating consultations early in the process of the
States' and Tribes' development of their part 132 submissions is appropriate
and necessary because under section 118 (c) (2) (C) of the CWA and § 132.5, the
States .and Tribes must complete their part 132 submissions to EPA generally by
September 23, 1996, and EPA must approve such submissions, or promulgate EPA
requirements, by March 23, 1997.  EPA believes the consultations can and
should occur simultaneously with the States' and Tribes' efforts to develop
their part 132 submissions.

      As a result of the consultation that has taken place with the FWS
concerning the final Guidance, EPA believes that consultations can be
concluded rapidly and routinely concerning part 132 submissions from States or
Tribes that are consistent with the final Guidance.  EPA will make every
effort to expedite exchange of information and consultation between FWS staff
members, State or Tribal representatives, and EPA Regional Office staff
members to help determine whether any additional modifications to a particular
submission are necessary, such as site-specific modifications.  If such
situations occur, EPA will notify the State or Tribe as soon as possible.

      EPA has consulted with the FWS regarding the methodologies for
development of criteria and values in the final Guidance.  Based upon
available data and information, EPA and the FWS have concluded that criteria
and values consistent with the final Guidance generally will result in water
quality that is not likely to jeopardize the continued existence of listed
species or destroy or adversely modify critical habitat.  Thus, in the absence
of a need for a site-specific modification, discussed below, States and Tribes
adopting criteria and values, methodologies, policies, and implementation
procedures consistent with the final Guidance will be approvable by EPA as
consistent with the ESA.

      EPA recognizes that information may become available--either prior to or
subsequent to approval of State and Tribal submissions under part 132--
indicating that more stringent criteria/values may be necessary to protect
certain species.  Therefore, EPA has modified procedure l.A of appendix F of
the final Guidance to provide that Great Lakes States and Tribes need to
develop more stringent site-specific modifications of criteria and values to
protect listed or proposed species, where such modifications are necessary to
ensure that water quality is not likely to cause jeopardy to listed species.
As discussed above, this provision is necessary to clarify the role of site-
specific modifications in protecting endangered or threatened species listed
or proposed under the ESA,  Procedure l.A also encourages States and Tribes to
adopt more stringent site-specific modifications of criteria or values where
necessary to protect candidate (Cl) species.

      Procedure l.A of appendix F provides that such site-specific
modifications be made where they are necessary to ensure water quality is not
likely to cause jeopardy to listed or proposed species.  EPA interprets the
phrase "where necessary" to include geographic sites where the species may be
affected, and where critical habitat has been designated.  Procedure 1 of
appendix F allows flexibility in interpreting the geographic extent of
"sites."  Therefore, States, Tribes, and EPA in consultation with the FWS,
should establish the geographic extent of the site-specific modifications to
be large enough to provide adequate protection to the species, including
survival and recovery.  For example, a site-specific modification could be for
an area as small as a local stream reach where a listed species currently
resides, or as large as the entire portion of the State located within the
Great Lakes basin, depending on the needs of the species.

      EPA also interprets the phrase "where necessary" to mean where available
data indicate that listed or proposed species are especially vulnerable to
specific pollutants, such as being more sensitive than the species used to

-------
90     Water Quality Guidance for the Great Lakes System — Supplementary Information Document


develop existing criteria, or being more exposed than assumed in the Guidance
methodologies.

      The GLI Clearinghouse, discussed in section II.C.I of this document,
will facilitate the incorporation of new data on listed, proposed, or
candidate  (Cl)  species under the ESA, or surrogates for such species, into
State and Tribal prpgrams.  Any such data will be placed in the EPA-operated
Clearinghouse,- where they will be available to States, Tribes, and other
interested persons.  EPA will place special emphasis in the Clearinghouse on
data relevant to protection of such species.  EPA will alert States and Tribes
to data that indicate unusual sensitivity of such species in the Great Lakes
basin.  EPA and the FWS will also work with States and Tribes to identify
sites where special protection may be recommended.

      EPA will also work with States and Tribes to disseminate any new data
that become available in future years showing that a site-specific
modification to criteria is appropriate because a listed or proposed species
is especially sensitive to aquatic pollutants.  For example, if data relevant
to a listed aquatic species become available concerning a SMAV that is less
than the calculated FAV, or that would yield a lower criterion or value using
the EPA recalculation procedure or resident species procedure, a site-specific
modification would be necessary for waters where the species is found.  If the
State or Tribe already had numeric water quality criteria for the pollutant,
the State or Tribe would need to adopt a site-specific modification and submit
it to EPA for approval as currently specified in § 131.20.  EPA would then
consult with the FWS concerning the approval of this submitted revision to the
water quality criteria.  If the State or Tribe did not submit such a
modification, or did not adopt criteria protective of the listed species, and
did not demonstrate to EPA why such modifications are not necessary, EPA would
have the authority under section 303(c) of the CWA to disapprove the State
criteria for relevant areas and take necessary actions.  Under section 303(c),
EPA would then be required to propose Federal criteria in place of the State
or Tribal criteria.

      If the State or Tribe had not adopted criteria for the pollutant of
concern, it would nevertheless need to utilize the new data in developing
criteria or values to be used as appropriate in establishing water quality-
based effluent limits for the pollutant, in accordance with § 122.44(d),
§ 132.4, and adopted State or Tribal provisions consistent with the
implementation procedures in appendix F of the final Guidance.

      EPA has added provisions to final procedures I.A.I and 1.A.2 of appendix
F to clarify ways that the site-specific modifications required in procedure
l.A may be accomplished.  For aquatic life, procedure l.A.l(c) describes two
ways criteria may be modified to reflect endangered or threatened species that
are particularly sensitive to a pollutant.  For example, the Species Mean
Acute Value  (SMAV) for the listed or  surrogate species may be found to be
lower than the calculated Final Acute Value  (FAV) developed using the Tier I
methodology in appendix A.  In this situation, the lower SMAV would be
substituted for the' calculated FAV in developing the aquatic life acute
criterion.  Alternatively, the State  or Tribe may lower the criterion using
EPA's recalculation procedure or resident species procedure as provided in
Chapter 3 of the U.S. EPA Water Quality Standards Handbook, Second Edition
 (August 1994).  Procedures I.A.I and  l.A.2 are discussed further in section
VIII.A of this document.  During the  consultation process EPA will solicit the
views of the FWS concerning the choice of method for criteria modification.

      For wildlife, procedure l.A.2 describes a recommended approach for
modifying wildlife criteria to protect endangered or threatened  species.
Procedure l.A.2 is discussed further  in section VIII.A of this document.

      EPA has added provisions to appendix E and to procedures 2 and 3 of
appendix F of the final Guidance that restrict certain actions States and
Tribes may take to allow  lowering of  water quality in high quality waters, to

-------
                           Section II: Regulatory Requirements                        91
grant variances, and to allow mixing zones.  Section I.A of appendix E
specifies that any lowering of water quality in high quality waters shall be
prohibited if such lowering of water quality will be likely to cause jeopardy
to listed species.  It is discussed further in section VII of this document.
Procedure 2.A of appendix F specifies that a variance to a water quality
standard may not be granted that would result in water quality likely to cause
jeopardy to listed species.  It is discussed further in section VIII.B of this
document.  Procedure 3 of appendix F includes provisions that mixing zones in
or outside the context of a TMDL may be allowed, at a minimum, only to the
extent that water quality would not likely cause jeopardy to listed species.
these provisions are discussed further in section VIII.C of this document.

      EPA has determined that each of these restrictions is necessary to
clarify how the requirements of the ESA affect decisions regarding
implementation of water quality standards.  EPA believes including these
provisions in the final Guidance will be helpful to States, Tribes, and
dischargers to understand the provisions that EPA believes will need to be
included in part 132 submissions to meet ESA requirements.

      Under section 7 of the ESA, EPA is obligated to consult with the FWS
concerning actions carried out, authorized, or funded by EPA that may affect
listed species.  As discussed above, EPA will consult with the FWS concerning
EPA's actions to approve State and Tribal submissions under § 132.5.  EPA also
intends to consult on approval actions under 40 CFR part 131 for water quality
standards in the Great Lakes basin, including criteria modifications and water
quality standards variances, that may affect listed species, submitted as
revisions to a State's or Tribe's water quality standards.  The FWS and EPA
recognize that to accomplish timely implementation of standards that may
affect listed species, early involvement and technical assistance by the FWS
is needed.  For example, EPA and the FWS will meet during the triennial period
for reviews of water quality standards under section 303(c) of the CWA,
preferably during the period when EPA and a State or Tribe discuss the extent
of an upcoming review.  EPA and FWS agree that meetings, discussions, and
exchanges of information throughout the process--from State and Tribal
development, adoption, submission of standards, through EPA final action--will
facilitate the consultations required by the ESA.  EPA will also invite and
encourage the States and Tribes to participate actively in the consultations.

-------
                               Section ID: Aquatic Life                             93
                           .  AQUATIC LIFE
A.    Summary of Final Rule

      The final Guidance is identical to the Guidance proposed on April 16,
1993, except for the following changes:

      -- States and Tribes are permitted to use either the averaging period
and frequency of exceedance specified for the proposed Tier I criteria
concentrations, or some other scientifically acceptable averaging period
and/or frequency of.exceedance.   If EPA determines that it is necessary to
promulgate criteria for a State or Tribe, it will promulgate averaging periods
of one hour  (Criterion Maximum Concentration) and four days (Criterion
Continuous Concentration) and a frequency not to exceed the criteria of more
than once-in-three-years (both CMC and CCC).

      -- Aquatic life criteria for the nine metals proposed in Tables 1 and 2
of part 132 are expressed as dissolved concentrations and total recoverable
concentrations.  Conversion factors for the proposed total recoverable metals
criteria were derived and publicly noticed on August 30,  1994 (59 FR 44678) .
EPA adjusted these conversion factors and used them in the calculations to
convert the proposed criteria from total recoverable concentrations to
dissolved concentrations for the final Guidance.  If a State or Tribe fails to
adopt aquatic life criteria for metals, EPA will promulgate dissolved
criteria.

      - - The final Guidance clarifies and elaborates on the option of using
indicator parameter limits under 40 CFR 122.44(d)(1)(vi)(C).   A full
discussion of the use of the indicator parameter option and its role in the
final Guidance appears in section VIII.E.2.f of this document (Reasonable
Potential Procedure).  As discussed in that section of this document the
States or Tribes are still required to adopt provisions consistent with the
Tier II methodology for aquatic life consistent with the methodology in
appendix A.  As described in procedure 5 of appendix F to part 132, States and
Tribes will then be required under the final Guidance to implement the Tier II
values through water quality-based effluent limitations (WQBEL)  based on such
values when, under procedure 5 of appendix F to part 132,  such limits are
determined to be required.   When deriving limits to meet Tier II values,
States and Tribes have the option of using an indicator parameter limit in
lieu of a Tier II-based limit.  If an indicator parameter is used, the State
or Tribe must ensure that the indicator parameter will attain the "applicable
water quality standard"  (as described in 40 CFR 122.44(d) (1) (vi) (C) ) .  The
"applicable water quality standard" in this instance would be the State's or
Tribe's narrative water quality criteria that protects aquatic life,  as
interpreted using the Tier II methodology.

B.    Final Tier I Methodology

      The final Guidance includes methodologies for deriving Tier I criteria
and Tier II values for the protection of aquatic life in the Great Lakes
System.  The Aquatic Life Tier I methodology is an adaptation of the
"Guidelines for Deriving Numerical National Water Quality Criteria for the
Protection of Aquatic Organisms and Their Uses"   (U.S.  EPA,  1985)   (1985
National Guidelines), developed under section 304(a) of the Clean Water Act
(CWA).   A copy of the 1985 National Guidelines is available in the
administrative record for this rulemaking.  The 1985 National Guidelines can
also be obtained through the National Technical Information Service (PB 85-
227049) .   Provisions consistent with the final Tier I methodology must be used
when deriving Great Lakes aquatic life criteria for use in State and Tribal
water quality standards.  The final Guidance also contains Tier I criteria for

-------
94     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

16 pollutants that are based on the latest scientific knowledge and derived
using an extensive aquatic toxicological data base.

      As stated above, the proposed Tier I methodology was an adaptation of
the 1985 National Guidelines which has previously undergone scientific  peer
review and public review and comment, and had been revised where appropriate.
EPA stated in the proposal that issues already addressed by EPA in response to
previous comments on the 1985 National Guidelines would not be addressed in
this proceeding  (58FR20852).  Therefore, the following discussion focuses on
those portions of the Tier I methodology which differ from the 1985 National
Guidelines.

1.    Required Data

      a.    Proposal:  In the proposed Guidance, two ambient criteria
concentrations to protect aquatic life for any given pollutant are derived
(the Criterion Maximum Concentration  (CMC) and the Criterion Continuous
Concentration  (CCC)).  In order to derive a CMC for a pollutant, it is
necessary that acceptable acute toxicity tests exist for aquatic animals in at
least eight families which represent differing habitats and taxonomic groups.
The CCC is the lowest of the Final Chronic Value  (FCV) or the Final Plant
Value  (FPV)  (see appendix A for a complete description of how to derive a CCC
or CMC).  The aquatic life methodologies to derive both Tier I criteria and
Tier II values did allow the use of some data for freshwater species that do
not reside in the Great Lakes System.

      b.    Comments:  Several commenters recommended that the proposed
Guidance derive different criteria for coldwater versus warmwater assemblages
of aquatic species, and possibly for other receiving water categories  (based
on inclusion or exclusion of different species in the underlying data set).
Several commenters also recommended limiting the toxicity data set to species
residing in the Great Lakes basin.

      EPA recognizes the potential differences between coldwater versus
warmwater aquatic species and that there are varied aquatic species
assemblages across the Great Lakes System.  Because of this, EPA has included
provisions in the final Guidance to calculate site-specific modifications to
criteria to adjust for  site-specific differences in water quality
characteristics, species, etc.  EPA believes it is better to derive site-
specific criteria where necessary than to try to define and protect generic
assemblages of aquatic  species  (such as warmwater, coldwater, etc.).  Site-
specific criteria can be developed to protect the types of species at each
site and, therefore, are more precise.  If a State or Tribe wishes to modify
the criteria on a site-specific basis to account for a different species
assemblage at the site, the recalculation procedure as described in appendix L
of the U.S. EPA Water Quality Standards Handbook, Second Edition - Revised
 (U.S. EPA, 1994) may be used.  The 1994 Water Quality Standards Handbook is
available in the administrative record for this rulemaking and can also be
obtained through the Water Resource Center, 401 M. Street, SW  (RC4100),
Washington, DC 20460 or by calling  (202) 260-7786.

      EPA disagrees that toxicity data should be  limited to those species
residing in the Great Lakes System.  EPA believes that species nonresident to
the Great Lakes  System  can represent untested resident species and may  used as
surrogates for those resident species not yet tested.  This approach is
consistent with  the 1985 National Guidelines.

       c.     Final Guidance:  The data requirements in the final Guidance for
deriving Tier  I  criteria are the  same as the proposed Guidance.  EPA would
like to clarify  that all data used in the derivation of Tier I criteria or
Tier II values must be  from  species which are resident to North America.

-------
                               Section ffl: Aquatic Life                            95
2 .    Commercially and Recreationallv Important Species

      a.    Proposal :  The 1985 National Guidelines stated that the Final
Acute Value  (FAV) must be set equal to the lower 95th percentile value, or the
Species Mean Acute Value for a species of commercial or recreational
importance.  The proposed Tier I methodology differed from the 1985 National
Guidelines by specifying that the FAV should be lowered only for a species
that is recreationally or commercially important to the Great Lakes System.
For example, the proposed CMC for cadmium was lowered from 2.3 M9/L to 2.1
     to protect the Chinook salmon.
      b.    Comments :  Several commenters noted that the commercially and
recreationally important species used to lower some of the criteria do not
reside in all receiving waters of the Great Lakes System.  Those commenters
suggested that the criteria should be lowered to protect such species where
they reside, and should not be lowered where they do not reside.

      EPA agrees that some of the recreational and commercially important
species used to lower criteria do not reside in all receiving waters of the
Great Lakes Basin.  However, many of these species are located over a large
area of the Basin and migrate throughout the lakes and tributaries.  EPA
believes it is reasonable to adjust the criteria for these species System-wide
instead of site-specifically as advocated by the commenters.  The current
National program and the final Guidance allow States and Tribes to modify the
aquatic life criteria on a site -specific basis to account for the situation
where the species does not reside at a particular site.  To accomplish this, a
State or Tribe may choose to modify the criteria using the recalculation
procedure described, above for those waters where the commercially or
recreationally important species does not "occur at the site."

      c.    Final Guidance :   The final Guidance requires the lowering of the
FAV for commercially or recreationally important species to the Great Lakes
System for the reasons stated above .

3 .    Ecologically Important Species

      a.    Proposal :  In the proposal, EPA invited comment on whether the
Tier I methodology should define "ecologically important" species and include
provisions to lower' the FAV for such species.  EPA also invited comment on
whether "ecologically important" species were adequately protected by the
proposed Guidance without adding additional provisions.

      b.    Comments :  Many commenters expressed a preference not to include a
provision to lower the FAV to protect "ecologically important" species because
they believed that a reasonable definition could not be developed.  A few
commenters favored lowering the FAV to protect "ecologically important"
species.  A few commenters favored treating all species as "ecologically
important," but no commenter offered any other definition of "ecologically
important" species..  Some commenters also recommended adding the concept of
"culturally important."  Many of these commenters named certain species as
"culturally" or "ecologically important;" however, these suggestions did not
identify any new species meriting protection since the named species are also
recognized as commercially or recreationally important species.  Several
commenters suggested that a provision for "ecologically important" species
could be used to provide protection for threatened and endangered species .

      EPA agrees with commenters that it is not necessary to include a
provision for lowering the FAV for "ecologically important" species.  EPA
believes that the Tier I methodology as proposed, utilizing all available
data,  adequately protects "ecologically important" and "culturally important"
species.  Further, EPA is unable to develop meaningful and workable
definitions for these- terms that would distinguish ecologically or culturally
important species from other species .

-------
96     Water Quality Guidance for the Great Lakes System - Supplementary Information Document

      EPA agrees that threatened and endangered species should be protected,
but does not believe the use of "ecologically important" species is the best
mechanism.  Instead, EPA has added a provision to procedure 1 of appendix F to
part 132 specifying that States and Tribes must adopt site-specific criteria
modifications to protect threatened and endangered species that are listed or
proposed under the Federal Endangered Species Act.  This provision is
discussed further in sections II.G. and VIII.A. of this document.  EPA has
determined that site-specific modifications to ensure protection to threatened
or endangered species are more appropriate than modifying .the State-wide or
Tribal aquatic life criteria or values as done for commercially or
recreationally important species, since threatened or endangered aquatic
species tend to be isolated to specific sites.  States, Tribes, and EPA in
consultation with the FWS, should establish the geographic extent of the site-
specific modifications to be large enough to provide adequate protection to
the listed species.  If a threatened or endangered species listed under the
Federal Endangered Species Act is found throughout the State's or Tribe's
waters, then it may be more prudent to adopt a State-wide or Tribal criteria
or value rather than develop numerous site-specific modifications.

      c.    Final Guidance:  For the reasons stated above, the final Guidance
does not include a provision for lowering the FAV for "ecologically important"
species nor provides a definition for "ecologically important."  However,
under CWA section 132.4 (i) which allows States and Tribes to adopt more
stringent provisions, a State or Tribe may independently choose to lower the
FAV to protect "ecologically important," "culturally important," or any other
group of species that it defines.

4.    Elimination of Final Residue Value

      a.    Proposal:   For chronic exposures, EPA proposed to modify the 1985
National Guidelines approach by deleting the option of using the Final  Residue
Value  (FRV) in deriving the Criterion Continuous Concentration  (CCC).   The CCC
was proposed as the lower of the Final Chronic Value  (FCV) or the Final Plant
Value  (FPV).

      b.    Comments:  Many commenters supported eliminating the FRV.  A few
commenters advocated retaining the FRV because it yields lower criteria.

      EPA agrees with the  comments that the FRV should not be utilized  in the
calculation of the CCC.  As explained in the proposal  (58FR20851), the  FRV is
intended to prevent concentrations of pollutants in aquatic species from
affecting the marketability of those  species or affecting the wildlife  that
consume aquatic  life.   Consequently,  EPA believes use of the FRV is redundant
with the methodologies  for human health and wildlife criteria contained within
the final Guidance.  EPA believes that those methodologies will yield more
appropriate criteria to protect humans and wildlife.  Furthermore, the  human
health and wildlife criteria derived  for dieldrin, endrin, and mercury  are
more stringent than the National aquatic life  criteria found in  Table III-4.

       c.    Final Guidance:  The  final Guidance does not require the use of
the FRV in the derivation  of the CCC  for the reasons stated above.

5.    Acute-Chronic Ratios

       a.    Proposal:   The Acute-Chronic Ratio (ACR) is a method of relating
acute  and  chronic toxicities.  To derive an ACR, comparable acute and chronic
toxicity  studies which  have  been conducted  under similar conditions for a
given  species are used.   From  comparable measurements of acute  and  chronic
values, an ACR is calculated by  dividing the measured acute value by the
measured  chronic value.   EPA proposed to follow the  1985 National Guidelines
by requiring ACRs for at  least three  families  of aquatic animals.  The  Final
Acute-Chronic Ratio (FACR) must  be either the  geometric mean of some or all  of
the species ACRs or another  value  appropriate  for  sensitive  species.

-------
                               Section HI: Aquatic Life                            97
      The proposal expressed a. preference for the use of freshwater ACRs when
deriving a Final Chronic Value  (FCV),  but allowed the use of ACRs for
saltwater species in the derivation of the FCV when the data set for deriving
a FCV does not contain three ACRs for freshwater species.  EPA invited comment
on the preference for freshwater ACRs in calculating a FCV to protect species
in the Great Lakes System.

      b.    Comments:   Some commenters preferred using saltwater ACRs only
when there were not-an adequate number of freshwater ACRs, while other
commenters wanted to use fresh and saltwater ACRs interchangeably, and others
suggested not using any saltwater ACRs.

       EPA agrees with those comments suggesting that saltwater ACRs should
only be used when there are an inadequate number of freshwater ACRs.  As
stated in the proposal, because the Great Lakes are freshwater lakes, the
preference is to use freshwater ACRs when deriving FCV for freshwater animals.
However, there may be situations where the freshwater data needed to derive a
freshwater ACR is limited.  In these situations, EPA believes it is
appropriate to use saltwater ACRs in place of the freshwater ACRs (see 1985
National Guidelines',  pgs.  14 and 15).

      c.    Final Guidance:  The final Guidance retains the provision that
saltwater ACRs may be used to derive a FCV when the data set does not contain
three ACRs for freshwater species as specified and if the freshwater and
saltwater ACRs are within a factor of 10 (as described in the aquatic life
Tier I criteria methodology, appendix A, VI.K.2 and the 1985 National
Guidelines).

6.    Bioavailability

      a.    Proposal:   The criteria for metals in Tables 1 and 2 of part 132
were expressed as total recoverable concentrations.  EPA invited comment on
whether the bioavailability of contaminants was adequately addressed using
site-specific modification approaches, as well as alternatives to address the
issue of expressing toxicity of both bioavailable and total contaminant
concentrations.

      Subsequent to the proposal, EPA issued a memorandum to all EPA Regional
Water Management Division Directors providing policy and guidance on the
interpretation and implementation of aquatic life criteria for the management
of metals  (The Office of Water Policy and Technical Guidance on Interpretation
and Implementation of Aquatic Life Metals Criteria (Prothro, 1993) .   The
memorandum covered a number'of areas including the expression of aquatic life
criteria, total maximum daily loads, permits, effluent monitoring, compliance,
and ambient monitoring.  With regard to expression of aquatic life criteria,
the memorandum recommended that State water quality standards be based on
dissolved metals because dissolved metal concentrations more closely
approximate the bioavailable fraction of metal in the water column than does
total recoverable metal concentrations.  However, because the present National
aquatic life criteria were developed using total recoverable measurements, it
is necessary to use, a conversion factor to convert the total recoverable metal
concentrations to equivalent dissolved metal concentrations.  The Prothro
(1993)  memorandum suggested conversion factors for 10 metals.

      Given EPA1s policy and guidance and the comments received on this issue,
EPA made available data to convert the proposed total recoverable metals
criteria to the comparable dissolved metal criteria on August 30, 1994 (59 FR
44678) .   EPA invited comment on the procedure used to derive conversion
factors for the nine proposed metals criteria in the "Draft Report:   Results
of Simulation Tests Concerning the Percent Dissolved Metal in Freshwater
Toxicity Tests  (Stephan, 8/1994) .  EPA also invited comment on the data
collected utilizing-the procedures as well as1the calculated conversion
factors for seven of the nine proposed metals.  However, this draft report did

-------
98     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

not contain data or conversion factors for cadmium or mercury(II) because the
testing vas not yet complete for those metals at the time of the notice.

      b.    Comments:  Numerous commenters stated that the proposed criteria
for metals overestimate the bioavailability of metals and that recent
technical and policy developments of EPA suggest the criteria are
overprotective.  These same commenters urged EPA to modify the proposed
Guidance to base the numeric metals criteria on the dissolved metal fraction
instead of the total recoverable metals approach used in the proposed
Guidance.  Some commenters, however, stated a preference for expressing the
criteria as total recoverable concentrations.  Many commenters asked EPA to
maintain the flexibility for the State or Tribe to choose the form of
pollutant used to develop criteria.
      Comments received from the August 30, 1994, Federal Register notice
generally favored using dissolved metals criteria for the final Guidance.
Many commenters agreed that repeating all of the toxicity tests to generate
new dissolved metals criteria was too resource-intensive and that the
simulation tests contained in the report were technically preferable.   Some
commenters also stated that dissolved metals criteria may allow potential
increases in metals loadings and potential impairment to the aquatic ecosystem
by ignoring fate and transport of particulate bound metals.

      Some commenters suggested that, except for chromium(III) and lead, EPA
should treat the current total recoverable metal criteria as dissolved
concentrations  (i.e., make the conversion factor 1.0).  Many commenters
requested the opportunity to comment on any new data used to derive the
conversion factors for cadmium and mercury(II).

      Although the draft report (Stephan, 1994) did not discuss implementation
of the dissolved metals criteria,  EPA received considerable comment regarding
implementation.  Generally commenters felt that it was important for EPA to
specify in the final Guidance how the dissolved criteria should be
implemented.  Commenters cautioned EPA against adoption of the dissolved
criteria without guidance on Water Effect Ratios  (WERs), modelling,
translation for limits, clean analytical and sampling methods and use of
appropriate historical data.  Some commenters suggested that EPA adopt  the
Prothro  (1993) memorandum into the final Guidance and others suggested  that
existing guidance was insufficient and EPA should update and/or revise
existing implementation guidance.  Comments were received regarding use of the
acidification method in place of the total recoverable method in the
simulation tests and the implications to analytical testing for compliance
purposes.

      EPA agrees with comments that, in general, the dissolved metal fraction
more closely approximates the bioavailable fraction of metal in the water
column than does total recoverable metal.  Aquatic life criteria are designed
to protect aquatic organisms from water column toxicity.  The primary
mechanism for water column toxicity is adsorption at the gill surface which
requires metals to be in the dissolved form.  The use of the dissolved  form of
the metal will, therefore, better approximate the toxicity to the aquatic
organism.

      This does not suggest, however, that the expression of metals criteria
as total recoverable is not scientifically defensible, nor does  this imply
that State and Tribes are required to adopt  the dissolved metals criteria.
EPA agrees with commenters that States and Tribes should be allowed to  choose
the form of pollutant for which to develop criteria.  EPA, while stating a
preference for dissolved metals criteria in  the final Guidance,  realizes that
there may be situations, such as  consideration of sediments or food chain
effects, where a State  or Tribe believes the expression of metals criteria as
total recoverable is preferable.  EPA will allow  the States and  Tribes  the
flexibility to adopt total recoverable criteria for metals as stated in the
Prothro  (1993) memorandum.

-------
                               Section ID: Aquatic Life                             99
      EPA recognizes that many of tne conversion factors published in the
draft report are very close to 1.0.  EPA chose to create experimentally
derived conversion factors rather than to assume that the existing total
recoverable criteria to be 100 percent dissolved.  EPA believes that this
approach is more technically sound.

      EPA recommends use of "The Office of Water Policy and Technical Guidance
on Interpretation and Implementation of Aquatic Life Metals Criteria"  (U.S.
EPA, 1993)  and the report titled "Interim Guidance on Determination and Use of
Water Effect Ratios for Metals" (U.S.EPA, 1994) for guidance on implementation
of dissolved metals criteria.   U.S.EPA (1993) contains guidance on dynamic
modelling and translators (Attachment #3),  and monitoring (Attachment #4).
U.S. EPA (1994) presents an effluent-specific approach for calculating a total
recoverable permit limit from a dissolved criterion (see pages 116 and 128-130
of U.S. EPA, 1994).  This approach is based on the percent of the total
recoverable metal in the effluent experimentally determined as described on
pages 112 and 125 (U.S. EPA, 1994).  A similar approach can be used to
calculate a permit limit for criterion expressed as free cyanide; in this case
the calculation is based on the percent of the total cyanide in the effluent
that becomes free cyanide in the downstream water.  EPA will continue to
update implementation guidance as needed in the future.

      For purposes of this rulemaking, Attachment #2 of U.S. EPA  (1993) is
superseded by the conversion factors and data in the report titled "Derivation
of Conversion Factors for the Calculation of Dissolved Freshwater Aquatic Life
Criteria for Metals" (U.S. EPA, 1995).  These conversion factors are also
found in Table III-l.

      EPA used the acidification method in the simulation tests, in place of
the total recoverable method,  because the metals potentially bound within the
test water would have been released by acidifying the samples.  The
acidification method is generally more accurate because it produces smaller
coefficients of variation than the total recoverable method.  In whole
effluent, metals may form stronger bonds to other components of the effluent.
Hence, for measuring total metals in effluent the total recoverable method is
the appropriate method.  The steps of heating and digesting the sample in the
total recoverable analytical method were not deemed necessary for purposes of
the simulation tests.  Moreover, the fraction of metal in the dissolved phase
may be different between effluent and surface waters.   In order to account for
all the metal being discharged into surface waters, and then calculate the
fraction that is dissolved,  the permitting authority must know the total
amount of metal in effluent.  Use of the acidification method was appropriate
for the simulation tests, but this method is not being considered for
regulatory use.

      Due to resource constraints,  EPA did not complete additional testing
required to update the conversion factors for mercury given in earlier EPA
guidance (Prothro, 1993) .  "The Office of Water Policy and Technical Guidance
on Interpretation and Implementation of Aquatic Life Metals Criteria"
(Prothro, 1993) did not contain conversion factors for the CMC for selenium or
for the for selenium and mercury.   Prothro (1993) , however,  did contain a
conversion factor of 0.85 for the National mercury CMC.   EPA made available
data to convert the proposed total recoverable metals criteria to the
comparable dissolved metal criteria on August 30, 1994 (59 FR 44678) for
selenium and mercury.  The conversion factor for the CMC for mercury came from
Prothro  (1993)  and this value was proposed in the August 30, 1994, notice for
the mercury CCC.

      Conversion factors were derived for selenium, even though Prothro  (1993)
stated that it is not appropriate to adjust the CMC and CCC for selenium or
the CCC for mercury because these are bioaccumulative chemicals.  Regardless
of whether these metals bioaccumulate, the important consideration is the
exposure that relates to the effect on which the CMC or CCC is based.  The CMC
is based on acute toxicity and so a relevant consideration is the

-------
100    Water Quality Guidance for the Great Lakes System ~ Supplementary Information Document

bioavailability of the metal in the water column; therefore, a total
recoverable CMC may be converted to a dissolved CMC if an appropriate
conversion factor is used and if there are no unacceptable risk management
considerations.  For a CCC, the exposure can be from the water column and from
the food, with the food chain consisting of some organisms whose primary
exposure is to pollutants in the water column and other organisms whose
primary exposure is to pollutants in the sediment.  It appears that exposure
to the sediment contributes substantially to the concentration of mercury in
the food chain, but does not contribute substantially to the concentration  of
selenium in the food chain.  Therefore, it is as acceptable to convert  the
total recoverable CCC for selenium to a dissolved CCC as it is to convert,  for
example, the total recoverable CMC for copper to a dissolved CMC.  In
contrast, it is not acceptable to convert a total recoverable CCC for mercury
to a dissolved CCC if the CCC for mercury is based on mercury residues  in
aquatic organisms.  It can, however, be acceptable to convert a CCC for
mercury to a dissolved CCC if the CCC is based on the toxicity of mercury to
aquatic organisms because the important exposure is through the water column;
as before, a dissolved criterion may be derived if an appropriate conversion
factor is used and if there are no unacceptable risk management
considerations.  Although a conversion factor cannot be used to derive  a
dissolved CCC for mercury if the CCC is based on mercury residues in aquatic
organisms, this does not necessarily mean that losses due to fate and
transport processes, such as volatility, cannot be taken into account in the
derivation of a permit limit.

      EPA believes that use of the CMC conversion factor to convert the total
recoverable mercury* CMC and CCC is technically preferable to having no
conversion factor for the mercury(II) CCC or requiring the CCC for mercury  be
expressed as total recoverable.  EPA will make available any new information
regarding alternative conversion factors for the mercury(II) CCC.

      EPA has completed testing to determine acceptable conversion factors  for
cadmium.  Because these conversion factors are substantially different  from
those in the Prothro  (1993) guidance, EPA will request public comment on the
data used to derive these factors.  EPA intends to amend the final Guidance
for the aquatic life cadmium criteria after comment is received on the  data
for the cadmium conversion factors.

      One commenter suggested that EPA consider the relationship of the
dissolved concentration of lead to hardness.  In the simulation tests it was
found that the dissolved concentration consistently fluctuated with variations
in hardness.  Therefore, EPA has amended the conversion factors for lead to
account for this hardness relationship.  Although Tier I criteria are not
provided for lead, EPA has included a conversion factor for lead in the final
report  (Stephan, 1995) for computing dissolved Tier II values or Tier I
criteria.

-------
                               Section ffl: Aquatic Life
101
                                  Table  III-l
1
Metal
Arsenic (III)
Cadmium
Chromium (III)
Chromium (VI )
Copper
Mercury (II)
Nickel
Selenium (IV)
Zinc
Recommended Conversion Factors
CMC
1.000
0.850
0.316
0.982
0.960
0.850
0.998
0.922
0.978
CCC
1.000
0.850
0.860
0.962
0.960
0.850
0.997
0.922
0.986
These recommended conversion factors are given to three decimal places  and are
not rounded off because  they are intermediate values in the calculation of
dissolved criteria.'

-------
102    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


      c.    Final Guidance:  For the reasons stated above, the final Guidance
contains metals criteria expressed as dissolved concentrations and as total
recoverable concentrations.  EPA converted the proposed criteria for total
recoverable metals to their comparable dissolved criteria using the conversion
factors found in Table III-l.  The final conversion factors and supporting
data are contained in the Final Report: "Derivation of Conversion Factors for
the Calculation of Dissolved Freshwater Aquatic Life Criteria for Metals"
(U.S. EPA, 1995) .  The conversion factors for the final GLI mercury(II) CMC
and CCC are 0.85 from "The Office of Water Policy and Technical Guidance on
Interpretation and Implementation of Aquatic Life Metals Criteria" (Prothro,
1993) .   EPA has also included the conversion factors from the Prothro  (1993)
guidance for cadmium in the final Guidance.  EPA intends to amend the final
Guidance for the aquatic life cadmium criteria after comment is received on
the data for the cadmium conversion factors.  States and Tribes may adopt
criteria as total recoverable metals.  The total recoverable analytical
methods found in 40 CFR § 136 must be used for purposes of compliance
monitoring in NPDES permit limits.

7.    Averaging Period/Frequency of Exceedance

      a.    Proposal:  The proposed Guidance, consistent with the current
National guidance, required that, except possibly where a locally important
species is very sensitive, aquatic organisms and their uses should not be
affected unacceptably if the following conditions are met:  for chronic
criteria, the four-day average concentration of a chemical does not exceed the
CCC or Secondary Continuous Concentration  (SCO more than once every three
years on the average; for acute criteria, the one-hour average concentration
of a chemical does not exceed the CMC or Secondary Maximum Concentration  (SMC)
more than once every three years on the average.

      Averaging periods are time periods over which ambient concentrations are
to be averaged to determine whether criteria are exceeded.  If the mean
ambient concentration of a pollutant exceeds the criteria over the averaging
period, adverse impacts on the resident aquatic life could occur.  Averaging
periods are one means of accounting for the exposure time required to elicit
toxic effects.

      An allowable frequency for exceeding  the criteria is incorporated into
the  criteria because it is not necessary for concentrations to be below
criteria at all times in order to adequately protect aquatic ecosystems.
Also, it is not generally possible to ensure that criteria are never exceeded.
Frequently, concentrations above criteria may occur without corresponding
impacts on  the aquatic biota if the duration is less than the averaging
period.  This is dependent on the magnitude and duration of the exceedance.

      b.    Comments:  A few commenters questioned the validity of the one-
hour and four-day averaging periods, and the once-in-three-year frequency
stating that these components may be overly stringent and inappropriate for
chemicals  such  as bioaccumulative chemicals of concern.  Commenters stated
that a  four-day averaging  period  is  too short for bioaccumulative chemicals,

      EPA agrees that the  averaging periods proposed in the Tier I and II
methodologies will protect ecosystems  from  the effects of toxicants to which
they are  applied but that  other averaging periods may also be appropriate for
certain pollutants.  The one-hour averaging period proposed for the CMC and
SMC  and the four-day averaging period  or duration for the CCC and SCC  were
based on  fast acting toxicants.  Because different pollutants cause effects to
aquatic organisms at different rates,  these assumptions might be overly
protective for  some pollutants in some situations.

      EPA's rationale for  the once-in-three-year frequency proposed for acute
and  chronic criteria and values  is presented in the 1985 National Guidelines
and  is  also explained in  appendix D  of the  "Technical Support Document for

-------
                               Section HI: Aquatic Life                            103
Water Quality-Based Toxics Control" (U.S. EPA, 1991) which is included in the
docket for this rulemaking.

      c.    Final Guidance:  EPA believes that attainment of the one-hour or
four-day averaging periods and the once-in-three-year frequency will protect
ecosystems from adverse effects of toxicants to which they are applied.
Consistent with the available data, however, EPA agrees with commenters that
alternate averaging periods and frequencies may also be appropriate for
certain pollutants.  Consequently, the final Guidance allows the use of
alternative averaging periods and frequencies if States or Tribes can
demonstrate to EPA that they are scientifically defensible.

      In evaluating whether an alternative averaging period is scientifically
defensible, EPA will pay particular attention to those toxicity tests which
account for the time needed to elicit effects.  Greatest emphasis will be
placed on results for sensitive species  (e.g., those affected at relatively
low concentrations).  In assessing lethality from acute or chronic toxicity
tests, EPA will pay particular attention to those relationships (a) between
effect concentrations and time of exposure, or (b) between time to death and
concentration.  In assessing sublethal (e.g., growth and reproductive)
effects, EPA will consider data from growth or reproductive tests comparing
time-varying exposure with steady exposure.  In such tests, evidence would be
considered supportive of a longer-term averaging period where time-variable
and steady concentrations yield the same growth or reproductive effects.

      In considering alternative frequencies of exceedance, EPA recognizes
that a substantial excursion above a criterion is likely to affect more taxa
more severely than a marginal excursion,  and that the setting of the allowable
frequency for marginal excursions implicitly sets even lower frequencies for
more severe excursions.  Consequently, in evaluating whether an alternative is
scientifically defensible, EPA will consider assessments that either
explicitly or implicitly account for  (a)  the probable frequency of lethal
events for an assemblage of taxa covering a range of sensitivities to
pollutants; (b) the probable frequency of sublethal effects for such taxa;  (c)
the differing effects of lethal and sublethal events in reducing populations
of such taxa; and  (d) the time needed to replace organisms lost as a result of
toxicity.  For an assemblage of taxa for which toxicity data are available,
such an assessment should yield an overall measure of toxic impacts that would
result from the alternative excursion frequency being proposed.  One such
measure of toxic impacts could be the projected aggregate population deficit,
compared to a system unaffected by pollutant toxicity.

      Alternatively, EPA will consider information from well-designed field
biological surveys indicating the relationship between biological quality and
chemical criteria excursion frequency, provided that confounding factors such
as bioavailability are appropriately taken into account and providing that the
field surveys are adequately sensitive.

8.     Final Tier I Criteria

      a.    Proposal:  The proposal provided a detailed discussion of how EPA
selected the 16 pollutants for which Tier I criteria were calculated.  EPA
chose not to propose specific numeric criteria for ten pollutants  (aldrin,
aluminum, chlordane, chlorpyrifos, DDT, endosulfan, heptachlor, lead, PCBs,
and toxaphene) for which EPA has National aquatic life criteria for the
reasons stated in the proposal (See 58FR20852).   EPA requested comment on the
proposed alternative of requiring States and Tribes to adopt the current
National criteria for those 10 pollutants.

      EPA proposed acute criteria for 16 pollutants and chronic criteria for
15 pollutants  (See Tables 1 and 2 of proposed part 132) .  Footnotes within the
tables delineated which proposed criteria were hardness or pH dependent and
those criteria were listed at a hardness of 50 mg CaCO3/L  or a  pH  of  6.5.

-------
104    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


      b.    Comments:  Several coiranenters asked EPA to perform the testing
needed to develop, or to develop, Tier I criteria for the ten remaining
pollutants.  Numerous commenters argued against using National Criteria where
the data set for the pollutant did not meet the proposed Tier I aquatic life
data requirements.  Some commenters favored using Tier II methods for these
pollutants.  One commenter thought that EPA was obligated to require adoption
of National aquatic' life criteria  (where Tier I criteria do not exist) by
States and Tribes to meet the "no less restrictive" clause of the Critical
Programs Act.

      EPA disagrees that it should be solely responsible for developing the
additional data to derive Tier I criteria for the 10 pollutants with National
criteria.  As discussed in section II.C. of this document, EPA in cooperation
with States and Tribes will establish a Clearinghouse to assist States and
Tribes in developing numeric Tier I water quality criteria and Tier II water
quality values.

      EPA agrees with commenters that States and Tribes should not be required
to adopt the current National criteria for the 10 pollutants that do not meet
the minimum data requirements for a Tier I criteria.

      In the preamble to the proposed Guidance, EPA explained that the
proposed Great Lakes criteria for aquatic life would not be less restrictive
than the National criteria because States and Tribes would be required  to
generate Tier I criteria or Tier II values to regulate these 10 pollutants.
EPA's position has not changed even though, as explained below, the final
Guidance provides States and Tribes the option of regulating pollutants
lacking a sufficient data set to derive Tier I criteria by either using Tier
II values or establishing limits based on an indicator parameter.  In the
National program, where a State or Tribe must regulate a pollutant that lacks
a promulgated numeric criterion, the State or Tribe may base a permit limit on
either a numeric value or an indicator parameter.  Consequently, EPA does not
consider its decision to allow States and Tribes to use indicator parameters
in place of Tier II values to make the final Guidance less restrictive  than
the National program for these 10 pollutants.

      EPA also notes that the final Guidance contains numeric Tier I criteria
for human health or. wildlife for chlordane, DDT, PCBs, and toxaphene that are
more stringent than the current National aquatic life criteria for the  same
pollutants.  These criteria will help ensure that the final Guidance will
protect aquatic life species as effectively as the National program for these
four pollutants.

      Several commenters requested that EPA provide the final acute and
chronic equations within the final rule.  EPA agrees that these equations
should be part of Tables 1 and 2.  EPA historically has provided criteria for
pollutants that are hardness dependent in the form of an equation that
accounts for hardness as part of the National criteria statement.

      c.    Final Guidance:  For the  reasons stated above, EPA has determined
that it is appropriate  to require States and Tribes to regulate these 10
pollutants by developing Tier II values.  The final Guidance includes the
acute and chronic hardness-dependent  criteria for cadmium, chromium(III),
copper, nickel, and  zinc.  The hardness and pH equations are found in Table
III-2 as well as each individual criteria document for the above chemicals.

      Table III-3 presents the CMCs calculated using the Tier I methodology
for aquatic life.  CMCs for metals are expressed as dissolved concentrations
and total recoverable concentrations.  Conversion factors utilized to convert
the metals criteria* from total recoverable  concentrations to dissolved
concentrations are found in Table  III-l.  For comparison the CMCs of existing
National criteria are also included.  Hardness-dependent CMCs for cadmium,
chromium(III), copper,  nickel, and zinc are found in Table III-2.  A pH-
dependent CMC for pentachlorophenol is also found in Table III-2.

-------
                               Section HI: Aquatic Life                           105
      A technical support document, Great Lakes Water Quality Initiative
Criteria Document for the Protection of Aquatic Life in Ambient Water  (EPA
820-B-95-004) (Final Criteria Documents), (available in the administrative
record to this rulemaking) presents the derivation of each of the final Tier I
CMCs and the toxicity studies from which the criteria were derived.  The final
Guidance requires that the numeric criteria in Table 1 to part 132 modified as
appropriate to reflect site-specific conditions  (or more stringent criteria)
be adopted by the Great Lakes States and Tribes and incorporated into their
ambient water quality standards.  The specific requirements on how these
criteria are to be incorporated into State and Tribal water quality standards
are discussed in section II  (Regulatory Requirements) of this document.

      Table III-4 presents final CCCs calculated using the final Tier I
methodology for aquatic life.  Final CCCs for metals are expressed as
dissolved concentrations and total recoverable concentrations.  Conversion
factors utilized to convert the metals criteria from total recoverable
concentrations to dissolved concentrations are found in Table III-l.  For
comparison the CCCs'in existing National criteria are also included.
Differences between proposed and final Great Lakes Tier I CMCs and CCCs are
described later in this section.

      The derivation of each of these final CCCs, and the toxicity studies
upon which they are based, are also discussed in Final Criteria Documents.
This document is available in the administrative record for this rulemaking.
Hardness-dependent CCCs for cadmium, chromium(III), copper, nickel, and zinc
are found in Table III-2.  A pH-dependent CCC for pentachlorophenol is also
found in Table III-2.  Criteria may be calculated at different concentrations
of hardness  (measured as mg/L of CaCO3  or pH.   For example,  using the
equations given in Table III-2, the total recoverable CMCs for zinc at a
hardness of 10,  50, 100, or 200 mg/L CaC03 are  17,  67,  120,  or 220  ptg/L
respectively.  The hardness equations used to calculate hardness dependent
freshwater criteria for metals  (i.e., cadmium,  chromium(III), copper, nickel,
and zinc)  can be used at any hardness.  Most of the data used to develop these
hardness formulas were in the range of 25 mg/L to 400 mg/L CaCO3  and the
formulas are therefore most accurate in this range.  Irrespective of this data
set, for waters with a hardness less than 25 mg/L CaCO3 criteria  should be
calculated using the actual ambient hardness of the surface water.  Limiting
use of the equation only to hardness above 25 could result in underprotective
criteria where the actual ambient hardness is below 25 mg/L CaC03.   The
majority of waters nationwide have a hardness of less that 400 mg/L CaCO3.   If
however, the hardness is over 400 mg/L CaCO3, two options  are  available for
the State or Tribe:  1)  use 400 mg/L CaC03 for  the criteria calculation or 2)
require an analysis to calculate a water-effect ratio for the site with
hardness above 400 mg/L CaCO3 and modify the  final criteria concentration
using the calculated water-effect ratio.  Use of a water-effect ratio in this
instance is recommended at a hardness above 400 mg/L CaCO3 because  other
confounding factors,  which may cause this hardness, can also affect the
toxicity of the metal.

      The final Guidance requires that the numeric criteria in Table 2 of part
132 (or more stringent criteria) be adopted by the Great Lakes States and
Tribes and incorporated into their ambient water quality standards.  The
specific requirements on how these criteria are to be incorporated into State
and Tribal water quality standards are discussed in section II of this
document.

-------
106   Water Quality Guidance for the Great Lakes System — Supplementary Information Document
                              Table III-2

                        Hardness and pH Equations'
Chemical
Cadmium
Chromium (III)
Copper
Nickel
Pentachlorophenol
Zinc
Final Acute Equation
gl-128 (tahMdncM) -3.6867
gO.819 (to budncM) + 3.7256
gO.9422 On huJnow) - 1.700
gO.846 On Imrinew) + 2.255
e 1. 005 (J>H)- 4.869
gO.8473 (In hMdnew) + 0.884
Final Chronic Equation
e0.7852 (ta budneu) - 2.715
pO-819 (In hudKM) -1-0.6848
gO.8545 (IntanJneM) - 1.702
gO.S4« (tn hardneu) -f 0.0584
gl.005 (pH) -5.134
gO.8473 (In hudncw) + 0.884
           These  equations  are  for  criteria expressed  as  total
           recoverable concentrations.• Conversion factors should be
           applied to the resulting value if  criteria expressed as
           dissolved concentrations are desired.

-------
                         Section HI: Aquatic Life
107
                            TABLE III-3
       Acute Ambient Water Quality Criteria for Aquatic Life
Chemical
Arsenic (III)
Cadmium b
Chromium (II I) b
Chromium (VI )
Copper b
Cyanide , free
Dieldrin
Endrin
Lindane
Mercury (II)
Nickel b
Parathion
Pentachlorophenol °
Selenium
Zinc b
Great Lakes
Final CMC*
(dissolved)
340
1.8
320
16
7.0
n/a
n/a
n/a
n/a
1.4
260
n/a
n/a
18
65
Great Lakes
CMCs"
(total)
339.8
2.1
1000
16.02
7.3
22
0.24
0.086
0.95
1.694
260
0.065
5.3
19.34
67
National
CMC"
360
1.8
980
16
9.2
22
2.5 «
0.18 "
2.0 d
2.4
790
0.065
5.5
20
65
      All values are in /ig/L.

      The toxicity of this chemical is hardness-related;  the criterion
      expressed is at a hardness of 50 mg/L.

      The criterion for this chemical is pH dependent;  the criterion
      expressed is at a pH of 6.5.

      This value is an FAV that was calculated according  to the 1980
      guidelines.  Although the CMC = FAV/2 in the  1985 National
      Guidelines, there is no CMC in the 1980 guidelines  and the
      procedure used to derive the FAV is different from  that used in
      the 1985 National Guidelines.
Note: The term "n/a" means not applicable.

-------
108
Water Quality Guidance for the Great Lakes System — Supplementary Information Document
                                   Table III-4
            Chronic Ambient Water Quality Criteria for  Aquatic Life
Chemical
Arsenic (III)
Cadmium b
Chromium (II I) "
Chromium (VI)
Copper b
Cyanide, free
Dieldrin
Endrin
Mercury (II)
Nickel b
Parathion
Pentachlorophenol c
Selenium
Zinc b
Great Lakes
Final CCC"
(dissolved)
150
1.2
42
11
5.0
n/a
n/a
n/a
0.77
29
n/a
n/a
4.6
66
Great Lakes
CMCs*
(total)
147.9
1.4
49
10.98
5.2
5.2
0.056
0.036
0.9081
29
0.013
4.05
5
67
National
CCC*
190
0.66
120
11
6.5
5.2
0.0019 d
0.0023 d
0.012 d
88
0.013
3.5
5.0
59
      '     All values in

      b     The toxicity of this chemical is hardness-related;  the criterion
            expressed is at a hardness  of 50 mg/L.

      c     The toxj.city of this chemical is pH related; the  criterion
            expressed is at a pH of  6.5.

      d     Based upon Final Residue Value.

      Note:   The term "n/a" means  not applicable.

-------
                               Section ffl: Aquatic Life                           109
9 .     Tier I Criteria

      a.    Proposal:   The proposal contained CMCs  (i.e., acute criteria) for
sixteen pollutants calculated using the proposed Tier I methodology for
aquatic life.  CMCs were proposed for arsenic(II), cadmium, chromium(III),
chromium(VI), copper,  cyanide, dieldrin, endrin, lindane, mercury, nickel,
parathion, pentachlorophenol,  phenol, selenium, and zinc.  CCCs  (i.e., chronic
criteria)  were proposed for fifteen of the sixteen chemicals listed above.
EPA did not propose a CCC for lindane because the minimum data requirements
were not met.

      b.    Comments:   Several commenters noted the lack of information or
lack of explanation in the proposed aquatic life criteria documents.  For
example, it was pointed out that EPA did not present the ACRs in the
arsenic(III) document.  In addition, commenters noted several small errors
within the documents,  such as mathematical calculation errors.

      Several commenters questioned the data used in the derivation of
criteria.   For example, for mercury EPA chose a 96-hour I,CX for Cranqonvx
which was two to three orders of magnitude lower than the 48-hour IiCK.
Several commenters indicated that some criteria derived were unrealistically
stringent, but generally did not present data to support this opinion.

      c.    Final Guidance:  EPA agrees with comments that the criteria
documents could have been more descriptive in the derivation of intermediate
values such as the FACR and in how the final CMCs and CCCs were calculated.
In response to these comments EPA has attempted to better explain how the
criteria were derived.  EPA notes that the most recent National aquatic life
criteria document and the final GLI document together contain the complete
data set for the chemical.  In addition, the final criteria for endrin,
cadmium, mercury, arsenic(III), lindane, nickel, zinc, copper, dieldrin, and
parathion were changed from the proposal for the reasons cited below.  The
proposed criteria concentrations for chromium(III), chromium(VI), cyanide, and
selenium are the same as the proposal.  Some explanatory language was added to
all the documents to better explain how the criteria were derived and the
documents were slightly reformatted.

      Some data were removed in response to comments.  Chemicals for which
data were deleted are cadmium, endrin, and mercury.  The proposed endrin
document stated that some acute toxicity data from the EPA endrin criteria
document were not used in development of the GLI criteria because the test
protocols did not meet current acceptable toxicity testing procedures.  Some
of these data, however, were not deleted and was used to calculate the
proposed endrin criteria.  EPA has deleted this data from the criteria
derivation for the final GLI endrin criteria.  This resulted in a slight
change from the proposed criteria.

      In reviewing the cadmium document EPA ascertained that the
concentrations of cadmium were not measured in a chronic test with Moina.
This data was deleted along with several acute and chronic tests conducted in
river water (Spehar'and Carlson, 1984).  The Spehar and Carlson  (1984) tests
were deleted because variables within the water which might affect the
toxicity of cadmium are unknown.  These changes resulted in raising the CCC by
nearly a factor of two.  In the proposed cadmium document, the range of
Species Mean Acute Value  (SMAVs) was greater than a factor of five for the
genus Morone.  Because of this wide range, EPA set the Genus Mean Acute Value
(GMAV) for Morone equal to the lowest SMAV for that genus.

      In the final cadmium document, EPA also set that the GMAV for Daphnia
equal to the lowest SMAV because the range of SMAVs was greater than a factor
of five.  EPA also has made the Genus Mean Chronic Value (GMCV) for Daphnia
equal to the lowest SMCV for this genus.  These changes also contributed to
the1 the higher CCC for cadmium.  The CMC did not change as a result of these
corrections.  EPA believes the approach used for selecting the GMAV is more

-------
110    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

appropriate than the approach used for selecting the GMCV when there  is a wide
range in the SMAVs or SMCVs.

      In the mercury document, the 96-hour LCjo for Cranqonvx was deleted and
the 48-hour LCy, was not used because  these values were from the same  test.
EPA determined that the concentration measured was too great a change from 48
to 96 hours, making, the test or measurements suspect.  This change resulted in
an increase in both the CMC and CCC.

      Some of the data which was listed in the references and used to derive
the criteria did not appear in the data tables.  For example, the ACRs used in
the derivation of the arsenic(III) CCC did not appear in Table 3 as
referenced.  EPA did include all of this data in the references and most of
the information was located in the National arsenic criteria document.  EPA
has amended the final arsenic(III) GLI criteria document to include all data
referenced.  References and data within the tables were examined and  similar
corrections were made as necessary to the documents for arsenic(III), copper,
dieldrin, and endrin.

      Minor mathematical and rounding errors were found in the proposed
criteria arsenic(III), cadmium, endrin, lindane  (acute), nickel, and  zinc.
Simple mathematical errors in addition, subtraction, multiplication and
division were corrected where found.  In addition, all the documents  were
checked to ensure that intermediate values were not rounded.  The criteria for
lindane and nickel did not change as  a result of these corrections.

      A new FACR was calculated for zinc.  The proposed CCC was derived with a
FACR which contained a saltwater ACR.  Upon additional review of the  data, EPA
determined that this saltwater ACR was not needed.  The FACR for zinc was
recalculated based only on freshwater ACRs.  This changed the CCC some what
and made it equal to the CMC.

      EPA found similar data trends  (wide ranges in SMAVs) in the copper,
dieldrin, and parathion data sets.  EPA consistently applied the procedure of
making the GMAV or GMCV equal to the  lowest SMAV or SMCV  (respectively) when
the range of SMAVs or SMCVs is greater than five.  EPA believes  that  this is
consistent with  section XI.B. of appendix A to part 132 which asserts that
appropriate modifications of the methodology shall be used consistent with
sound scientific evidence where warranted.  EPA believes that this
modification is warranted to ensure protection to the more sensitive  species
within a genus.  In this situation, the mean of the data could  result in
underprotection  for these species if  the genus is very sensitive.  EPA,
therefore,  consistently applied this  procedure.  Applying this procedure to
copper, dieldrin, and parathion did not result in changes to the proposed
criterion concentrations.

      After the Tier I criteria for phenol were proposed, EPA reevaluated the
data used to derive the criteria.  EPA has determined that at this time there
is insufficient  data of adequate quality to finalize the CCC for phenol.  EPA
also realized that not all  of the data which could be used to derive  the CMC
and CCC for phenol were made available for public comment at the time of the
proposal.   Because of the withdrawal  of some of data used to derive the
proposed criteria and the  inability for public comment on other data  available
prior to proposal, EPA will not finalize the proposed criteria  for phenol.
EPA will review  the- additional data which was not used in the proposal  and
make it available through  the clearing house once established.

10.   Potential  Changes to  National Guidelines

      a.    Proposal:  In  the proposal, EPA noted that the 1985 National
Guidelines  were  being revised.  To  date, no revisions have been proposed  for
the  1985 National Guidelines.

-------
                               Section ffl: Aquatic Life                            111
      b.    Comments:   Several commenters asked for clarification about what
will happen when the 1985 National Guidelines are revised.

      c.    Final Guidance:  Due to current resource constraints on EPA,
revisions to the 1985 National Guidelines are not anticipated in the near
future.  When revised National Guidelines are proposed, EPA will announce
whether the revisions should replace or supersede any portion of the final
Guidance.  Until such time, Great Lakes States and Tribes are required to use
the final Tier I methodology herein.

C.    Final Tier II Methodolocrv

      The proposed Guidance required the use of a Tier II value where a State
or Tribe lacked the-full eight families of toxicity data needed to set a
chemical-specific Tier I criterion for a pollutant.  As proposed, the Tier II
methodology would allow derivation of a chemical-specific value using toxicity
data for as few as one species.  It contained adjustment factors to compensate
for the missing data.   EPA proposed to require use of methodologies consistent
with the Tier II methodology to interpret narrative criteria (such as the
prohibition on the discharge of no toxic chemicals in toxic amounts) in all
cases where a State or Tribe determined using proposed procedure 5 of appendix
F to part 132, that an effluent discharging into the Great Lakes system
contained a pollutant in an amount that causes, has the reasonable potential
to cause, or contributes to an excursion above any water quality standard,
including the narrative criterion for water quality.  EPA also proposed to
require the use of methodologies consistent with the Tier II values as the
applicable water quality standard for use in "reasonable potential"
determinations under procedure 5 of appendix F to part 132 where there is no
Tier I criterion nor sufficient data to calculate Tier I criteria.

1.    Requirement for use in Interpreting the Narrative Toxics Criterion

      Under the current National permitting requirements, where a State or
Tribe determines that the discharge of a pollutant causes, has reasonable
potential to cause,, or contributes to an excursion above a State's or Tribe's
narrative criterion, the State or Tribe must, in the absence of a numeric
criterion for that pollutant, interpret the narrative criterion and include
water quality-based effluent limitations (WQBELs)  in the permit that are
derived from and comply with the narrative as interpreted.  Depending on the
case-specific circumstances, such limits may currently be required for the
specific pollutant of concern and/or for whole effluent toxicity  (WET)  (see 40
CFR section 122.44(d)(1)(vi)).

      a.    Proposal:   The Tier II methodology developed for the proposed
Guidance provided a consistent means for interpreting narrative criteria that
protect aquatic life for all toxic pollutants with a very small number of
data.  EPA proposed to require States and Tribes to use the Tier II
methodology to interpret their narrative criteria and where necessary to
establish WQBELs.  Under procedure 6 of appendix F to part 132, EPA also
proposed the specific circumstances under which WET limits would be required
in a permit.  Both EPA and the Initiative Committees thought that requiring
permit limits based on Tier II values would increase consistency among Great
Lakes States and Tribes.  The Steering Committee also hoped that the
relatively stringent Tier II values would motivate some permittees to develop
the toxicity data needed to derive a Tier I criteria.

      At the same time,  EPA recognized that a facility with a WET limit could
argue that it had already reduced each of the individual pollutants in that
effluent sufficiently to protect the tested aquatic species in the receiving
water body (which often would be identical to the single species required as
the minimum data for a Tier II value),  and that further reductions in the
amounts of individual  pollutants based on a chemical-specific Tier II goal for
the water body were unnecessary.  Consequently, EPA asked for comment on the

-------
112    Water Quality Guidance for the Great Lakes System -- Supplementary Information Document


need for requiring both permit limits based on Tier II values and permit
limits for WET, as well as other options for harmonizing the two requirements.

      b.    Comments:  Many commenters suggested that the Whole Effluent
Toxicity (WET) criteria be used in place of the aquatic life Tier II values
and furthermore, that the Tier II methodology be published as merely guidance
to the States and Tribes.  Some commenters suggested that the Tier II values
and WET protocols overlap substantially.  These commenters recommended  that
both the Tier II methodology and guidance require that WET protocols be
utilized to interpret the narrative criteria, in lieu of the proposed Tier II
values.  Some commenters thought that values derived using the Tier II  method
would actually create less consistency among States and Tribes.  Other
commenters preferred the proposed option of requiring use of both the Tier II
values and WET.  ,

      EPA disagrees that WET be used in place of the Tier II aquatic life
methodology and that the Tier II methodology be published as guidance.  EPA
notes that the Tier II aquatic life methodology offers some significant
practical benefits for both the regulated community and the States and  Tribes.
For the regulated community, the chemical-specific Tier II approach offers the
advantage of allowing the permittee to focus immediately on a single
contaminant for the purposes of designing effluent treatment.  In contrast,
WET often leads to a facility conducting fairly extensive investigations to
identify the cause of adverse effects on the tested organisms and to develop
an effective approach to reducing the effects.

      EPA notes that an individual discharger will not always need to have
both Tier II and WET limits in its permit in order to protect the narrative
water quality standards.  However, EPA maintains as it did in its rulemaking
promulgated on June 2, 1989 at 40 CFR 122.44(d)(1)(vi)(C), that once a  finding
is made that the discharge of a pollutant causes, has the reasonable potential
to cause, or contributes to the excursion above the narrative criterion,
reliance on WET alone, in lieu of chemical-specific limits must only be done
where the discharger can demonstrate that WET sufficiently guards against
excursions above the applicable water quality standard  (i.e., the Tier  II
interpretation of the narrative standard).  EPA continues to believe that the
use of the Tier II methodology ie important for deriving Tier II values to
determine whether a pollutant has the reasonable potential to exceed a  Water
Quality-Based Effluent Limit  (WQBEL).

      c.    Final Guidance:  The final Guidance specifies the Tier II
methodology as the methodology States and Tribes will have to use in
interpreting their narrative water quality criteria for the protection  of
aquatic life.  Procedures 5 and 6 of appendix F to part 132 specify the
conditions under which chemical-specific and/or WET limits will be required
for an individual discharger.  Both kinds of limits will not always be
necessary or required for a discharger.  As described in section VIII.E of
this document, and consistent with 40 CFR 122.44(d)(1)(vi)(C), limits on
indicator parameters may be used in lieu of limits on the pollutant of  concern
when implementing the narrative criteria, including, a WET limit in lieu of a
chemical numeric limit.  However, when an indicator parameter is used,  the
State or Tribe must ensure that the indicator parameter will attain the Tier
II interpretation of the "applicable water quality standard"  (as described in
40 CFR 122.44(d)(1)(vi)(C)).

      It should be noted that a WET limit can not be used in lieu of a
chemical-specific limit where a total maximum daily load  (TMDL), developed
pursuant to 40 CFR 130.7 and approved by EPA, and corresponding wasteload
allocation specifies the acceptable level for that chemical in a discharger's
effluent.  A State, Tribe  or permittee in the Great Lakes System that wished
to base permit limits on a TMDL, would need to  generate a Tier I criterion or
a Tier II value  that could then be apportioned  among contributing sources(40
CFR 122.44(d)(1)(vii)(B) ).

-------
                               Section HI: Aquatic Life                            113
      Finally, it should be noted that Tier II values derived using the Tier
II methodology are the appropriate means of determining whether a pollutant is
or may be discharged into the Great Lakes System at a level which causes, has
the reasonable potential to cause, or contributes to an excursion above a
narrative criterion.  WET testing is not an appropriate substitute for this
chemical-specific determination.  EPA does not believe that the Tier II method
by itself would cause less consistency among the Great Lakes States and
Tribes.  Currently, States have different mechanisms to translate narrative
criteria.  EPA believes that the Tier II methodology enables States and Tribes
to translate their narrative criteria with the same mechanism.

2.    Data Requirements

      a.    Proposal:  The proposed Tier II methodology utilized all available
data and provided for derivation of a Tier II value when data sufficient to
derive a Tier I criterion are not available.

      b.    Comments:  Several commenters were concerned that the data
requirements for the Tier II methodology were not adequate.  Some commenters
recommended that EPA require a minimum of three to seven of the eight
taxonomic families specified.  A few commenters further suggested that EPA, at
a minimum, require data for a daphnid, rainbow trout and fathead minnow.

      c.    Final Guidance:  EPA believes that the Tier II minimum data
requirements are sufficient given the purpose of the Tier II methodology.  As
described in the proposal (58FR20835), the Initiative Committees sought to
ensure consistency among States in the Great Lakes System as to how limited
toxicity data are used to interpret narrative standards.   The proposed aquatic
life Tier II methodology fulfilled this goal.  Section 303(c)(2)(B) of the CWA
specifies that States and Tribes shall adopt criteria for all toxic pollutants
for which presence in the affected waters could reasonably be expected to
interfere with designated uses adopted by the State or Tribe.  The minimum
data required for the Tier II methodology promotes consistency in how this
requirement is implemented.  The proposal provided a standardized process for
utilizing available data to derive values for purposes of interpreting
narrative standards, thereby achieving greater consistency among the States
and Tribes in this activity.

3.    Other Methods.for Tier II Values

      a.    Proposal:  EPA's Science Advisory Board (SAB), in its report,
"Evaluation of the Guidance for the Great Lakes Water Quality Initiative,"
suggested the use of short-term and short-cut chronic toxicity tests to derive
a Tier II value.  The SAB believed this could overcome the cost of completing
standard chronic toxicity tests.  EPA invited comment on the appropriateness
of short-term and short-cut chronic tests for the derivation of Tier II
values.

      b.    Comments:  Many commenters did not feel short-term or short-cut
chronic tests were appropriate for criteria or value derivation.  Other
commenters supported the use of short-cut chronic tests,  but did not provide
any examples of acceptable short-cut tests.  One commenter did recommend a
short-term chronic test (7-day fathead minnow test, EPA/600/4-89/001).

      At this time, EPA does not consider short-cut chronic tests appropriate
for the derivation of a Tier I criterion or Tier II values.  EPA has serious
concerns over whether short-cut chronic tests can accurately predict effects
from long-term exposures.   A short-cut chronic test is a short-term exposure
to an aquatic organism during a sensitive life stage(s) which might yield
toxicity results similar to those obtained from complete life-cycle, partial
life-cycle, or early life-stage tests.  Although this type of test might yield
results that are similar to those from a complete chronic test for some
chemicals, significant exceptions can occur.  No commenter pointed to any
studies in which a short-cut chronic method was compared to a complete life-

-------
114    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


cycle toxicity test for the same species.  Without such validation or
verification of a short-cut chronic test, EPA is hesitant to allow its  use as
a replacement for a life-cycle, partial life-cycle or early life-stage  test.

      EPA does, however, think that the 7-day Ceriodaphnid test, which
measures survival, growth and reproduction,  is acceptable for use in deriving
Tier I criteria and Tier II values, because  this is a life-cycle test
according to the American Society for Testing and Materials (See ASTM
Definition in the Administrative Record).  EPA does not consider this test a
short-cut test because it is a life-cycle test.  The 7-day duration is
appropriate for a life-cycle test with species in the genus Ceriodaphnia,
whereas a 21-day duration is appropriate for species in the genus Daohnia.

      c.    Final Guidance:  The final Guidance does not allow data from
short-cut tests to be used in deriving a Tier I criterion or Tier II value.
However, the final Guidance does allow use of the 7-day Ceriodaphnid life-
cycle test for derivation of both Tier I criteria and Tier II values.

4.    Adjustment Factors

      a.    Proposal:  The proposed Tier II methodology uses adjustment
factors obtained in the statistical analysis described by Host,  et al.  (1991)
in the draft paper, "Analysis of Acute and Chronic Data for Aquatic Life,"
(which may be found in the administrative record for this rulemaking) to
derive Tier II values from data for one to seven of the eight taxonomic
families required for Tier I calculations.   Depending upon the number of Tier
I minimum data requirements satisfied by the data base, different adjustment
factors are applied to the lowest Genus Mean Acute Value to arrive at the
Secondary Acute Value (SAV).  These adjustment factors are intended to  relate
the results of one to seven toxicity tests to a FAV.  EPA invited comment on
the selection of an 80th percentile in establishing the proposed adjustment
factors.  EPA also invited comment on the use of a set of lower  adjustment
factors to be used only where daphnid data were available as opposed to the
higher adjustment factors that would be necessary if data for the specified
daphnids are not required.

      b.    Comments:  Several commenters also suggested that EPA set up a
hierarchy of toxicity testing requirements for Tier II.  This would establish
the order in which data are added to the Tier II data set.  Several of  these
commenters suggested that EPA establish the  hierarchy where more sensitive
species would be tested first.

      It would be very difficult to set up a hierarchy of toxicity testing
requirements for Tier II because there is a  range of sensitivities among
species that satisfy any one minimum data requirement and because the species
sensitivity varies with the pollutant under  consideration.  EPA  believes it
would be speculative to create such a hierarchy and impose it on diverse
pollutants which elicit varying toxic responses.

      Many commenters stated that the 80th percentile was too conservative and
recommended that EPA establish median or 50th percentile adjustment factors.
Some commenters  supported  the use of  the 80th percentile adjustment factors.
A few commenters recommended that the level  of protection for adjustment
factors be set at the 95th percentile to provide the same level  of protection
as a Tier I criterion.  Many of the comments supported the use of adjustment
factors with daphnid data  rather  than those  without.

      c.    Final Guidance;  When deriving Tier II values for chemicals for
which limited  data are  available  for  establishing permit limits, EPA desires
to be highly certain that  adequate protection is afforded.  EPA  must be highly
certain that Tier II values provide a level  of protection greater than  or
equal to a Tier  I criterion.  EPA believes  that use of a procedure that would
result  in less protection  than use of the  80th percentile would  cause Tier II
values  to be significantly less protective  than Tier I criteria. Use of the

-------
                               Section HI: Aquatic Life                           115
50th percentile would allow less protection than a Tier I criterion in 50
percent of the cases.  EPA does not believe that 50th percentile factors would
ensure that adequate protection is afforded in most cases.  If the Tier II
value is thought to be unnecessarily low, the State, Tribe, or discharger can
generate the data necessary to allow derivation of a different (usually
higher)  Tier II acute value or a Tier I FAV.

      Possible factors were calculated in a number of ways in the draft report
on which the proposed adjustment factors were based.  The proposed factors
were calculated as "overall 80th percentiles," i.e., the percentiles were
calculated as percent of the simulations rather than as percent of the
chemicals.  After further consideration, EPA has decided that the most
appropriate adjustment factors to use are those that were calculated as
"medians of the 95th percentiles" on page 62 of the draft report because these
percentiles were calculated as percent of the chemicals (Host, et al., 1991) .
These final adjustment factors (with daphnid required) are given in Table III-
5.  In addition, some minor anomalies occur in the overall 80th percentile
values,  but no such anomaly occurs in the medians of the 95th percentiles.
Further, the calculation of the assumed ACR is based on the percent of the
chemicals.

      It might seem that the adjustment factors and the assumed ACR should be
based on the same percentile.  Because the assumed ACR will often be used with
a Tier II acute value, it does not seem reasonable to increase the level of
protection by having both be at the 95th percentile; in this case it might be
appropriate to have the ACR at a lower percentile.  In some cases, however, an
assumed ACR will be used with a Tier I FAV; in these cases it might be
appropriate to have the assumed ACR be the 95th percentile.  It does not seem
prudent to use two different values for the assumed ACR, and so EPA has
retained the value of 18, which is based on the 80th percentile (see the next
section).  The final Guidance also retains the use of adjustment factors "with
daphnid data."

-------
116    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


                                    Table III-5

                 Adjustment Factors (with Daphnid Data Required)


                             Sample Size
                     (Number  of  Data Requirements Fulfilled)

Percentile     1234567


   95          21.9  •  13.0    8.0   7.0    6.1     5.2     4.3

-------
                               Section HI: Aquatic Life                           117
5.    Assumed ACRs

      a.    Proposal:  In the proposal, EPA requested comment on the use of
assumed ACRs in place of experimentally derived ACRs, and particularly on the
use of 18 as the assumed ACR.  When there are less than three experimentally
derived ACRs, the proposal required the use of enough assumed ACRs of 18
 (along.with experimentally derived ACRs) to bring the total number of measured
and assumed ACRs to three  (See appendix A.XIII. of proposed part 132).

      b.    Comments:  Several commenters supported the use of assumed ACRs
and the proposed assumed ACR of 18.  One of those commenters suggested that
the assumed ACR of 18 be used as a cap.  No comments were received which
questioned the use of assumed ACRs.  Commenters suggested using alternate
assumed ACRs developed by the European Centre for Ecotoxicology and
Toxicology.  These ACRs are median values generated for various chemical
classes  (e.g., pesticides, metal/organometals, etc.).

      EPA agrees with commenters that use of assumed ACRs are appropriate when
experimentally derived ACRs are not available.  EPA disagrees with that ACRs
should be developed using median values.  EPA does not believe it is
appropriate to base ACRs on median values whether they are generated across
chemical classes or for a single class of chemicals, because the median will
be too low for half of the chemicals.

      c.    Final Guidance:  The final Guidance retains the use of assumed
ACRs when insufficient experimentally derived ACRs exist.  A final assumed ACR
of 18, based on an 80th percentile shall be used when there are insufficient
measured ACRs.  EPA believes that it is inappropriate to alter experimentally
derived ACRs capping them at 18.  The final Guidance requires experimentally
derived ACRs when available, but assumed ACRs shall be used where data are
lacking.  Since the value of 18 is the 80th percentile, twenty percent of the
chemicals are expected to have larger ACRs.  The final Guidance does not
preclude a State or Tribe from choosing an assumed ACR which is more stringent
 (or greater than 18).

D.    Comparison with the CWA and EPA's National Guidance

      Section 118(c)(2)(A) of the CWA requires the water quality criteria for
the Great Lakes to be "no less restrictive" than the National criteria and
guidance.  As EPA explained in the proposal, it will not promulgate Tier I
criteria for several pollutants for which National criteria exist.  EPA
continues to believe that this decision does not make the Great Lakes criteria
"less restrictive" than National criteria.  Moreover, EPA notes that its
decision to change the adjustment factors used to compute Tier II values makes
the Tier II values more closely resemble Tier I criteria.  Tier II values will
now be as or more restrictive as Tier I criteria in 95 percent of all cases,
rather than 80 percent.

      In the proposal EPA also explained that the Great Lakes Tier I criteria
for four pollutants that appeared to be less conservative than the National
criteria for the same pollutants were not actually less restrictive.  EPA
continues to rely on that analysis.  Similarly, EPA continues to rely on the
proposal's conclusion that the Tier II methodology is not less stringent than
the more general narrative standards allowed under the National program.

      In response to comments on the proposal EPA has decided to express all
Tier I criteria for metals in terms of dissolved rather than total recoverable
concentrations.  This change results in criteria that are more appropriate
but,  arguably, less-stringent than the National criteria (which are expressed
as total recoverable concentrations).  As explained above in the discussion of
this issue, EPA has already begun to allow the use of dissolved values for
metals under the National program.  EPA intends to continue revising the
National program to be more consistent with the Great Lakes program on this
issue.

-------
118    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


E.    Conformance with the Great Lakes Water Quality Agreement

1.    Tier I Aquatic Life Criteria and Methodology

      The Great Lakes Water Quality Agreement  (GLWQA)sets out both general and
specific "objectives" for water quality in the Great Lakes.  The specific
objectives include a list of concentration levels for individual pollutants
set out in Annex 1 to the GLWQA.  The Annex indicates that some of these
levels were chosen to protect aquatic life.   (Other levels protect either
human health or wildlife.)  Some of the Annex 1 concentrations for the
protection of aquatic life, however, appear to be more conservative than the
corresponding final Tier I criteria for aquatic life.  EPA nevertheless
believes that the final criteria, as well as the methodology from which they
were derived, conform with the provisions and objectives of the GLWQA.

      In the first place, the language of the GLWQA itself shows that neither
the general not the specific objectives are legally binding, precise
standards.  Article III  (general objectives), Article IV  (specific objectives)
and Annex 1  (specific objectives) all describe their provisions as
"objectives," or goals, rather than requirements.  Further, both Article II
and Annex I use the permissive term "should" rather than the mandatory
"shall."  Article IV contains some provisions that appear more mandatory, but
they do not relate to the achievement of specific water quality requirements.
Moreover, it describes the Annex 1 numbers as  "desired" levels rather than
requirements.  Consequently, EPA thinks it is reasonable to believe that the
framers of the GLWQA did not intend the Annex 1 numbers to become enforceable,
mandatory requirements under U.S. or Canadian law.   (The U.S. Department of
State reached the same conclusion where it analyzed the GLWQA for the Office
of Management and Budget in  1978.)

      EPA also believes that the Great Lakes Critical Programs Acto of  1990
did not change the legal status of the GLWQA's objectives or provisions.
Section 118(c)(2)(A) of the CWA directs EPA to adopt guidance that  "conforms"
with the objectives of the GLWQA.  "Conform" means to "make similar" or "bring
into harmony," not to  "duplicate" or "follow precisely."  The very  same
sentence in  section 118 of the CWA shows that Congress knew how  to  require a
closer relationship-when it wanted to impose one:  it requires EPA's guidance
to be  "no less restrictive than the provisions of  [the CWA] and National water
quality criteria and guidance."   Clearly, Congress did not impose the same
standard with respect  to the general and specific objectives of  the GLWQA.

      EPA also finds it  significant that section 118(c)(2)(A) of the CWA
directs EPA to develop numeric  limits for the Great Lakes waters rather than
simply to incorporate  the  GLWQA numeric values.  EPA believes that  Congress
would have  been very explicit if  it had intended to deprive EPA  of  the
authority to exercise  its  own judgement on  the technical and scientific issues
involved.   Moreover, the  legislative history shows that Congress knew and
approved of the ongoing  work of  the Great Lakes Initiative Committees.  S.
Rep. 101-339.  101st Cong., 2d Sess. at 18  (June 27., 1990); 136  Cong. Rec.
S15616  (Oct. 17,  1990).   Since  the legislative history  so prominently
acknowledges the  committees' work, it is reasonable  to  assume that  Congress
expected EPA to develop  its  own  criteria.   Consequently, EPA does not believe
that "conformance" with  the  GLWQA requires  the numeric  criteria  proposed to be
 identical to or no less  restrictive  than the GLWQA objectives, including
 individual  Annex  1 values.   Rather,  EPA's  guidance as a whole needs to  further
 the objectives of the  GLWQA.  In EPA's judgement, the final Tier I  aquatic
 life criteria  are consistent with the general objective that Great  Lakes
waters  should  be  "free from  materials...that...will  produce conditions  that
 are toxic or harmful  to  ...aquatic  life."   Great Lakes  Water Quality Agreement
 of 1978, Art.  Ill, para,  (d).

       In  the preamble  to the proposed rule,  EPA stated  that  it would seek  to
 revise some of the  Specific  Objectives of  the Agreement.   EPA, however, has
 reconsidered this approach.   In light of  all the  strong evidence that  the

-------
                               Section ffl: Aquatic Life                           119
Specific Objectives were meant to be goals rather than requirements, EPA
believes revisions are unnecessary.  It does not intend to pursue them at this
time.

2.     Tier II Values and Methodology

      Tier II is a conservative methodology designed to establish
environmentally protective limits on the discharge of pollutants into the
Great Lakes System.  The methodology will regulate the discharge of certain
pollutants which, in certain Great Lakes States, may have been regulated by
narrative criteria rather than specific numeric criteria.  The Tier II
methodology is consistent with the general objective of the GLWQA cited above.
Moreover, it serves as a translator mechanism for that "narrative" objective.
The Tier II methodology will enhance protection of aquatic life in the Great
Lakes basin and will serve the GLWQA's purpose of promoting consistency in the
regulation of toxics in the Great Lakes basin.  Therefore, it also "conforms"
to the GLWQA.

-------
                          Section IV: Bioaccumulation Factors                       121
          IV.   BIOACCUMULATION FACTORS
A.    Summary of Final Rule

      The final Guidance is similar in substance to the Guidance proposed on
April 16, 1993, except for the following changes:

      - - Taking into account the freely dissolved  concentration of a chemical
in the derivation of bioaccumulation factors  (BAFs) for organic chemicals.

      -- Use of the equation,  baseline BCF  =  KOW in place  of the equation
originally proposed, when predicting a bioconcentration factor (BCF) from a
chemical's octanol-water partition coefficient  (KQW) for organic chemicals.

      -- Use of a model adapted from Gobas  (1993)  rather  than the model
discussed in the proposal to determine food-chain  multipliers (FCMs) for
organic chemicals,  which provides FCMs for  the  entire range of KOWS,  rather
than defaulting to one for log KQWS greater than six.

      -- Addition of an option for predicting a BAF based on the biota-
sediment accumulation factor (BSAF)  methodology as the second most preferred
method in the hierarchy for derivation of BAFs  for organic chemicals.

      -- The use of standard lipid values of  3.10  percent in edible tissue of
trophic level 4 fish and 1.82  percent in edible tissue for trophic level 3
fish for use in determining human health BAFs for  organic chemicals in place
of the lipid value originally proposed.   The  use of standard lipid values of
10.31 percent in whole body of trophic level  4  fish and 6.46 percent in whole
body for trophic level 3 fish for use in determining wildlife BAFs for organic
chemicals in place of the value originally  proposed.

      -- BAFs for individual PCB congeners, weighted by their relative
concentrations in salmonids, which is the predominant route of exposure,
rather than the mean of the nine most common  congeners or calculation from
concentrations of total PCBs in fish and ambient water.

B.    Explanation of Final Provisions

1.    BAFs

      Aquatic organisms are exposed to chemicals through  the water that they
live in, the food that they eat,  and contact  with  sediment.  Chemicals enter
these organisms via gills, epidermis,  or the  gastrointestinal tract; this
uptake of chemicals process is called bioaccumulation.  For certain chemicals,
uptake through the food chain is the most important route of exposure.  As
lower trophic level organisms are consumed  by higher trophic level organisms,
the tissue concentrations of some chemicals increase through one or more
trophic levels so that residues in organisms  at trophic level 3 and 4 might be
many orders of magnitude greater than the concentration of the chemical in the
ambient water.  While the concentration in  the  ambient water may be too low to
affect the lowest level organisms,  this biomagnification process can result in
exposure concentrations for the consumers of  top trophic  level aquatic
organisms that are above the fish tissue concentrations that correspond to
current Clean Water Act (CWA)  section 304(a)  water quality criteria by several
orders of magnitude (58 FR 20816).   Table 1-1 in the proposed Guidance  (58 FR
20816)  compares the measured concentrations of  PCBs and three pesticides found
in fish tissue (De Vault 1993a)  with the fish tissue concentrations
corresponding to current CWA section 304 (a) water  quality criteria at a 10"5
risk level.  Consumers of these fish would  be consuming fish that contain up

-------
122    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

to 200 times the levels of PCBs calculated in the 304(a) criteria  to
correspond to a 10~5 risk.

      In the final Guidance, as in the proposal, EPA relies on BAFs to  reflect
the propensity of a-chemical to accumulate in the tissues of aquatic
organisms, accounting for exposure from all sources of a chemical.  In  order
to properly account for exposure to a chemical, both the wildlife  criteria and
the human health criteria and values have been developed to incorporate
appropriate BAFs.  In addition, the human health BAFs are used to  identify
Bioaccumulative Chemicals of Concern  (BCCs) which warrant increased attention,
and more stringent controls, within the basin.  See discussion of  BCCs  in
section II.C.8 of this document.

      As discussed in the proposal, bioaccumulation refers to the  net
accumulation of a substance by an aquatic organism from its ambient water,
sediment and food.  A BAF represents the ratio  (in L/kg) of a substance's
concentration in the tissue of aquatic organisms to its concentration in the
ambient water in situations where both the organism and its food are exposed
and the ratio does not change substantially over time.  Measured BAFs are
based on field data.

      A BCF  (in L/kg) is the net accumulation of a substance by an aquatic
organism from the ambient water only through gill membranes or other external
body surfaces.  BCFs are determined either by measuring bioconcentration in
laboratory tests  (comparing fish tissue residues to chemical concentrations in
test waters), or by.predicting the BCF from the octanol-water partition
coefficient  (Kow or p)  of a chemical.   The equation,  log BCF = 0.79 log KOW -
0.40, was used in the proposal to relate the BCF and the octanol-water
partition coefficient for a chemical  (see 58 FR 20859) .

      EPA's 1991 National guidance documents, the "Technical Support Document
for Water Quality-based Toxics Control" and draft "Assessment and  Control of
Bioconcentratable Contaminants in Surface Waters," recommend a methodology for
estimating the BAF where a field-measured BAF is not available.  This
methodology predicts the BAF by multiplying the BCF by a factor which accounts
for the biomagnification of a chemical through trophic levels in a food chain.
The numerical factor which represents the magnitude of this biomagnification
through the food chain is called the food-chain multiplier  (FCM) in these 1991
National guidance documents.  In the 1991 documents and the proposed Guidance,
EPA calculated the FCMs using a model of the step-wise increase  in the
concentration of an organic chemical from phytoplankton  (trophic level  1)
through the top predatory fish  (trophic level 4) of a food chain  (Thomann,
1989)  (see 58 FR 20859 for a discussion of the Thomann model).

2.    Measured and Predicted BAFs

a.    Hierarchy of Methods.

      i.    Proposal:  The proposed Guidance listed three methods  for deriving
BAFs for organic chemicals, described below  in order of decreasing preference:

       (1)   A BAF measured in the field, preferably in fish collected in the
Great Lakes which are at the top of the food chain;

       (2)   A BAF predicted by multiplying a BCF measured in the laboratory,
preferably on a  fish species indigenous to the Great Lakes, by the food-chain
multiplier;  and,

       (3)   A BAF predicted by multiplying a BCF calculated from the log KOW
 (using the equation, log BCF =  0.79 log KOW  - 0.40 (Veith and Kosian, 1983))
by the food-chain multiplier.

      Subsequent  to the proposed  Guidance, EPA requested comment in  a Notice
of Data Availability  (59 FR 44678) on incorporation of a BAF derived from  the

-------
                           Section IV: Bioaccumulation Factors                       123
BSAF methodology as the second method in the hierarchy for deriving BAFs for
organics.

      ii.  • Comments: Many commenters stated that field-measured BAFs and
predicted BAFs for a chemical do not correlate well and in most cases the
predicted BAF for a.chemical overestimates the field-measured BAF.  Because of
this the commenters'advocated using only field-measured BAFs when deriving
human health and wildlife criteria.

      Several commenters argued that BCFs should be used instead of BAFs
because of the high degree of variability and site-specificity in field-
measured BAFs; the current state of the science does not support the
transition from BCF to BAF; and biomagnification might be due to data
anomalies and oversimplification of the food chain rather than trophic
transfer of a chemical.

      Many commenters wanted EPA to discuss the uncertainty associated with
the BAFs because they were concerned that all the uncertainties inherent in
each portion of the methodology could produce an aggregate uncertainty larger
than the BAF.

      Several commenters questioned the use of a combination of the Veith and
Kosian  (1983) regression equation with the Thomann model (1989), and suggested
using either the whole Thomann approach or testing the validity of the
combination approach.

      Many commenters suggested using the BSAF or Bioavailability Index  (BI)
for derivation of criteria in place of the BAF.  Along the same lines, many
commenters advocated using the concentration in sediments instead of ambient
water when deriving criteria because sediments are the primary repository of
chemicals with log KOWS greater than four and using minute,  trace water
concentrations of chemicals with log KOWS greater than four in the derivation
of BAFs will result in inappropriately high values.

      Finally many commenters wanted the flexibility to revise the BAF when
new data become available, while others stressed the need for communication
between EPA and the regulatory agencies to share data on BAFs.

      EPA does not agree that the field-measured BAFs and predicted BAFs for a
chemical do not correlate well.  The adaptation of the Gobas model for
estimating FCMs (see discussion in section IV.B.4 below on the Gobas model)
reduces much of the uncertainty and variability associated with comparing
field-measured BAFs and predicted BAFs.  A comparison of the BAFs predicted by
the Gobas model (1993) against the field-measured BAFs from Oliver and Niimi
(1988)  for the 52 chemicals which have field-measured BAFs for at least three
fish shows that differences between the mean BAFs are less than a two-fold for
46 of the 52 chemicals, and less than a three-fold for 51 of the 52 chemicals
(Zipf,  1995) .  EPA concludes that when field-measured BAFs are not available,
the model used in the final Guidance acceptably predicts BAFs for the Great
Lakes System.

      EPA partially agrees with commenters who advocate using field-measured
BAFs when deriving criteria.  In the proposal, Tier I criteria and Tier II
values for human health and wildlife were differentiated based on the quantity
and quality of toxicological data only.  After reconsideration, EPA has
decided to differentiate the Tier I criteria and Tier II values for human
health and Tier I criteria for wildlife based on the quantity and quality of
both the toxicological and bioaccumulation data.  The minimum toxicological
data for human health is discussed in section V and for wildlife in section
VI.   The new minimum BAF data required to derive Tier I human health criteria
for organic chemicals include either:  (a) a field-measured BAF;  (b) a BAF
derived from the BSAF methodology, or  (c) a chemical with a BAF less than 125
regardless of how the BAF was derived.  The new minimum BAF data required to
derive Tier I wildlife criteria for organic chemicals include either:  (a) a

-------
124    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


field-measured BAF, or (b) a BAF derived from the BSAF methodology.  For all
inorganic chemicals, including organometallies such as mercury, the minimum
BAF data required to derive a Tier I human health and wildlife criteria
include either:  (a) a field-measured BAF or  (b) a laboratory-measured BCF.
For the majority of inorganic chemicals, the BAF is equal to the BCF (i.e.,
FCM = 1) because there is no apparent biomagnification or metabolism.  The
basis for these new requirements is explained below.

      Requiring the use of field-measured BAFs or field-measured BSAFs when
deriving Tier I criteria for organic chemicals with predicted BAFs greater
than 125 eliminates for these Tier I criteria concerns about the effect of
metabolism on the BAF.  This is the case because field studies measure
chemical concentrations in the tissues of the fish that are exposed to the
chemical from food, ambient water and sediment.  The measured concentrations
in the fish inherently account for the effect of metabolism from all sources
of exposure.  On the other hand, the concentration of the chemical in fish
tissue from laboratory-measured BCF accounts for exposure from ambient water
only.  Consequently, BCFs do not account for the effect of metabolism of
chemicals accumulated from exposure through the diet of the aquatic organism.
Metabolism may either increase or decrease the concentration of a chemical and
its by-products in the tissue of an aquatic organism.  EPA's BAF methodology
uses a model  (Gobas, 1993) to predict the accumulation of a chemical from food
sources.  The model of Gobas (1993), however, does not account for the effects
of metabolism; in other words,  the entire concentration of the chemical input
to the model plus the concentration biomagnified through the food chain is
predicted to accumulate in the fish tissue.  Consequently, BAFs based on
multiplying a laboratory-measured BCF times a FCM (predicted from Gobas, 1993)
may either under- or overestimate the amount of a chemical a fish will
bioaccumulate.  In addition, predicted BAFs that are obtained by multiplying a
predicted BCF by a FCM make no allowance for metabolism.

      EPA has decided that Tier I criteria, which must be adopted into State
regulations and become the goal for permit limits throughout an entire State,
should accurately account for the effects of metabolism.  Accordingly, EPA is
requiring the use of field studies for determining a BAF to be used in the
derivation of a Tier I criterion for human health.

      Tier I human health criteria may also be derived for chemicals with BAFs
less than 125 regardless of which of the four methods specified in the final
Guidance is used to derive the BAF.  For chemicals with a BAF less than 125,
exposure from consumption of fish is less than or equal to the exposure from
consumption of drinking water.  This assumes a fish consumption rate of 15
grams per day and drinking water consumption of two liters per day.  Thus for
these chemicals the effects of metabolism by aquatic organisms are not as
significant a determinant of ultimate human exposure as for chemicals with
larger BAFs.  Therefore, for organic chemicals with a BAF less than 125, all
four methods specified in the final Guidance can be used to obtain a BAF used
to derive Tier I criteria.

      EPA has decided, notwithstanding the fact that predicted BAFs will at
best only partially account for the effects of metabolism, to allow derivation
of Tier II human health values using BAFs based on BCFs and the FCM.  EPA's
decision is based on several factors.  First, as described elsewhere in this
document, available information indicates a very good correlation between
predicted BAFs based on the methodology in the final Guidance and field-
measured BAFs.  Therefore, it appears that the error introduced by using
predicted BAFs rather than field-measured BAFs may be relatively small for
many chemicals.  Second, Tier II values by definition are based on a less than
ideal data base.  Acknowledging that  the data available is not perfect for
Tier II chemicals, EPA believes that  it is nevertheless important to protect
human health  from potential adverse effects  resulting from their discharge to
surface waters.  The Tier II methodology, including use of predicted BAFs,
specifies the best available protocols for assessing protective ambient levels
for these chemicals for which Tier  I  data are not available  This will provide

-------
                           Section IV: Bioaccumulation Factors                       125
consistency among Great Lakes States and Tribes in the interpretation of
narrative human health criteria when establishing water quality-based effluent
limitations  (WQBELs).  To the extent that a discharger of a chemical for which
a Tier II value is established on the basis of a predicted BAF questions
whether a field-measured BAF would not provide a more accurate measure of
bioaccumulation, that discharger is free to conduct an adequate field study
and request that it-be utilized by the appropriate regulatory authority in
establishing a-Tier II value or Tier I criterion.  Finally, for those
chemicals where available data indicates that metabolism is likely to
considerably increase or decrease a predicted BAF, appendix B section VIII of
the final Guidance allows States and Tribes to modify the predicted BAFs
accordingly.  EPA believes that derivation of an "effective FCM" for certain
chemicals, as described in the proposed Guidance, may in many cases reasonably
account for the metabolism.  However, States and Tribes are not limited to the
use of an "effective FCM" in correcting a predicted BAF for metabolism.

      EPA disagrees with commenters that BCFs should be used instead of BAFs.
Bioaccumulation is what occurs in nature, and is what determines the total
concentration of chemicals in aquatic organisms that are consumed by humans
and wildlife.  For some chemicals the biomagnification of a chemical through
the food chain can be substantial.  Using BCFs, which only account for
exposure from the ambient water, could substantially underestimate the
potential exposure to humans and wildlife for some of these chemicals and
result in criteria or values which are underprotective.  The use of BAFs,
which account for uptake from all sources, will ensure that the potential
exposure from these chemicals is adequately accounted for in the derivation of
human health and wildlife criteria.  Using BAFs is the most comprehensive and
scientifically valid approach.  As mentioned in the proposal (58 FR 20858),
BAFs have been used in deriving human health criteria development since 1980.

      EPA recognizes that field-measured BAFs will have some variability from
site to site.  In recognition of this, EPA allows the derivation of site-
specific BAFs as discussed in procedure 1 of appendix F to part 132.  Although
there might be some variability in field-measured BAFs, it does not invalidate
their usefulness in estimating the potential exposure to humans and wildlife,
nor does it imply that BAFs are less accurate than BCFs in predicting that
exposure.

      In addition, EPA does not agree with commenters that biomagnification
might be due to data anomalies and oversimplification of the food chain.
There is ample evidence of biomagnification occurring because of trophic
transfer of chemicals.  The importance of uptake of chemicals through the diet
and the potential for a stepwise increase in bioaccumulation from one trophic
level to the next in natural systems has been recognized for many years
(Hamelink, et.  al.,  1971) .   Many researchers have noted that the BAFs of some
chemicals in nature exceed the BCFs measured in the laboratory or estimated by
log KOW models  (e.g.,  Oliver and Niimi 1983,  Oliver and Niimi  1988,  Niimi
1985, Swackhammer and Hites 1988).

      Further,  EPA believes that the state of the science supports the use of
BAFs.  EPA's Science Advisory Board  (SAB) in its December 16,  1992, report on
the Evaluation of the Guidance for the Great Lakes Water Quality Initiative
stated that the BAF procedure is more advanced and scientifically credible
than existing BCF procedures and that the use of the BCF, FCM,  and BAF
approach appears to be fundamentally sound (EPA-SAB-EPEC/DWC-93-005).  In a
subsequent SAB report on August 12, 1993 on the ongoing revision of the
methodology for deriving National Ambient Water Quality Criteria for the
protection of human health  (EPA-SAB-DWC-93-016), the Drinking Water Committee
reported on a similar BAF methodology.  Although cautioning that its
"criticisms should not be taken as a recommendation to relax standards or to
ignore the potential for bioaccumulation where it is known to play an
important role," the Drinking Water Committee'also stated that "for the time
being, the Agency should focus attention on BCFs rather than BAFs, because of
the higher likelihood of collecting an adequate BCF data base."  In evaluating

-------
126    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

the two SAB committee reports, it is  important to keep in mind that the first
committee was reviewing the proposed  BAF methodology for the Great Lakes
Guidance, while the second committee  was reviewing a similar proposed
methodology that would be applicable  nation-wide.  Thus, the second
committee's recommendation that a sufficient BAF data base may not be
available at the present time to derive BAFs for chemicals for nation-wide
criteria guidance does not imply that sufficient information is not available
to develop BAFs for regional water quality standards in the Great Lakes'.
Indeed, to rely on BCFs in the Great  Lakes System would be directly contrary
to the Drinking Water Committee's exhortation that their criticism not be
taken as a recommendation "to ignore  the potential for bioaccumulation where
it is known to play an important role."  EPA has revised the BAF methodology
where possible at this time to take into account the concerns raised by both
SAB Committees, and. believes after careful review of SAB and public comments
that use of BAFs in the final Guidance represents the most scientifically
defensible approach for accounting for chemical uptake by aquatic biota in the
Great Lakes System.

      EPA also recognizes there were  some uncertainties in application of the
proposed BAF methodology and has addressed these in the final Guidance.  For
example, to reduce the uncertainty in predicting the biomagnification of
chemicals, EPA is using a model in the final Guidance that uses Great Lakes
specific parameters and includes a benthic food chain component to estimate
FCMs.  In addition, the final Guidance takes into account the freely dissolved
concentration of a chemical in the derivation of BAFs for organic chemicals.
Taking the freely dissolved concentration into account will eliminate much of
the variability associated with specific waterbodies because most of the site-
specific differences in bioaccumulation arises from the partitioning of the
chemical to the particulate organic carbon and dissolved organic carbon of the
water column.  However, professional  judgement is still required throughout
the derivation of BAFs and a degree of uncertainty is still associated with
the determination of any BAF, BSAF, BCF or KOW.   Despite this uncertainty,
EPA maintains that BAFs are the most  useful measure of the exposure of an
aquatic organism to all chemicals.

      EPA agrees with those commenters who suggested the use of the BSAF
approach for deriving BAFs and has modified the proposed hierarchy of methods
for deriving BAFs to include a BAF derived from the BSAF methodology as the
second preferred method after field-measured BAFs.  The BSAF provides a method
by which the concentrations of chemicals in the sediment are related to the
concentrations in fish tissue.  The concentrations of chemicals with log KQWS
greater than 6.5 are greater in the sediment than in the water column and more
readily measured; therefore use of the BSAF reduces the uncertainty associated
with relating concentration in fish tissue to the concentration  in the water
column for these chemicals.  This is  particularly true for chemicals with
higher KOWS since these generally show a greater affinity for sediments.  The
BI is the same method as the BSAF and the terms can be interchanged.  For
further details on deriving BAFs using the BSAF methodology, and the data
supporting the approach, see the final Great Lakes Water Quality Initiative
Technical Support Document for the Procedure to Determine Bioaccumulation
Factors  (EPA 820-B-95-005)  (BAF TSD)  for BAFs which is available in the public
docket for this rulemaking.

      In response to commenters who wanted EPA to develop sediment criteria
instead of water criteria, EPA asserts that ambient water quality criteria
provide a measure of acceptable chemical levels in the medium in which  non-
benthic aquatic organisms primarily exist, and are therefore an  important
measure of acceptable environmental conditions for these organisms.  EPA
agrees that sediment criteria may also provide useful indices of acceptable
chemical levels, especially when the  greatest risks are known to be associated
with toxic effects  to benthic organisms.

      EPA has  changed the final Guidance in response to commenters' concerns
with the use of the Veith and Kosian  regression equation in  combination with

-------
                           Section IV: Bioaccumulation Factors                       127
the Thomann model  (1989).  In the final Guidance, the equation baseline BCF =
KQW that is used to predict BCFs is equivalent to the equation BCF" = KQW for
log KQWS less than three that is used in the Gobas 1993 model to predict FCMs.
Thus there is no longer a need to validate the use of the regression equation
by Veith and Kosian  (1983) in combination with the Thomann model  (1989).

      Finally, EPA agrees with commenters that it is important to revise BAFs
when new data become available and anticipates that the Clearinghouse
discussed in section II.C.I will provide the mechanism through which new data
are evaluated and disseminated.  In addition, the final hierarchy of data
preference allows for the incorporation of new data.  For example, if a field-
measured BAF is calculated for a chemical for which only a predicted BAF was
previously available, preference would be given to the field-measured BAF.

      iii.  Final Guidance:  EPA revised the proposed hierarchy of methods for
deriving BAFs based on public comments.  The final Guidance lists four methods
for deriving BAFs for organic chemicals, listed below in order of decreasing
preference: a BAF measured in the field, in fish collected from the Great
Lakes which are at or near the top of the food chain; a BAF derived using the
BSAF methodology; a BAF predicted by multiplying a BCF measured in the
laboratory, preferably on a fish species indigenous to the Great Lakes, by the
FCM; and a BAF predicted by multiplying a BCF calculated from the KOW by the
FCM.

b.    Field-Measured BAFs

      i.    Proposal:  As discussed above, the proposal stated a preference
for using field-measured BAFs over predicted BAFs.  This preference stems from
the fact that field-measured BAFs automatically account for any
biomagnification and metabolism that might occur.  The proposal also stated
that field-measured BAFs should be based on fish species, preferably living in
the Great Lakes at or near the top of the aquatic food chain.  In its December
16, 1992 report, the EPA's SAB stated that field-measured BAFs must be
interpreted very carefully, and it should be recognized that they might
contain substantial errors and variability due to several factors (see 58 FR
20860).

      ii.   Comments:  Many commenters stated that the field-measured BAFs
might contain substantial errors and expressed a preference for the more
established BCF.  Other commenters were concerned that the field-measured BAFs
were based on a single data source, that limiting the data base to the Great
Lakes for calculation of field-measured BAFs ignores a wealth of information,
and that it was important to show that the BAFs measured in one lake
accurately predict fish tissue levels in other Great Lakes.  Several
commenters suggested developing guidelines for measuring and/or evaluating
field-measured BAFs. which would include provisions for the simultaneous
collection of water and fish tissue and accounting for fish mobility, as well
as discussing the importance of the sampling location and temporal change.

      EPA acknowledges that there can be errors in determining field-measured
BAFs,  as with any field-measurements, and has attempted to minimize these
potential errors when deriving BAFs for the final Guidance by carefully
screening the data used to calculate the BAFs.  EPA continues to contend that
a field-measured BAF is a more accurate gauge of what is occurring in nature
than a laboratory-measured or -predicted BCF because the BAF measures the
actual impacts of biomagnification, bioavailability, concentration in the
sediment,  growth dilution, and metabolism rather than predicting them through
use of a model.

      EPA also agrees with the SAB's comments and with commenters concerned
about the difficulty of collecting and interpreting field-measured BAFs.  EPA,
however, thinks that States and Tribes can adequately use and interpret field
studies.  To assist them in this task, EPA plans to provide guidance
concerning the determination and interpretation of field-measured BAFs before

-------
128    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

the States and Tribes are required to adopt water quality standards  consistent
with the final Guidance.  This will provide interested parties with  a  set  of
procedures that will assist them in collecting and interpreting  the  field-
measured BAFs.

      The majority of data used to calculate the field-measured  BAFs in the
final Guidance came from the data of Oliver and Niimi  (1988).  This  data set
is generally recognized as being the most complete set of data available in
the Great Lakes for estimating field-measured BAFs.  EPA acknowledges  that the
data from Oliver and Niimi come from Lake Ontario, but believes  that the data
can be used to predict BAFs in other Great Lakes because the values  take into
account the percent lipid and are based on the freely dissolved  concentration
of the chemical in the ambient water.  Taking the lipid content  into account
allows the data to be applied to other fish species.  Derivation of  the BAFs
on a freely dissolved basis from field data eliminates the site-specific
nature of the BAFs caused by the amounts of dissolved and particulate  organic
carbon present at the field site and therefore, allows the use of  the  derived
BAFs in the other Great Lakes.

      Using data from the Great Lakes is preferable to using information from
other bodies of water because it better represents the physical, chemical, and
hydrological conditions present within the Great Lakes.

      iii.  Final Guidance:  The final Guidance requires that field-measured
BAFs be the preferred method for deriving BAFs because of their  ability to
account for biomagnification, growth, metabolism, concentration  in the
sediment, and bioavailability.

c.    Field-Measured BSAFs

      i.    Proposal:  As discussed in the August 30, 1994 Notice  of Data
Availability (59 FR 44678), BSAFs can be used for measuring bioaccumulation
directly from concentrations of chemicals in surface sediments or  to estimate
BAFs.  Because BSAFs are based on field data and incorporate the effects of
metabolism, biomagnification, growth, concentration in the sediment, and
bioavailability, BAFs derived from the BSAF methodology will incorporate the
net effect of these factors.  The BSAF approach is particularly  beneficial for
developing water quality criteria for chemicals such as polychlorinated
dibenzo-p-dioxins  (PCDDs) and polychlorinated dibenzofurans  (PCDFs)  and
certain biphenyl congeners which are detectable in fish tissues  and  sediments
but are difficult to measure in the ambient water.

      BSAFs are measured by relating  lipid-normalized  concentrations of
chemicals in an organism to organic carbon-normalized  concentrations of the
chemicals in surface sediment samples associated with  the average  exposure
environment of the organism.  The BSAF is defined  as:

                                           C/
                                 BSAF = —-L


      where:

            C,    =  the lipid-normalized concentration of the chemical in
                     tissues of the biota  (/zg/g lipid) .

            Csoc   =  the organic carbon-normalized concentration of  the
                     chemical in the surface sediment  (jig/g  sediment organic
                     carbon).

For further explanation,  see  the  BAF  TSD.

       ii.   Comments:   Commenters generally supported the  use  of the BSAF
methodology in deriving BAFs.   Some  commenters  stated that while chemicals in
the sediment may contribute  to  bioaccumulation,  the  proposed BSAF model is

-------
                           Section IV: Bioaccumulation Factors                       129
only valid if limited to a chemical- and site-specific context.  Some
commenters suggested that EPA more fully address the issue of relative
concentrations of a chemical in the sediment and water column.

      Other commenters questioned the validity of data for reference chemicals
used when deriving BAFs from the BSAF methodology since the data are from a
single unpublished study.  Others raised questions about the availability of
the study.

      EPA acknowledges that BAFs derived from field-measured BSAFs require
chemical-, species- and site-specific data.  EPA is not intending to place an
undue burden on the States, Tribes or dischargers to generate data.  If
pertinent data are available, EPA has set forth a methodology in which they
can be used.  Basing the BSAF on lipid-normalized concentrations and organic
carbon-normalized concentrations eliminates many of the species- and site-
specific characteristics, and therefore the BSAF is applicable to a wider area
than that at which it was measured.

      In response to commenters'  concerns about the relative concentrations of
a chemical in the sediment and the ambient water, EPA believes that fish are
exposed to organic chemicals through contact with water, food and, and to some
extent,  sediment.  At steady-state, the concentrations of these chemicals in
water or surface sediment, although numerically quite different, are equally
useful for prediction of bioaccumulation.in fish.  When concentrations of some
chemicals are temporally variable and/or nondetectable in water, BSAFs can
provide the most reliable field-measurement of bioaccumulation.  BAFs are
needed,  however, to calculate water quality criteria.  Fortunately, the BSAF
methodology inherently includes a measure of the disequilibrium that usually
occurs between the sediment-water distribution of the chemicals by using
reference chemicals for which a field-measured BAF is available.  The relative
concentrations of the chemical in the sediment and water are therefore
accounted for in the BSAF.  The BSAF method translates the bioaccumulation and
disequilibrium information presented by the BSAF into a BAF through comparison
to reference chemicals with similar sediment-water disequilibrium at the same
site.  In this method the reference chemicals provide key relationships
between measured BSAFs and BAFs for the ecosystem.

      EPA understands the commenters' concerns with the availability of the
data used for deriving BAFs from the BSAF methodology.  All the BSAF, BAF and
KQW data used to derive the BAFs  based on the  BSAF methodology were presented
in the August 30, 1994 Notice of Data Availability (59 FR 44678).  EPA
acknowledges that some of the fish and sediment analytical data have not been
published. However, the use of the Oliver and Niimi data (1988), which have
been published,  to demonstrate correlations and numerical similarity between
BSAFs and BAFs calculated from the two independent data sets should provide
additional assurance of the applicability of the unpublished data for
demonstration of the BSAF method for deriving BAFs from BSAFs.

      iii.  Final Guidance:  In the final Guidance, a BAF derived from the
BSAF methodology is added as the second most preferred method in the hierarchy
for derivation of BAFs for organic chemicals.

d.    Measured and Predicted BCFs

      i.    Proposal:   The proposed Guidance preamble (58 FR 20859) discussed
three analytical techniques that can be used to measure organic chemicals in
tissue and water for the purposes of establishing a laboratory-measured BCFs:
gas chromatography, high pressure liquid chromatography or radio-labeled
organic chemicals.

      The proposed Guidance predicted BCFs from the octanol-water partition
coefficient (KQW or P)  of a chemical using the equation:

            log BCF =0.79 log KQW - 0.40 (Veith and Kosian,  1983)

-------
130    Water Quality Guidance for the Great Lakes System ~ Supplementary Information Document


Subsequent to the proposed Guidance, EPA requested comment in the August 30,
1994, Notice of Data Availability  (59 FR 44678) on the use of an alternative
equation from the proposal to predict BCFs.  The equation was:

                                 BCFf  = Kow

where the BCF,W is reported on a lipid-normalized basis using the freely •
dissolved concentration of the chemical in the water.

      ii.   Comments:   Some commenters indicated that use of radio-labeled
organic compounds would overestimate the laboratory-measured BCF.  Other
commenters stated that difficulties with radio-labeled materials are easily
avoided in actual practice.

      Several commenters stated that predicted BCFs should not be used  for
chemicals that are suspected of metabolism.  Many commenters supported  the
change from the Veith and Kosian  (1983) equation to the equation BCF" = KOW/
but recommended that EPA specify the limitations of its use;  commenters
wanted EPA to be explicit about the assumed percent lipid in derivation of the
BCF.  Several commenters requested development of guidelines for measuring
KOW-

      EPA partially agrees with the commenters who recommended against  the use
of radio-labeled organic chemicals for measuring BCFs because the organism may
also accumulate a metabolite of the parent compound, thereby overstating the
actual BAF.  Attempts to measure the amount of radio-labeled compound obtained
through tissue analysis, instead of measuring the radioactivity of the  fish,
have not been definitive.  There is also the possibility of contamination of
the labeled compound.   Because of these concerns, EPA has decided that  BCFs
for organic chemicals may be based on measurement of radioactivity only when
the BCF is intended to include metabolites or when there is confidence  that
there is no interference due to metabolites.

      EPA partially agrees with commenters who suggest that predicted BCFs
should not be used for chemicals that are suspected of metabolism.  EPA stated
in the Technical Support Document for the Procedure to Determine
Bioaccumulation Factors  (EPA-822-R-94-002) that the relationship BCF" = KQW
is applicable to organic chemicals which are either slowly or not metabolized
by aquatic organisms.   Since predicted BCFs do not account for metabolism,
they will not be used in the derivation of Tier I human health and wildlife
criteria unless the predicted BAF is  less than 125.  Predicted BCFs, however,
can be used in the derivation of Tier II human health values if no laboratory-
measured BCF data are available.  For a more detailed discussion on Tier II
values, see section IV.B.2.a.ii of this document.

      In response to commenters' questions regarding the assumed percent
lipid, the 1983 average percent lipid associated with the Veith and Kosian
equation was 7.6.  The BCF in the equation BCF" = KOW is reported on a  lipid
normalized basis, and therefore by definition assumes 100 percent lipid.  This
BCF is then adjusted to the percent lipid appropriate for a given trophic
level.  See section III.B.3 of this document for further explanation.

      EPA is in the process of developing guidelines for measurement of KOWS.
In  the interim period,  section III.F  of appendix B to part 132 lists the
analytical technique priorities for deriving  KOWS.  KOWS are an integral
factor in developing BAFs used in derivation of human health criteria and
values and wildlife criteria; therefore, EPA believes that providing guidance
on  the acceptability of KOWS will result in more consistent criteria.

-------
                           Section IV: Bioaccumulation Factors                       131
      iii.  Final Guidance:  In the final Guidance, predicted BCFs are derived
from a chemical's KQW using the following equation:

                             Baseline BCF = Kow

See the final BAF TSD for details.

      The BCF .based on this equation provides a more consistent and
scientifically defensible basis for predicting BAFs than the equation  (Veith
and Kosian, 1983) used in the proposal.  The theoretical basis presented by
Mackay (1982) and the experimental data referenced in the August 30, 1994,
Notice of Data Availability  (58 FR 44678), suggest that octanol is a
reasonable surrogate for lipids.

      For laboratory-measured BCFs, EPA continues to strongly recommend the
use of the procedural and quality assurance requirements specified in the
American Society for Testing Materials (ASTM) (1990) "Standard Practice for
Conducting Bioconcentration Tests with Fishes and Saltwater Bivalve Molluscs,"
and in the EPA guidance contained in Stephan et al. (1985), "Guidelines for
Deriving Numerical National Water Quality Criteria for the Protection of
Aquatic Organisms and Their Uses."

e.    Inorganic Chemicals

      i.     Proposal:  The proposed Guidance limited the methods for obtaining
a BAF for inorganic chemicals to a field-measured BAF or a laboratory-measured
BCF.  This is because no method is available for reliably predicting BCFs or
BAFs for inorganic chemicals.  BCFs and BAFs for some inorganic chemicals vary
between species, and from one tissue to another within a species.  The
proposal included BAFs for 17 inorganic chemicals, including mercury.

      ii.   Comments:  Several commenters indicated that there was a wide
range of variability in laboratory-measured BCFs for inorganic chemicals.
Other comments indicated that in order to check the reliability of measured
data, modelling should be allowed for inorganic chemicals.  Other commenters
advocated site-specific modifications for metals based on each Lake's
characteristics.

      EPA acknowledges that there is some variability in laboratory-measured
BCFs for inorganic chemicals, and that there are real differences between
species.   For example, saltwater molluscs have substantially higher BCFs for
many metals than do freshwater and other saltwater fishes and invertebrates.
Variability is not a major 'concern for the majority of inorganic chemicals,
however,  because their BCFs are low for species consumed from the Great Lakes.
For example, the BAF for ten of the 17 inorganic chemicals in the list of
chemicals in Table 6 of part 132 is below ten.  This indicates that even
though there may be some variability in laboratory-measured BCFs, the impact
on the final criteria will be small.

      EPA agrees with commenters that modelling should be allowed if the model
has been shown to be appropriate.   EPA advocates but does not require the use
of modelling to verify the reliability of the measured data for inorganic
chemicals. In addition,  EPA agrees that site-specific modifications should be
allowed for BAFs and provides for this in procedure 1 of appendix F to part
132.

      iii.  Final Guidance:  After review of the comments, EPA  decided that
the proposed Guidance for derivation of BAFs for inorganic chemicals will not
be amended.  The most accurate measurement of bioaccumulation for inorganic
chemicals are field-measured BAFs  and laboratory-measured BCFs.

-------
132    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

3.     Lipid Values

      Consistent with the existing National guidance, the proposed Guidance
relied on the assumption that an organism's ability to bioaccumulate organic
chemicals is related to it's lipid content.

      Therefore to determine a BAF for organic chemicals for use in deriving
wildlife Tier I criteria and human health Tier I criteria and Tier II values
it is necessary to know the percent lipid content of the organisms being
consumed.  Humans typically eat fish  fillets which usually have  lower lipid
content than the whole fish consumed  by wildlife.  EPA's current recommended
National percent lipid values for calculation of human health criteria  is 3
percent  (U.S. EPA, 1980}.  There is currently no National guidance for
wildlife and thus there is no recommended percent lipid value.

      In the proposal, EPA requested  comments on a variety of issues including
what solvent should be used in the measurement of percent lipids.  Some
commenters advocated the use of a standardized extraction method and a
consistent system to measure lipid content, while a few suggested use of
methylene chloride as the extraction  solvent.  No rationale for  selecting
between the solvents which have been  proposed was presented by commenters.
EPA believes that the data is inconclusive  as to which extraction method and
solvent to use and has therefore has  not made a recommendation on which method
and solvent to use in the measurement of percent lipids.  EPA will be
providing additional guidance on which extraction method(s) and  solvent(s) to
use in the guidance on the determination and interpretation of field-measured
BAFs.

a.    Lipid Value for Human Health BAFs

      i.    Proposal:  The proposed Guidance used a lipid value  of 5.0  percent
in edible tissue for use in determining human health BAFs for organic
chemicals.  Percent lipid data for edible tissue  (mostly skin-on fillets) were
gathered from the fish contaminant monitoring programs in Michigan, Wisconsin,
Ohio, Indiana, New York and Minnesota.  The use of skin-on data  to determine
the lipid values provided an extra margin of safety to the many  anglers who
remove the  skin  from the fillet.

      In selecting the lipid value for human health BAFs, lipid  data for the
following fish groups were considered: salmonids only; salmonids and non-
salmonid game fish; and all fish  (game and  nongame species).

      The mean lipid value for salmonid and non-salmonid game fish of 5.02
percent  lipid was proposed because this option best represented  the range of
species  typically cpnsumed by people  in the Great Lakes basin.   Also
considered  was the mean lipid values  weighted by human consumption patterns.
The resulting consumption-weighted mean for all sport-caught game fish  was
4.72 +  2.42 percent  lipid.  Because these results were not  statistically
different from the unweighted mean, use of  the unweighted mean of 5.02  percent
 (rounded to 5.0  percent) was proposed for the human health  BAFs.

      ii.    Comments:  A few commenters did not agree with  the use of skin-on
filets,  and suggested using  skin-off  filets for lipid measurement with  a
corresponding lipid  value  of four percent.   Other commenters stated that  the
five percent lipid value did not  provide  an adequate  margin of safety nor
protect  high-risk subpopulations  (e.g., Tribes and subsistence fisherpersons)
and  suggested using  an 11  percent lipid value.  Other commenters suggested
using three percent  for  the  tributaries of  the Great  Lakes, which is
 consistent  with  the  National guidelines dated November  28,  1980  (45 FR  79347) .
 Some commenters  stated that  the  data  did  not exist to determine  which fish
 species were consumed and  the  corresponding percentage  of  lipid  for high-risk
 groups.

-------
                           Section IV: Bioaccumulation Factors                       133
      In order to further examine whether the five percent lipid value was
appropriate, EPA conducted additional analysis of the data from a second fish
consumption survey conducted by West, et al.  (1993)  (see section V of this
document, Human Health, for a complete discussion of this study).  EPA
requested comments on the appropriateness of the data presented in the study
in a Federal Register notice on August 30, 1994  (59 FR 44678).  The results
from this analysis indicate that the consumption-weighted mean percent lipid
value for trophic level 4 fish is 3.10 and 1.82 for trophic level 3.  EPA
believes that the use of the West et al.  (1993) survey to estimate the percent
lipid used for deriving BAFs is an improvement on the methods utilized in the
proposal because the West survey allows a determination of the actual fish
species consumed and the rate of consumption.  When this information is
coupled with the information on percent lipid values for these fish, it is
possible to derive a more accurate reflection of the grams of lipid from fish
that are consumed by humans from each trophic level.  EPA acknowledges that
the West study only covered anglers in the State of Michigan, but concludes it
represents the best study to use for deriving consumption-weighted mean
percent lipid values for trophic levels 3 and 4.   States and Tribes can derive
alternative percent lipid values to be used in the derivation of site-specific
BAFs whenever they have the information needed to revise the derivation.

      EPA does not agree with those commenters advocating the use of skin-off
fillets for deriving percent lipid values.  Although many people remove the
skin and other fatty tissue when they prepare their fish for cooking, the
study by West et. al.  (1993) indicates that about 37 percent of anglers in
Michigan continue to prepare fish with the skin on even though they are aware
of the State fish advisories recommendation to trim fat and/or skin from the
fish.

      EPA also disagrees with those commenters advocating the use of a three
percent lipid value for tributaries.  Due to the mobility of the prey and its
host, it is difficult to characterize their territorial range.  The
consumption weighted mean percent lipid for the respective trophic levels
represents an overall average for the Great Lakes System.  The fish lipid data
used to determine the percent lipid values were gathered from fish contaminant
monitoring programs in the Great Lakes System  (including its tributaries)  and
represent the species consumed by people in the West et al. survey.  EPA also
does not agree that the lipid values should be increased to 11 percent
representative of lake trout, a species with the highest lipid value.  In the
majority of the cases people consume a variety of species and not simply lake
trout,  as evidenced by the West survey.  The lipid values selected for use in
deriving BAFs represent the wide variety of fish consumed by sport anglers in
the Great Lakes System.  In cases where it can be documented that a
subpopulation consumes fish with an average lipid content higher than those
prescribed in the final Guidance, then it may be appropriate for a permitting
authority to increase the lipid value in deriving a site-specific criterion
for waters fished by the subpopulation.  The permitting authority should
evaluate all aspects of exposure, including amount consumed, before altering
just one factor such as percent lipid, since the values for these variables
are interrelated.

      iii.  Final Guidance:   EPA has specified the use of a consumption-
weighted mean percent lipid value for trophic level 4 fish of 3.10 and 1.82
for trophic level 3 in edible tissue for use in determining human health BAFs
for organic chemicals in the final Guidance.

b.    Lipid Value for Wildlife BAFs

      i.    Proposal:   A lipid value of 7.9 percent for wildlife BAFs, based
on consumption of whole fish, was included in the proposal.  The lipid value
for the wildlife BAFs was determined using whole fish lipid data from the U.S.
Fish and Wildlife Service National Contaminant Biomonitoring Program and the
Canadian Department of Fisheries and Oceans.  The 7.9 percent lipid value was
the mean of lipid values for all fish, game and nongame, in the entire Great

-------
134    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

Lakes System.  Data for all fish were used because wildlife typically are
nondiscriminatory consumers of fish.

      ii.   Comments:  A few comments indicated that the percent  lipid  value
should be based on fattier fish.  Other commenters felt that 7.9  percent lipid
was based on a reasonable analysis and assumptions, but would like more
information on. prey preferences and whether high lipid organs are
preferentially eaten by wildlife.

      iii.  Final Guidance:  In the final Guidance, the percent lipid for the
actual prey species consumed by the representative wildlife species is  used to
estimate the BAF for the trophic levels at which wildlife consume.  The
percent lipid is based on the consumption patterns of wildlife and cross-
referenced with fish weight and size and appropriate percent lipid  (see final
TSD for BAFs).  This approach is a more accurate reflection of the lipid
content of the fish consumed by wildlife species than the approach used in the
proposal.

      EPA has required use of a percent lipid value for trophic level 4 fish
of 10.31 and 6.46 for trophic level 3 in whole fish for use in determining
wildlife BAFs for organic chemicals in the final Guidance.

4.    FCMs

      a.    Proposal:  As discussed in the proposed Guidance, when a field-
measured BAF is not available, a predicted BAF can be calculated  by
multiplying a laboratory-measured or predicted BCF by a food-chain multiplier.
The FCM accounts for the biomagnification of a chemical through trophic levels
in the food chain.  The FCMs in the proposal were based on a model by Thomann
(1989)  (see 58 FR 20859 and 20861 for a more complete description of the model
and its uses).

      In the August 30, 1994 Notice of Data Availability  (59 FR 44678), EPA
requested comment on use of a food-chain model by Gobas  (1993) which, unlike
the Thomann 1989 model, includes both benthic and pelagic food chains,  thereby
estimating exposure of organisms to chemicals from both the sediment and the
water column.

      b.    Comments:  Many commenters cited the 1992 SAB report  contention
that the FCM model  (Thomann, 1989) has not been adequately peer reviewed or
sufficiently validated to be used in a regulatory framework.  Commenters also
questioned the applicability of one model to a diverse ecosystem  such as the
Great Lakes and encouraged EPA to discuss alternate models.

      Many commenters criticized the use of the Thomann model for not
adequately accounting for metabolism, biotransformation, degradation,
persistence, or seasonal or temporal variability.  In addition, commenters
argued that the model is extremely sensitive to certain input parameters such
as the lipid content and that the input parameters for the FCMs that EPA
calculated were taken directly from Thomann's paper instead of using Great
Lakes specific assumptions.  Other commenters questioned the model assumption
that the system is at steady state, while others supported this assumption due
to the difficulty in describing the ecosystem parameters that would affect the
likelihood of reaching equilibrium.

      Many comments stated that the Thomann model had little application for
chemicals with log KQWS greater than 6.5 because it did not consider sediment
as a route of exposure.  Commenters suggested using alternative food-chain
models such as Thomann  (1992) which incorporates sediment as a route of
exposure; or Thomann and Connolly  (1984) which has Great Lakes specific input
parameters.

      Some commenters on the August 30, 1994 Notice of Data Availability  (59
FR 44678) supported the use of FCMs derived from the Gobas model, assuming

-------
                           Section IV: Bioaccumulation Factors                       135
that no metabolism occurs, because it incorporates the concentration of the
chemical in both sediment and the water column.  Other commenters on the
Notice requested further explanation as to why some FCMs for trophic level 3
are greater than trophic level 4.  Other commenters stated that additional
field validation and documentation of both the Thomann 1989 and the Gobas 1993
model is needed before they can be used in a regulation.

      EPA acknowledges that the 1989 Thomann model does not account for
certain important processes such as metabolism.  Similarly, although the 1993
Gobas model includes a metabolic rate constant, it was set equal to zero due
to a scarcity of data for individual chemicals.  Based on EPA's review of the
commenters1 concerns regarding metabolism, EPA has decided in the final
Guidance to differentiate which BAF data are required to derive Tier I human
health and wildlife criteria for organic chemicals based on whether or not
metabolism is taken into account.

      EPA agrees with commenters that it is important to use Great Lakes-
specific parameters whenever possible and that there should be an attempt to
account for the most sensitive input parameters to the model.  In light of
these concerns, EPA has used Great Lakes-specific input parameters in the
Gobas model that is used to derive FCMs for the final Guidance.  In addition,
EPA selected the model of Gobas (1993) to derive FCMs in part because this
model, in contrast to the model of Thomann (1992), required fewer input
parameters and had input parameters which could be more easily specified.

      EPA agrees that assuming a system is at steady state might be a
simplifying assumption in some cases, but as noted by commenters it is very
difficult to determine the parameters affecting a system reaching equilibrium.
This is especially true in a system as large as the Great Lakes.  The model of
Gobas (1993)  when used with conditions that are not at steady-state will
predict BAFs which are very similar to those obtained from steady-state
conditions.  The concentrations of the chemicals in the water and sediment in
the Gobas model (1993),  are used as input parameters and therefore, the
disequilibrium between the water column and the sediments are included in the
model calculations.   The differences between measured and predicted BAFs when
conditions are changing depends upon the rate of change.  For the Great Lakes,
the rate of change for PCBs and other bioaccumulative chemicals is quite slow
because burial in sediments and volatilization into the atmosphere are the
major routes of removal for the chemicals from the ecosystem.  For all these
reasons, EPA concludes that it is reasonable to continue to use a model for
estimating the FCMs which assumes that the system is at steady state.

      EPA agrees that sediment should be considered as a route of exposure in
the model, especially for chemicals with log KOWS  greater than 6.5.   EPA
considers the model by Gobas (1993) an improvement on the 1989 Thomann model
because it incorporates the exposure of organisms to chemicals from the
sediment by including a benthic food-chain component.

      The FCMs for trophic level 3 in some cases are greater than those for
trophic level 4.  Potential causes of the higher concentrations (on a lipid
basis) in the trophic level 3 fish include: 1) growth rates which are much
slower than the predator fishes; 2) differing rates of depuration and
elimination of the chemical by the predator fishes.  Field-measured BAFs
derived from the data of Oliver and Niimi  (1988)  are, in general,  consistent
with the model results for smelt.  For example, DDT, PCB 66, PCB 70+76, PCB
56+60+81, and PCB 49 have field-measured log BAF™s for large smelt  (trophic
level 3) of 7.93,  7.88,  7.71, 8.12, and 7.66 and for piscivorous fishes
(trophic level 4)  of 7.78, 7.79, 7.56, 7.96,  and 7.13, respectively  (Oliver
and Niimi, 1988).

      EPA in developing the final Guidance FCMs (Table B.I) calculated the
trophic level 3 values by averaging  (geometric mean) the individual FCMs for
sculpin and alewife.  The FCMs for smelt were not included in the calculation
of the trophic level 3 values because these organisms are at a trophic level

-------
136    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

higher than 3 but less than 4.  The FCMs for trophic levels 3 and 4  in  the
final Guidance are composed of fish with diets consisting of solely  trophic
level 2 and 3 organisms,  respectively.

      EPA, in part, agrees with the comments that using one set  of modelling
conditions might not be totally representative of the entire Great Lakes
System.  However, numerous similarities do exist among the food  webs in the
five Great Lakes.  First, all of the Great Lakes have both benthic and  pelagic
food web components.  Second, all of the Great Lakes except for  Lake Erie have
salmonids as their piscivorous fish.  Third, all of the Great Lakes  have their
piscivorous fish occupying the fourth trophic level.  Fourth, all of the Great
Lakes have forage fishes occupying the third trophic level.  The food web used
in the development of the FCMs was based upon a four trophic level food web
with both benthic and pelagic food web components taken from Lake Ontario.
EPA has determined that enough similarities exist among the five Great  Lakes
to derive FCMs using one set of modelling conditions taken from  Lake Ontario.

      In selecting a model to use in developing FCMs, EPA did consider
alternative models,'ie.,  Thomann et al.  (1992), before selecting the model of
Gobas  (1993).  The model of Gobas (1993) required the specification  of  fewer
input parameters for benthic food-web components in comparison to the model of
Thomann et al. (1992).  The parameters required by the model of  Thomann et al.
(1992) for the benthic food web are not readily available and would  have
required assumptions or guesses for the appropriate values.  In  contrast, the
model of Gobas required no assumptions or guesses for the input  parameters
used with the benthic food-web components of the model.

      EPA does not agree with commenters suggesting that additional  validation
of the models is needed before use in the final Guidance.  EPA does
acknowledge that a model is not a perfect simulation of what is  occurring in
an aquatic ecosystem.  However, based on the comparison of field-measured BAFs
(from Oliver and Niimi, 1988) to predicted BAFs, the 1993 Gobas  model
acceptably predicts BAFs for the Great Lakes System.

      c.    Final Guidance:  For the reasons cited above, EPA decided to use
the 1993 Gobas model in the development of FCMs to be used in the final
Guidance.

      For chemicals with log KOWS greater than 6.5,  the proposed Guidance
recommended that a FCM of one should be used when no chemical-specific  data
was available.  In the final Guidance, this is no longer necessary because the
Gobas model allows the derivation of  FCMs for the entire range of K<,ws.

      The resulting FCMs for trophic  levels 2, 3, and 4 along with the  input
parameters for the model, are included in the final TSD for BAFs, and in
appendix B of part 132.

5.    Accounting for the Effect of Metabolism in Predicted BAFs

      a.    Proposal:  The proposed Guidance acknowledged that many  organic
chemicals that are taken up by aquatic organisms are transformed to  some
extent by the organism's metabolic processes, and that the rate  of metabolism
varies widely from one chemical to another.  For most, but not all,  organic
chemicals, metabolism  increases the depuration rate, decreases the BAF, and
reduces  the harmful  effects to the organism.

      Because accounting for metabolism  is  difficult, the proposed BAF
methodology included a provision  that allowed for predicted BAFs to  be
modified if justified by the data  (e.g., if  information  showed bioaccumulation
was reduced by metabolism).

       EPA requested  comments on suggested methods to adjust predicted BAFs  for
chemicals that are metabolized; the types of chemicals or chemical groups  for
which  the BAF might  be affected by metabolism; and  an approach to account  for

-------
                           Section IV: Bioaccumulation Factors                       137
metabolism by using an "effective FCM" when a field-measured BAF is not
available but a laboratory-measured BCF is available.

      b.    Comments:  Some commenters argued that the "effective FCM" should
not be used because it did not encourage data generation.  According to the
commenters, adopting a lower "effective KQW"  to back calculate an "effective
FCM" implies that elimination at all trophic levels would be enhanced by '
metabolism.  However, this may not be the case given that lower trophic levels
may lack metabolic capabilities possessed by higher trophic level animals.
Other commenters stated the "effective FCM" was appropriate for conservative
protection of fish consumers.  Several commenters stated that the GLI does not
address the effects of metabolism of pollutants, such as PAHs.  However,
according to the commenters, no adjustment of BAFs for metabolism should be
included in the GLI because no reliable methods are available to predict
potential decreases and increases in toxicity due to metabolism of pollutants.
Other comments on metabolism have been incorporated in section IV.B.2b.

      EPA acknowledges that metabolism is not incorporated in the FCMs or
predicted BCFs and that use of the "effective FCM" has limited applicability
and may generalize the effects of metabolism.  However, by including a BAF
predicted from the BSAF methodology as the second data preference, EPA is
including an additional method for calculating BAFs that accounts for
metabolism.  In addition, since only field-measured BAFs, BAFs derived from
the BSAF methodology, BAFs less than 125 can be used to derive Tier I criteria
for human health and wildlife,  metabolism is either accounted for in these
measurements or cannot substantially reduce the criterion.  Finally, EPA notes
that for a chemical such as aldrin, transformation to dieldrin, does not
reduce the risk of adverse impacts.

      c.    Final Guidance:   Because of the comments and these modifications
to Tier I data requirements, EPA has not required the use of the an "effective
FCM, " but recognizes that it is a valid method that could be used by States or
Tribes to account for metabolism.

6.     Bioavailability

      a.    Proposal:  In the proposed Guidance, the predicted human health
and wildlife BAFs for organic chemicals were based on the total concentration
of the chemical in water and bioavailability was not taken into account.  EPA
acknowledged in the proposal that for chemicals with log KQWS greater than
6.5, a substantial percentage of the total concentration can be associated
with particulate and dissolved organic matter in water and therefore be
unavailable for accumulation in the water column.  EPA requested comment on
deriving BAFs in terms of "freely dissolved"  chemical, (i.e., that which is
dissolved and not associated with other organic matter) to adjust for the
difference in bioavailability between the site water and the water on which
the predicted BAFs were based.

      In a subsequent Notice dated August 30, 1994 (59 FR 44678), EPA
requested comment on an equation which defines the relationship of a BAF
reported on the basis of the total concentration of the chemical in the water
to a BAF reported on the basis of the freely dissolved concentration of the
chemical in the water:

                             BAF; =  (ffd) (BAF/d)

where:

      BAF)   =     BAF (L/kg of lipid) reported on the basis of the lipid-
                  normalized concentration of chemical in the biota (kg/kg
                  lipid)  divided by the total concentration of the chemical in
                  the ambient water  (kg/L);

-------
138    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


      BAF,™ =      BAF  (L/kg of lipid) reported on the basis of the  lipid-
                  normalized concentration of chemical in the biota (kg/kg
                  lipid)  divided by the freely dissolved concentration  of the
                  chemical in the ambient water  (kg/L); and

      ffd    =     fraction of the total chemical in the ambient water that  is
                  freely dissolved.

      The fraction of the chemical in the ambient water that is freely
dissolved, ffd,  can be calculated using the KQW for the chemical and the
concentration of DOC and POC in the ambient
                    f   = 	±_
                    ,fd   .    ( DOC ) ( Kow )
water:                             10
                                               (POC) (Kow)
where:

      POC   =     concentration of particulate organic  carbon,  kg of  organic
                  carbon/L of ambient water;

      DOC   =     concentration of dissolved organic  carbon,  kg of organic
                  carbon/L of ambient water; and

      KOW   =     octanol-water partition  coefficient.

      b.    Comments;..  Some  commenters  stated that  using the  bioavailable
fraction of the chemical in  the ambient water would more accurately reflect
the fraction of the total chemical available to bioaccumulate in the  biota.
Other commenters  stated that it was not possible  to accurately measure the
concentration of  the chemical freely dissolved in water.   Commenters  also
stated that due to a lack of data, the  bioavailable fraction  cannot be
predicted and it  is necessary to use total concentration for  derivation of
BAFs.  Other commenters felt that EPA had  not been  explicit in the definition
of  "concentration in water"  and requested  a more  clearly defined statement.

      Several commenters wanted clarification on  the  suggested POC/DOC values
from Lake Superior and application of the  methodology and an  explanation of
why the baseline  BAFs  were converted to total concentration for the derivation
of criteria.  Other  commenters stated that EPA did  not  evaluate the potential
errors associated with the conversion equations,  variations over time or
throughout a waterbody.

      EPA agrees  with  commenters that taking into account the bioavailable
fraction of the chemical in  the ambient water would more accurately reflect
the fraction of the  total chemical available to bioaccumulate in the biota.
In  the Notice dated August 30, 1994  (59 FR 44678),  EPA  set forth the equation,
fw = 1/(1 + POC • KOW + DOC • Kow/10) , from which  the  fraction of the chemical
that is freely dissolved in  the water can  be calculated using the KQW for the
chemical and the  DOC and POC in  the  ambient water.  EPA acknowledges that the
freely dissolved  concentration of  a  chemical is difficult to  measure, however,
the KOW, DOC and  POC can be  measured or estimated and used to calculate the
freely dissolved  concentration.

      The baseline BAF is based  on the  freely  dissolved concentration of a
chemical, while the  BAF used in  the  derivation of the human health and
wildlife Tier I criteria will  reflect the  total  concentration of the chemical.
In  order to implement  the criteria,  the BAFs need to  be based on a total
concentration of  the chemical  in  the water column because analytical methods
in  40 CFR part  136  that are  used for compliance monitoring determine the total
amount of chemical  in  the water.   The  concentration of  POC and DOC estimated
from Lake Superior  from Eadie  et  al.  (1990) will  be used to calculate BAFs

-------
                           Section IV: Bioaccumulation Factors                       139
based on total concentration for derivation of human health Tier I criteria
and Tier II values and wildlife Tier I 'criteria.  EPA believes that the values
for POC and DOC are.protective of the entire Great Lakes System.  Other values
for POC and DOC may be used to derive site-specific criteria if scientifically
justified.

      Concentrations of POC and DOC are those discussed in the August 30, 1994
Notice and are specified  (see appendix B to part 132) and are accounted for in
the criteria.  The BAFs based on the concentration of the freely dissolved
chemical in the water when used properly will provide the same predicted
residue in aquatic organisms as the BAFs based on the concentration of the
total chemical in water.

      The calculation of the freely dissolved concentration of a chemical in
the water column assumes equilibrium conditions, and therefore variations over
time and throughout a waterbody are negligible.  In EPA's judgement, the
errors associated with the conversion equations are minimal in comparison to
the benefit of normalizing the site-specific parameters of POC and DOC in
calculation of the BAF.

      c.    Final Guidance:  EPA has decided to use the freely dissolved
concentration of organic chemicals in the derivation of baseline BAFs and the
total concentration of the chemical for derivation of Tier I human health and
wildlife criteria. tThe fraction of the chemical in the ambient water that is
freely dissolved, fH,  will be calculated using the KOW for the  chemical  and
the concentration of DOC and POC in the ambient water.  For further details
concerning this equation, see the final TSD for BAFs which is available in the
public docket for this rulemaking.  Basing the measured and predicted baseline
BAFs on the concentration of the freely dissolved chemical in water permits
the derivation of generic BAFs devoid of site-specific influences and
considerations, such as varying concentrations of POC and DOC and allows
consistent usage and derivation of the BAFs throughout the final Guidance.

7.    Calculation of Baseline BAFs

      a.    Proposal:   In the Technical Support Document for the Procedure to
Determine Bioaccumulation Factors - July 1994  (EPA-822-R-94-002) which
accompanied the August 30, 1994, Notice of Data Availability (58 FR 44678),
equations were set forth for calculating baseline BAFs.  Slight modifications
were made to the equations in order to correct for the errors in the original
equations.

      b.    Final Guidance:  In the final Guidance, a baseline BAF shall be
calculated from a field-measured BAF of acceptable quality using the following
equation:
                Baseline BAF =
                                   Measured BAFj
                                         ffd
- l
      where:

          BAF|   =  BAF based on total concentration in tissue and water.

          f,     =  fraction of the tissue that is lipid.

          ffd     =  fraction of the total chemical that is freely dissolved in
                   the ambient water.

      The trophic level to which the baseline BAF applies is the same as the
trophic level of the organisms used in the determination of the field-measured
BAF.

-------
140    Water Quality Guidance for the Great Lakes System — Supplementary Information Document
      A baseline BAF for organic chemical "i" shall be calculated from a
field-measured BSAF. of acceptable quality using the following equation:

              (BaSeUneBM,,.(BAF,,.
      where :
        ( BAFffd )
      (BSAF)i

      (BSAF)
                           BAF based on the measurement of freely dissolved

                           reference chemical in the water column.

                           BSAF for chemical "i".

                           BSAF for the reference chemical "r" .

                           octanol -water partition coefficient for chemical
      (KoW)r
                           octanol -water partition coefficient for the
                           reference chemical "r".
      The trophic level to which the baseline BAF applies is the same as the
trophic level of the organisms used in the determination of the BSAF.

      A baseline BAF for trophic level 3 and a baseline BAF for trophic level
4 shall be calculated from a laboratory-measured BCF of acceptable quality and
a FCM using the following equation:
             Baseline BAF  =  (FCM)
                                      Measured
                                                - 1
                                             -fd
where :

    BCF!

    f,

    fa


    FCM
                   BCF based on total concentration in tissue and water.

                   fraction of the tissue that is lipid.

                   fraction of the total chemical that is freely dissolved in
                   the ambient water.

                   the food- chain multiplier obtained from Table B-l by
                   linear interpolation for trophic level 3 or 4, as
                   necessary .
      A baseline BAF for trophic level 3 and a baseline BAF for trophic level
4 shall be calculated from a KQW of acceptable quality and a FCM using the
following equation:

          Baseline BAF =  (FCM)  (predicted BCF)  =  (FCM) (Kow)

      where :

            FCM    =  the food-chain multiplier obtained from Table B-l by
                      linear interpolation for trophic level 3 or 4,  as
                      necessary.
                   =  octanol-water partition coefficient.

-------
                           Section IV: Bioaccumulation Factors                       141
8.    Other Uses of BAFs

      In the final Guidance, human health BAFs are used to identify chemicals
of greatest concern within the Great Lakes System.  Chemicals identified as
BCCs are those for which extra controls are required as specified  in the final
implementation procedures and under the antidegradation procedures in the
final Guidance.  See discussion of BCCs in section II.G of this document.

9.    Individual BAFs

      a.    2.3.7.8-TCDD

      i.    Proposal:  In the proposed Guidance, the BAF^, for TCDD for
derivation of wildlife criteria was 79,000 for trophic level 3 and trophic
level 4.  The BAFf,*, in the August 30, 1994 Notice of Data Availability  (59
FR 44678) was 320,000 for trophic level 3 and 160,000 for trophic level 4.
The BAFy0%/ for TCDD for derivation of human health criteria in the proposed
Guidance was 50,000 for trophic level 4.  The BAFj0%, in the August 30, 1994
Notice of Data Availability  (59 FR 44678) was 101,000.  Differences between
the proposed values and those in the Notice were based on differences in the
type of data used. The proposed value is based on a laboratory-measured BCF,
while the value in the Notice is based on the BSAF methodology.

      ii.   Comments:  Several commenters stated that the documentation for
the proposed BAF for 2,3,7,8-TCDD was inadequate and that all relevant studies
must be cited and available for review.  Many commenters stated that the
proposal was inconsistent with the Interim Report on Data and Methods for
Assessment of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Risks to Aquatic Life and
Associated Wildlife (EPA, 1993) and argued that the inconsistency was due to
the fact the EPA was not being forthright with all the available information.

      Commenters stated that the basis for the trophic level 3 BAF for
2,3,7,8-TCDD is unexplained and that the scientific basis for what has been
done should be made, available prior to publication of the final Guidance.

      EPA has cited and made available all studies used in the derivation of
the BAF for 2,3,7,8-TCDD.  The data for the basis of the trophic level 3 BAF
was made available in the BSAF section of the Technical Support Document (see
EPA-822-R-94-002).  Any perception of inconsistency between the Interim Report
and the proposal may be due to the way the data are used and the evolution of
the BSAF method after the Interim Report was completed.  The BSAF method does
not require estimation of the concentration of 2,3,7,8-TCDD in water to
determine the BAF as was done in the Interim Report on Data and Methods for
Assessment of 2,3,7,8-Tetrachlorodibenzo-p-dioxin Risks to Aquatic Life and
Associated Wildlife- (EPA, 1993) .

      The BAF for 2,3,7,8-TCDD is greater for trophic level 3 fish than
trophic level 4 fish because the Lake Ontario BSAF data for smelt and trout
indicated this difference.  This may be attributable to growth dilution at
trophic level 4  (i.e., with trout).

      iii.  Final Guidance:  For 2,3,7,8-TCDD, the baseline BAF will be based
on a predicted BAF derived from the BSAF methodology; the baseline BAF for
trophic level 3 is calculated using the ratios of FCMs for log KQW = 7.02.
Both baseline BAFs pake into account the bioavailability of 2,3,7,8-TCDD in
the water column and is lipid normalized.  The baseline BAFs for 2,3,7,8-TCDD
for trophic level 3 and trophic level 4 are 9,360,000 and 9,000,000
respectively.  The BAFs for derivation of Tier I human health criteria for
2,3,7,8-TCDD are 48,490 for trophic level 3 and 79,420 for trophic level 4,
based on the total concentration of the chemical in the water column and
assuming 1.82 and 3.10 percent lipid, respectively.  The BAFs for derivation
of Tier I wildlife criteria for 2,3,7,8-TCDD are 172,100 for trophic level 3
and 264,100 for trophic level 4,  based on the total concentration of the

-------
142    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

chemical in the water column and assuming 6.46 and 10.31 percent  lipid,
respectively.

      b.    PCBs

      i.    Proposal:  The BAFj.,,, for PCB in the proposed Guidance was
1,000,000 for trophic level 3 and 2,800,000 for trophic level 4.  The
in the August 30, 1994 Notice of Data Availability  (59 FR 44678)  was 658,000
for trophic level 3 and 687,000 for trophic level 4.  The BAF15.0%,  for PCB  for
derivation of human health criteria in the proposed Guidance was  1,776,860  for
trophic level 4.  The BAFj0%< in the August 30, 1994 Notice of Data
Availability  (59 FR 44678) was 435,000 for trophic level 4.

      ii.   Comments:  Commenters were concerned that the apparent
disequilibrium between the sediment, water column and fish tissue
concentration makes predictions of BAFs difficult.  In place of the BAF,
commenters suggested use of the BSAF.

      Other commenters stated that the one BAF for all PCBs was not
appropriate but instead BAFs should be congener specific because  of varying
bioavailability and rates of uptake and metabolism.  One commenter suggested
developing BAFs for lower chlorinated and higher chlorinated PCBs.  Commenters
felt that development of a BAF based on the "most prevalent" congeners  did  not
satisfactorily address the problem that there is no single BAF for all  PCBs.
According to the commenters, data show that BAFs for different congeners  vary
widely.  Other commenters were concerned that the BAF in the August 30, 1994
Notice of Data Availability  (59 FR 44678) for trophic level 4 would decrease
the BAF four-fold and therefore result in a more lenient water quality
criteria and believe that EPA has not justified this proposed relaxation  in
PCB BAFs.

      Commenters criticized the Oliver and Niimi  (1988) data set  for PCBs
because of temporal and spatial variability in the data, use of unadjusted
whole fish values, and use of centrifuged water concentrations.

      EPA recommends the use of the BSAF, as detailed in section  B.2.c  above,
if a field-measured BAF is not available.  For PCBs, however, field-measured
BAFs are available for most congeners and therefore will be used  in the
derivation of the baseline BAF.  Field-measured BAFs inherently account for
the effects caused by a) metabolism, b) the disequilibrium of the chemical
between the sediment and the water column, and c) all the other naturally
occurring environmental processes.

      EPA has revised the calculation of the BAF for PCBs based on comments
that use of the most prevalent congeners in Schultz et al.  (1989) is
inappropriate.  In the  calculation of the revised BAF for PCBs, the BAFs  for
the individual congeners were assigned weights based on their concentration in
salmonids as reported by Oliver and Niimi  (1988), as suggested by a commenter.
The revised BAF is based on the concentrations of PCBs in salmonids in  the
Great Lakes and not  the concentration of PCBs in the water column since fish
are the predominant  route of exposure to humans and wildlife.  The BAF  is
based on field-measurements which inherently account for metabolism,
bioavailability, growth, concentration in the sediment and rate of uptake.  It
is necessary to calculate a single BAF in order to derive Tier I  human  health
and wildlife criteria because the corresponding toxicological data are  only
available for Aroclor 1260, a mixture of congeners.  Therefore development  of
a BAF for lower- anfl higher-chlorinated PCBs is not appropriate.

      The BAF for PCBs  was derived using field-measured data for  the congeners
based on their weighted concentration in salmonids.  The BAFs take into
account the bioavailability of the chemical in the water column and are lipid
normalized.   In derivation of the final BAF, the baseline BAF was converted to
a BAF based on the total  concentration of PCBs in the water column.

-------
                           Section IV: Bioaccumulation Factors                       143
      EPA acknowledges the commenters' concerns that the trophic level 4 BAF
decreased four-fold in the August 30, 1994 Notice of Data Availability  (58 FR
44678) as compared to the proposed value.  EPA has recalculated the BAF, as
described in the above paragraph, which resulted in a baseline BAF value of
116,600,000 for trophic level 4.

      EPA lipid normalized the fish tissue concentrations given by Oliver and
Niimi (1988) in order to avoid use of unadjusted whole fish values, and
calculated the freely dissolved concentration of a chemical in the water
column to account for the centrifuged water concentrations.  EPA has
determined that if these parameters are accounted for, much of the variability
is eliminated.

      iii.  Final Guidance:  The baseline BAF for PCBs is based on field-
measured BAFs for the congeners of PCBs based on their concentration in fish
and thus takes into account the freely dissolved concentration of PCBs in the
water column and lipid normalization.  The baseline BAFs for trophic level 3
and trophic level 4 are 55,280,000 and 116,600,000, respectively.  The BAFs
for derivation of Tier I human health criteria for PCBs are 520,900 for
trophic level 3 and 1,871,000 for trophic level 4, based on the total
concentration of the chemical in the water column and assuming 1.82 and 3.10
percent lipid, respectively.  The BAFs for derivation of Tier I wildlife
criteria for PCBs are 1,850,000 for trophic level 3 and 6,224,000 for trophic
level 4, based on the total concentration of the chemical in the water column
and assuming 6.46 and 10.31 percent lipid, respectively.

      c.    Mercury

      i.    Proposal:  The proposed mercury BAF was 60,000 for trophic level 3
and 130,000 for trophic level 4.  Subsequent to the proposal, EPA asked for
comment on a recalculated mercury BAF in the August 30, 1994 Notice of Data
Availability  (59 FR 44678) .  The recalculated BAF was 28,000 for trophic level
3 and 140,000 for trophic level 4.  The derivation of the recalculated BAF
differed from the proposed BAF for mercury in the following ways (see EPA-822-
R-94-002 for discussion of the data used):

      1)    The estimated percent of total mercury in water that is
methylmercury was changed from 25 percent to 17 percent.

      2)     The estimated biomagnification factors (BMFs) were changed as
follows:

      Trophic Level           New BMF           Old BMF
      2 to 3                  1.26              2.154
      3 to 4                  5.0               2.154

      3)     The estimated percent of the total mercury in fish that is
methylmercury was changed from 85.3 percent to 97.5 percent.

      ii.    Comments:  Many commenters stated that the derivation of the BAF
for mercury is not scientifically valid because the assumptions regarding the
bioavailability and fate of mercury in the water column were conservative and
EPA had not accurately assessed the sequestration of mercury in the
environment, especially in sediment.  Other commenters advocated site-specific
modifications for mercury, stating that a single BAF is oversimplistic and
ignores variations.

      Commenters stated that the mercury BAF in the August 30, 1994 Notice of
Data Availability (59 FR 44678)  did not utilize available field data
submitted.

      In development of a scientifically-defensible mercury criteria,
commenters stated that EPA must consider that fish accumulate the majority of
mercury via ingestion of contaminated sediments and food, not by gill uptake

-------
144    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


of dissolved mercury.  Commenters recommended using the BSAF method to
recalculate BAF for mercury.

      EPA does not agree that the BAF for mercury is not scientifically valid.
The assumptions used by EPA regarding the bioavailability and fate, with one
exception described below, were based on the best available data including  •
that submitted by commenters.  After review of the data, EPA revised the-
percent of total mercury which is found as methylmercury from 25 percent to 17
percent based on data of Gill and Bruland  (1990).

      EPA also does not agree that it did not accurately assess the
sequestration of mercury in the environment.  Sequestration of an inorganic
chemical in sediment is generally not important because BAFs are derived based
on the relationship between fish tissue concentrations and the water column.
For chemicals like mercury, methylation appears to occur in the sediment which
releases methylmercury into the water column.  This is different from the
transfer through the benthic food-chain component that is part of the Gobas
model used for organic chemicals.  If partitioning to the water column  occurs,
BAFs based on the bioavailable portion would be higher than the BAFs in the
final Guidance which are based on total.

      EPA agrees with the commenters that exposure of aquatic organisms to
mercury is primarily through food and sediment and has developed a
biomagnification factor for mercury based on field data.  This was necessary
because the 1993 Gobas model is only applicable to organic chemicals.   The
BSAF method as described in the final Guidance is also applicable to only
organic chemicals, and therefore cannot be used in the derivation of the BAF
for mercury.  Mercury's BAF is derived based on the percentage of methyl
mercury and amount of trophic uptake as explained in appendix D of the  July
1994 TSD for BAFs  (EPA-822-R-94-002).  EPA has provided the rationale for the
derivation of the BAF in the final TSD for BAFs which is available in the
public docket for this rulemaking.

      EPA has completed a more comprehensive national analysis of the data
concerning the bioaccumulation of mercury by fish, which is being peer
reviewed at this time.  EPA had intended to use in the final Guidance the
baseline BAFs contained in the initial draft of the report but EPA decided to
wait until the report has been peer-reviewed and completed.  EPA is planning
to issue a proposal* to revise the mercury criterion in the final Guidance to
reflect this new information.

      iii.  Final Guidance:  The baseline BAF for mercury is based on a
laboratory-measured BCF and a biomagnification factor.  The baseline BAFs for
both wildlife criteria and human health criteria and values for trophic level
3 are 27,900 and for trophic level 4 are 140,000.

C.    Conformance to the CWA. Great Lakes Water Quality Agreement and Great
      Lakes Critical Programs Act of  1990

      Section 118(c) of the CWA requires EPA to develop, "inter alia,"
guidance on minimum water quality limits to protect human health, aquatic life
and wildlife in  the Great Lakes System.  The Great Lakes Critical Programs Act
of 1990  (CPA) states that the final Guidance shall be no less restrictive than
the provisions of the CWA, National water quality criteria and National
guidance, and shall conform with  the  objectives and provisions of the Great
Lakes Water Quality Agreement  (GLWQA).  For reasons set out in the preamble to
the final Guidance  (see 58 FR 20858), EPA has determined that the final
Guidance meets these requirements.

D.    Adoption of Water Quality Standards Consistent with the Final Guidance

      The final  Guidance  for deriving BAFs  is included in appendix B to part
132.  The BAF TSD, which  discusses the basis for the final methodology  and
which sets  forth the data and considerations upon which the individual  BAFs

-------
                           Section IV: Bioaccumulation Factors                       145
are based, is available in the docket for this rulemaking.  Copies are also
available upon written request as described in section XIII of this document.

E.    References

      ASTM. 1990. Standard Practice for Conducting Bioconcentration Tests with
Fishes and Saltwater Bivalve Molluscs.  Designation E 1022 - 84.  Pages 606-
622.  In Annual Book of ASTM Standards.  Section 11, Water and Environmental
Technology, Volume 11.04.  American Society for Testing and Materials.  1916
Race Street, Phila., PA 19103.

      De Vault, D.S. 1993. Data on Contaminant Trends in Lake Trout
(Unpublished). Contains data for 1984-1990.

      Eadie, B.J., N.R. Morehead, and P.P. Landrum.  1990. "Three-phase
partitioning of hydrophobic organic compounds in Great Lakes waters."
Chemosphere, 20, 161-178.

      Gill, G.A., and K.W. Bruland.  1990.  Mercury Speciation in Surface
Freshwater Systems in California and Other Areas.  Environ. Sci. Technol.
24:1392-1400.

      Gobas, F.A.P.C.  1993, "A model for predicting the bioaccumulation of
hydrophobic organic chemicals in aquatic food-webs: application to Lake
Ontario."  Ecological Modelling, 69, 1-17.

      Hamelink, J.L., R.C. Waybrant and R.C. Ball. 1971.  A proposal: exchange
equilibria control the degree chlorinated hydrocarbons are biologically
magnified in lentic environments. Trans. Amer. Fish. Soc. 100: 207-214.

      Mackay, D. 1982.  Correlation of bioconcentration factors.  Environ.
Sci. Technol. 16:274-278.

      Niimi, A.J. 1985.  Use of laboratory studies in assessing the behavior
of contaminants in fish inhabiting natural ecosystems.  Water Poll. Res. J.
Canada 20:79-88.

      Oliver, B.C. and A.J. Niimi. 1983.  Bioconcentration of chlorobenzenes
from water by rainbow trout: correlations with partition coefficients and
environmental residues.  Environ. Sci. Technol. 17:287-291.

      Oliver, E.G. and A.J. Niimi. 1988.  Trophodynamic analysis of
polychlorinated biphenyl congeners and other chlorinated hydrocarbons in the
lake Ontario ecosystem.  Environ. Sci. Technol. 22:388-397.

      Schultz, D.E., G. Petrick, and J.C. Duinker, 1989.  Complete
Characterization of Polychlorinated Biphenyl Congeners in Commercial Aroclor
and Clophen Mixtures by Multidimensional Gas Chromotography-Electron Capture
Detection.  Environ. Sci. Technol. 23: 853-859.

      Swackhammer, p.L. and R.A. Kites. 1988.  Occurrence and bioaccumulation
of organochlorine compounds in fishes from Siskwit Lake, Isle Royale, Lake
Superior.  Environ. Sci. Technol. 22:543-548.

      Thomann, R.V. 1989.  Bioaccumulation Model of Organic Chemical
Distribution in Aquatic Food Chains.  Environ. Sci. Technol. 23:699-707.

      Thomann, R.V. and J.P. Connolly. 1984. Model of PCB in the Lake Michigan
lake trout food chain. Environ. Sci. Technol. 18: 65-71.

      Thomann, R.V., J.P. Connolly, T.F. Parkerton. 1992. An Equilibrium Model
of Organic Chemical-Accumulation in Aquatic Foodwebs with Sediment
Interaction.  Environ. Toxicol. Chem. 11: 615-629.

-------
146    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


      U.S. EPA. 1980. Seafood consumption data analysis.  Stanford Research
Institute International, Menlo Park.  California.  Final Report.  Task 11,
Contract No. 68-01-3887.

      U.S. EPA guidance contained  in  Stephan et al.  (1985)  "Guidelines for
Deriving Numerical National Water  Quality Criteria for the  Protection of
Aquatic Organisms and Their Uses."  NTIS  # PB85-227049.  U.S.  Department of
Commerce, 5285 Port Royal Road,  Springfield,  VA 22161.

      U.S. EPA. 1993. Interim Report  on  Data and Methods for Assessment of
2,3,7,8-Tetrachlordibenzo-p-dioxin Risks to  Aquatic Life and Associated
Wildlife. EPA/600/R-93/055.

      Veith, G.D. and P. Kosian. 1983.   Estimating Bioconcentration Potential
from Octanol/Water Partition  Coefficients.   Chapter 15 in PCBs  in the Great
Lakes.  Mackay, D., R.  Patterson,  S.  Eisenreich,  and M.  Simmons (eds.) Ann
Arbor Science.

      West, P.C. et al. 1989.   Michigan  Sport Angler Fish Consumption Survey:
A Report to the Michigan Toxic Substance Control Committee, University of
Michigan Natural resources Sociology  Research Lab.  Technical Report #1, Ann
Arbor, Michigan, MDMB Contract # 87-20141.

      Zipf, G.W..  Memorandum to Michael Cox.  1995.

-------
                              Section V: Human Health                           147
                      V.   HUMAN  HEALTH
A.    Summary of Final Rule

      The methodology for the development of human health criteria and values,
as outlined in appendix C of part 132,  is identical to the Guidance proposed
on April 16, 1993 except for the following changes:

      -- When deviations from values presented in the Integrated Risk
Information System (IRIS)  are anticipated or considered necessary,  it is
strongly recommended that such actions be communicated to the EPA Reference
Dose  (RjD) and/or Cancer Risk Assessment Verification Endeavor (CRAVE) work
group immediately.

      - - To determine the weight of evidence of carcinogenicity  for a
chemical, and to determine (on a case-by-case basis)  whether a Group C
chemical should be a Tier I criterion or Tier II value the following data (if
available)  shall be considered:  mutagenicity/genotoxicity (determinations  of
whether the chemical interacts directly with DNA); structure activity;
metabolism and mode of action.

      -- Clarification has been added to the discussion on additional
uncertainty factors which may be applied to short-term (28-day)  study results
used in the development of Tier II values.  In some cases, an uncertainty
factor, from 1-10,  may be needed to extrapolate from 28-day study results to
subchronic  (90-day) results.   This decision must be made on a case-by-case
basis.

      - - If the duration of a cancer bioassay is significantly less than  the
natural lifespan of the test animal, the slope may be adjusted on a
case-by-case basis to compensate for latent tumors which were not expressed
during the cancer bioassay.  This is a change from the proposal  which required
an adjustment to compensate for latent tumors if the cancer bioassays were
shorter than 78 weeks for mice and 90 weeks for rats.

      -- A default relative source contribution factor (RSC)  of  0.8 will  be
applied to all noncarcinogenic chemicals.  In the proposed Guidance,  an RSC of
0.8 was applied to noncarcinogenic bioaccumulative chemicals of  concern (BCCs)
and an RSC of 100 percent was applied to noncarcinogenic nonbioaccumulatives.
If actual exposure data exists, States and Tribes may use this data,  following
procedures outlined in the 1980 National Guidelines to calculate an RSC.

      -- A fish consumption rate of 15 grams will be applied in  the derivation
of criteria (3.6 grams for trophic level 3 fish and 11.4 grams for trophic
level 4 fish).
      - - The minimum data requirements for deriving a Tier I criterion have
been changed to include minimum data requirements for bioaccumulation factors
as well as minimum toxicological data.   For organic chemicals, field-measured
bioaccumulation factors (BAFs), predicted BAFs based on the biota-sediment
accumulation factor (BSAF), and BAFs less than 125 will be used  in calculating
Tier I criteria.  For inorganic chemicals, either a field-measured BAF or a
laboratory-measured BCF can be used to calculate Tier I criteria.

B.    Explanation of Final Provisions

      The final Guidance for human health is described below. As with the
final Guidance for aquatic life discussed earlier, EPA is proposing a two-tier
approach for human health.  Minimum data requirements for Tier I criteria and
Tier II values are discussed later in this section.

-------
148    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      Sample human health criteria have been calculated using the proposed
methodology for 18 chemicals.  The 18 chemicals chosen for criteria
development were selected from the Great Lakes Water Quality Initiative
(GLWQI) group of chemicals of concern listed in proposed 40 CFR part 132,
Table 6, to represent a broad cross- section of the types of chemicals  found
in the Great Lakes basin.  The intent of the chemical selection was to  test
the final methodology against a broad range of chemicals and to demonstrate
how the criteria development process will be carried out.  Selection from
among the chemicals of concern was made from the perspective of demonstrating
the final methodologies' applicability to all types of chemicals, rather than
on the basis of health risk priorities.

C.    Criteria Methodologies

      The  final Guidance establishes methodologies to derive human health
criteria which will not result in zero risk, but will provide a level of
protection likely to be without appreciable risk.

1.    Endpoints Addressed

      a.    Proposal:  The proposed Guidance addressed noncancer and cancer
effects only.  Organoleptic effects  (taste and odor), while of concern  from an
aesthetic standpoint, were not considered a significant health concern  and
therefore were not included in the proposal.  EPA asked for comment on  whether
EPA should require the Great Lakes States and Tribes to adopt Tier I criteria
identical to the existing National guidance for organoleptic substances
developed under section 304(a) of the Clean Water Act  (CWA) .

      b.    Comments:  Several commenters were in favor of the EPA proposal to
focus only on cancer and noncancer criteria development.  Several other
commenters disagreed, stating that existing National guidance for organoleptic
criteria should be required to be adopted by the States.  If organoleptic
criteria are not in place, these commenters argued, people may turn to  other
sources of drinking water  (e.g., bottled) or may turn away from Great Lakes
fish  (if they are tainted) in favor of other protein sources.  The commenters
also  stated such alternatives may have negative health effects: the quality of
bottled drinking water may be unknown and the health implications of eating
other  sources of pr&tein  (or sources of food) such as beef, chicken, pork etc.
may reduce one's exposure to bioaccumulated chemicals but may increase  one's
exposure to higher cholesterol.

       c.    Final Guidance:  The final Guidance will continue to focus  on
noncancer and cancer effects and will not require the adoption of Tier  I
organoleptic criteria.  While it is conceivable that some people may seek
alternative water and food sources due to organoleptic properties of chemicals
and such choices may themselves entail a risk of incurring adverse health
effects, EPA believes that the human health protection goals reflected  in the
CWA are best served.by  focusing on actual, health-related effects  (cancer and
noncancer effects) due  to the exposure to the aquatic resource itself.
Moreover, the current National criteria guidance developed for organoleptic
effects are available for use by Great Lakes States and Tribes in developing
criteria.   If States or Tribes want to set  criteria based on more stringent
organoleptic criteria,  as set forth in the  1980 Federal Register Notice of
Water Quality Criteria Documents  (U.S. EPA,1980), they are free  to do so..

2.     Mechanism  of Action

       a.    Cancer.

       i.    Proposal:   In the proposal, EPA regarded   carcinogenicity
generally as a non-threshold adverse  health effect.  Given  that  assumption,
"no effect" levels for  carcinogens  (other than  zero) were not established,
because even extremely  small doses are assumed  to potentially elicit a  finite
increase in the  incidence of cancer.  The proposed Guidance would have

-------
                              Section V: Human Health                           149
provided that a non-threshold mechanism be assumed for carcinogens unless data
exist that demonstrate a threshold mechanism.  Such a finding would have to
weigh all the studies for a particular chemical to determine whether a true
threshold effect is occurring across all sensitive test species.  The proposal
also recommended that States and Tribes confer closely with EPA prior to
submitting for EPA approval any criterion for a carcinogen that is based on
the assumption that a "safe" threshold dose exists for the chemical.  The EPA
proposal also stated that the Linearized Multistage Model  ("LMS") would be
used to extrapolate from actual animal bioassay data to the dose/response
relationship expected at low doses, unless it can be established on a
case-by-case basis that another model such as the "time-to-tumor" model or
ones based on modifications of the LMS model, are more appropriate.

      ii.   Comments:  In response to the EPA-held assumption that
carcinogenisis is a nonthreshold event, several commenters stated that EPA
relies too heavily on the nonthreshold model of carcinogenicity and rarely
factors into the weight of evidence such information as pharmacokinetic data,
genotoxicity and mutagenicity data, structure activity and mode of action.
With regard to the use of the LMS as a default cancer model, several
commenters supported the continued use of the LMS as a default cancer potency
model.  Other commenters stated that the use of the LMS model is outdated and
does not consider all the available mechanistic, pharmacokinetic and other
relevant data in assessing cancer potency.

      iii.  Final Guidance:  Based on the rationale below, EPA will continue
to include the use of the LMS model as a default model in the final Guidance,
unless it can be established on a case-by-case basis that another model is
more appropriate.  As the 1986 Guidelines for Carcinogen Risk Assessment
states:  When pharmacokinetic or metabolic data are available, or when other
substantial evidence on the mechanistic aspects of the carcinogenesis process
exist, a low dose extrapolation model other than the linearized multistage
procedure might be considered more appropriate on biological grounds.  When a
different model is chosen, the risk assessment should clearly discuss the
nature and weight of evidence that lead to the choice (U.S. EPA, 1986).

      EPA believes the nonthreshold assumption of carcinogenesis is a valid
assumption in the absence of data which indicate otherwise.  EPA also
reiterates that the methodology provides an opportunity to demonstrate a
threshold mechanism for a chemical and that on a case-by-case basis another
model (threshold or'non-threshold) can be applied to the cancer data if it can
be established that the other model is more appropriate.

      EPA disagrees with the comment that it does not consider all the data in
making judgements on potential carcinogens.  EPA considers all the available
data when conducting a cancer assessment for a chemical, including
pharmacokinetic, metabolic, mutagenicity/genotoxic and structure activity
data.  To ensure that this occurs with regard to development of criteria in
the future,  EPA has revised the human health methodology in appendix C to part
132 to provide for consideration of each of these specific pieces of evidence
when evaluating the-potential carcinogenic risk of a chemical.

      with regard to the choice of a cancer model, and specifically the
appropriateness of continuing reliance on the LMS, EPA believes that, in the
absence of adequate information to the contrary (such as information on the
mechanism of carcinogenic action), the LMS is the best of the mathematical
extrapolation models used for extrapolating from high dose to low dose.  As
stated in the 1986 Guidelines for Carcinogenic Risk Assessment:  When data and
information are limited,  and when much uncertainty exists regarding the
mechanism of carcinogenic action,  models or procedures which incorporate low
dose linearity are preferred when compatible with the limited information.  In
the absence of adequate information to the contrary, the linearized multistage
procedure will be employed (U.S. EPA, 1986).  Because of the uncertainties
associated with dose response, animal to human extrapolation, and the serious
public health consequences that could result if risk were under-estimated, EPA

-------
150    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


believes that it is prudent and consistent with public health goals of the CWA
to use the LMS to estimate cancer risk for ambient water quality criteria.
The LMS has been endorsed by four agencies in the Inter-Agency Regulatory
Liaison Group and was characterized as less likely to under-estimate risk at
the low doses typical of environmental exposure than other models that could
be used (Inter-Agency Regulatory Liaison Group, 1979) .

      b.    Noncancer.

      i.    Proposal:  The proposed Guidance would have provided that a
threshold mechanism of action be assumed in deriving criteria for protection
against noncancer effects, unless it is demonstrated on a case-by-case basis
that there is no threshold with respect to a given chemical's toxicity
effect(s).  This means there is a dose below which no-adverse effects should
be observed, or if an adverse effect is observed, the risk of deleterious
effect over the span of a lifetime is not appreciable.  EPA also recognized
that there may be exceptions to this principle:  for some non-carcinogenic
effects, no identifiable threshold of effects has been demonstrated.
Chemicals which may exert non-threshold non-cancer effects include genotoxic
teratogens and germline mutagens.

      ii.    Comments:  Several commenters agreed that EPA should allow for a
showing that a chemical has no threshold; however these same commenters did
not express an opinion regarding the default assumption of a threshold for
noncarcinogens.   One commenter believed all noncancer effects should be
considered nonthreshold events unless proven otherwise.  Other commenters
believed making such a showing would be difficult and possibly very costly.
Some commenters suggested that EPA apply a "weight of evidence" approach to
developing an RfD; that we not ignore studies showing no adverse effects while
focussing only on studies indicating adverse effects.

      iii.  Final Guidance:  For the reasons stated below, the final Guidance
will continue to assume that noncancer endpoints of toxicity exhibit a
threshold unless data indicate otherwise.  This is consistent with the 1980
National Guidelines which concluded for noncarcinogens that there is a dose
below which no adverse effects should be observed.  It is also EPA's
experience  (in developing RfDs) that noncancer effects are regarded almost
without exception as threshold events and that nonthreshold noncancer effects
are extremely rare and in some cases may be perceived as such due to a lack of
toxicological information  (e.g., extremely low doses have not been tested or
cannot be measured).  The possible exceptions to this rule, as stated in the
proposal,  are genotoxic teratogens and germline mutagens.  EPA reiterates that
in the rare instances that this type of chemical is encountered, it is
recommended that States and Tribes confer closely with EPA prior to
establishing a noncancer criterion on the basis of a non-threshold effect.
EPA believes the proposed language is adequate in allowing a demonstration of
nonthreshold noncarcinogenicity and is adopting this language in the final
methodology.  EPA has provided additional guidance in "Great Lakes Water
Quality Initiative Technical Support Document for Human Health Criteria and
Values" (EPA 820-B-95-007)  (Human Health TSD) on determining whether a
chemical demonstrates nonthreshold characteristics.

      With regard to the comment that EPA treat all chemicals as if they are
nonthreshold chemicals, EPA believes the overwhelming scientific evidence
supports the assumption that noncancer endpoints are threshold events.  EPA
most often assumes that noncarcinogenic and/or nonmutagenic changes have a
threshold, that  is, a. dose level below which a response is unlikely, because
homeostatic, compensating and adaptive mechanisms in the cell protect against
toxic effects at levels below this threshold.  Therefore, in EPA's judgement
it would be inappropriate to assume that noncarcinogens act by nonthreshold
mechanisms, when the overwhelming evidence suggests otherwise  (U.S. EPA,
1991) .

-------
                              Section V: Human Health                            151
      With regard to the comment that EPA should apply a weight-of-evidence
approach in developing noncancer criteria and values, EPA believes  it  does
employ a weight-of-evidence approach in that it evaluates all the data before
choosing a specific study upon which quantification  is based.  EPA  believes
its process for evaluating noncancer effects is similar to the cancer  weight -
of-evidence approach.  For example, in developing a  noncancer criterion, EPA
reviews all the toxicological effects data and determines which  study  appears
to reflect the most critical endpoint.  Negative studies  (i.e.,  those  that do
not elicit a responee) as well as positive studies  (i.e.,.those  that do elicit
a response) are considered.  IRIS coversheets describe the critical study
which serves as the basis for an RfD but also lists  all the critical and
supporting studies that were considered in the development of the RfD.  EPA
agrees that the main focus when evaluating the potential adverse effects of a
chemical are on studies that show adverse effects.   However, EPA believes this
is reasonable to ensure that humans are protected against potential adverse
effects.  EPA will continue to focus  on studies which are most  relevant to
the consideration of human risk assessment and with  the development of more
pharmacokinetic data, which can clearly identify metabolic/toxicokinetic
differences between species, human risk assessments  will become  easier.

3.    Choice of Risk Level

      a.    Proposal:  The proposal derived criteria which correspond  to a
plausible upper bound increased incremental risk of  developing cancer  of one
in 100,000 (10"5)  over a lifetime of exposure.  The choice of 10'5 risk  level
was recommended by the Initiative Committees and is  within a range  of  risk
levels  (i.e., 10"4 to 10"*)  that EPA considers to be adequately protective and
which EPA has historically considered acceptable in  making regulatory
decisions.  The majority of the Great Lakes States traditionally have  used a
10"5 risk level in setting their water quality criteria.  EPA asked  for comment
on the choice of risk level, and on alternate risk levels, such  as  10"* and
10"",  which could be adopted in the final Great Lakes human health criteria
methodology.

      b.    Comments:  Many commenters were in favor of the proposed risk
level of one in 100,000, while  other commenters were in favor of higher and
lower risk levels (one in 10,000 and one in a million).  Those who  favored one
in a million risk believed one in 100,000 was not stringent enough, and that
many of the exposure parameters in the criterion equation were not  protective
of high-end exposed individuals (e.g., high-end fish consumers). That  is, many
people who eat more*than 15 grams of fish per day would be protected at less
than 1 in 100,000 risk from cancer (that is, a higher risk).  Those that
favored a higher risk level, at one in 10,000, believed that EPA's  cancer
model, and other exposure assumptions are already so conservative that there
is no need to set criteria at the one in 100,000 risk level.  Other commenters
referred to EPA's Superfund program,  which uses a risk level of  one in 10,000,
and cited inconsistency among EPA programs.

      c.    Final Guidance:  The final Guidance continues to derive criteria
and values based on a 10'5 risk level.  EPA believes  that this is a  reasonable
decision in light of the fact that it reflects the policy preferences  of most
of the Great Lakes States.  EPA notes, however, that selection of this risk
level does not reflect a judgment that this is the only level of acceptable
risk that would achieve the human health protection  goals of the CWA.  Rather,
as noted above, EPA believes that ensuring protection in the range  of  10"* to
10"6 is acceptable and consistent with the CWA's objectives.  Commenters were
correct in noting that some identifiable subpopulations or groups consume more
fish than 15  grams/day, as assumed in the methodology, and thus  such consumers
may face a risk that is higher than 10"5.  However,   EPA believes these
consumers will nonetheless be adequately protected.   (See section V.B.S.e of
this document for detailed information on fish consumption in the Great Lakes
basin.)

-------
152    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      EPA disagrees with those commenters advocating the use of  a  one  in
10,000 risk level.  While the model used to estimate cancer risk has many
conservative assumptions that are likely to overestimate actual  risks,  EPA
believes this conservatism is appropriate given the severity of  the potential
outcome  (cancer) and the uncertainties inherent in extrapolating from  the  high
doses administered to laboratory animals to the low doses experienced  by
humans.  Moreover, setting the target risk level at 10"4 for average consumers
would not include any margin of safety that might be appropriate to ensure the
protection of high-end consumers.  With regard to the commenter's  concern  with
the conservative nature of the exposure assumptions, EPA attempts  to select
reasonable assumptions given the need to provide adequate protection to
humans.

       EPA disagrees with the commenter who asserted that EPA requires the use
of a 10"4 risk level in the Superfund program, and that the risk  level  used in
the proposed Guidance's human health methodology is therefore inconsistent
with the Superfund program.  Under EPA's Superfund program, acceptable
exposure levels are generally those falling within the  10"4 to 10"* risk range.
See 40 C.F.R. 330.430  (2) (i) (A) (2) .

      Finally, EPA disagrees that the final Guidance is inconsistent with
other EPA programs regarding the use of a 10"5 risk level.  As stated above,
EPA believes the 10"4 to 10"* risk level will provide adequate protection of
public health.  In any particular regulatory action under its various
authorities, EPA targets a level of public health protection that  effectively
implements EPA's statutory duties and which is tailored to the specific
circumstances being addressed by the regulatory action.  As stated previously,
EPA believes the approach in the final Guidance is consistent with EPA's
historical practices in its public health protection programs.

      As noted in section II of this document, levels of some pollutants fish
tissue in the Great Lakes basin currently demonstrate that applicable  criteria
are being exceeded, and baseline risks associated with these exposures in  some
cases exceed the 10"4 to 10"* risk range that EPA considers adequately
protective.  In particular, groups that generally consume more fish on average
than the general population  (e.g.. Native American subsistence fishers and low
income minority sport anglers) are at greatest risk, which EPA has estimated
may be as high as 3.7 x 10"2 for Native Americans in Lake Michigan  due
primarily to PCBs.  The purpose of the human health criteria and the criteria
methodology contained in the final Guidance is to ensure that the  adoption of
water quality standards that, where they are attained would provide adequate
public health protection.  Obviously, where standards are not yet  attained,
actual risk levels will be higher and steps will have to be taken  to meet
applicable standards.  Nonetheless, EPA evaluates the protectiveness of the
criteria and the methodology assuming that criteria are met in order to ensure
that protective water quality standards will be established in the Great Lakes
basin.

      With regard to comments that EPA should adopt a lower risk level (i.e.,
10-6), EPA notes that States and Tribes are free to adopt a more stringent
approach than that contained in the final Guidance.  Given that  a  plurality of
States have utilized the 10-5 risk level, EPA believes that the  choice of  risk
level is a reasonable one.

      With regard to criteria for protection against non-cancer  effects, these
criteria are derived so as to prevent hypothetically exposed individuals
 (i.e., those consuming pollutant-bearing fish and drinking water at the rate
assumed  in the criteria-derivation formulas explained below) from  receiving a
dose of  the chemical above that which is calculated to correspond  to no
appreciable risk of adverse effect, based on a threshold model of  chemical
activity.  The issue of risk level is therefore not relevant to  such threshold
effects  (except  in these rare instances where a State or Tribe may find, and
EPA would agree, that a noncancer effect exhibits no threshold).

-------
                              Section V: Human Health                           153
      For the reasons stated above, the final Guidance will continue to derive
criteria and values based on a 10'5 risk level.

4.    Acceptable Dose

      The proposed Guidance used the term risk associated dose  (RAD) to
represent the dose associated with a one in 100,000 plausible upper bound risk
of developing cancer from lifetime exposure to a carcinogen.  The term
Acceptable Daily Exposure  (ADE) -was used to represent the dose of a
noncarcinogen expected to result in no appreciable risk of adverse health
effects from lifetime exposure.

      a.    RAD.

      The proposed Guidance discussed the steps for determining a Risk
Associated Dose  (RAD)  (see 58 FR 20865) .  EPA also pointed out that many of
the steps in developing a RAD may have already been conducted by EPA for a
particular chemical and the results made available to the public through EPA's
Integrated Risk Information System (IRIS). It was recommended that IRIS be
consulted when developing a RAD.   (The reader is referred to section V.C.4.C
for a description of IRIS and its utility in developing RADs.)

      i•    Biologically Relevant versus Sensitive Species

      (A).  Proposal:  EPA pointed out that when animal studies are used to
estimate effects on humans, data from species most biologically relevant to
humans are generally preferred  (i.e., a species in which pharmacokinetics
and/or toxic mechanisms of action appear closely related to humans).  In the
absence of data to distinguish the most relevant species, data from the most
sensitive animal species tested, i.e., the species exhibiting a carcinogenic
response at the lowest administered dose, given a relevant route of exposure,
should be used.

      (B).  Comments:  The majority of commenters argued that the most
biologically relevant species rather than the most sensitive species should be
used. Commenters stated that use of the most sensitive species must be weighed
against the assumption that humans are as or more sensitive than the most
sensitive species. A minority of commenters agreed with EPA's proposed
language to use the most sensitive species as a default, in the absence of
data.

      EPA agrees that the most biologically relevant species to humans should
be used when data is available for the chemical on the mechanism of the
carcinogenesis as well as the pharmacokinetics of the test species relative to
humans.   However, in many cases there will be limited data to assess how well
an animal model reflects human toxicological response for a chemical. In these
situations,  EPA believes it is prudent to use data from the most sensitive
species.  This is supported by the 1986 Guidelines for Carcinogen Risk
Assessment which states: In the absence of appropriate human studies, data
from a species that responds most like humans should be used, if information
to this effect exists...because it is possible that human sensitivity is as
high as the most sensitive responding animal species, in the absence of
evidence to the contrary,  the biologically acceptable data set from long-term
animal studies showing the greatest sensitivity should generally be given the
greatest emphasis, again with due regard to biological and statistical
consideration (U.S.'EPA, 1986).

      (C).   Final Guidance:  For the reasons stated above, the final Guidance
is retaining the proposed language that data from species most biologically
relevant to humans is preferred.  However,  in the absence of data to
distinguish the most relevant species, data from the most sensitive species
will be used.   This provision applies to both cancer and noncancer effects.

-------
154    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      ii.    Less than Lifetime Adjustment Factor.

      (A).   Proposal:  In the proposed Guidance,  EPA included   a. factor
(L/Le)3, where  L is the natural  lifespan of  the test species and Le is the
duration of the study.  This adjustment factor is necessitated when an  animal
study being used to derive a dose response relationship is not of sufficient
duration to measure cancer development over the natural lifespan of the test
species.  The rationale for such a factor is described in detail in the 1980
National Guidelines (45 PR 79352).  Under the proposal, the slope factor
adjustment was applied to mouse and rat data if the study duration  (Le) was
less than 78 weeks for mice or 90 weeks for rats, by multiplying the  cancer
slope factor by the factor (L/Le)3.   EPA requested comment on whether the use
of this adjustment factor for studies with less than lifetime duration  was
appropriate.

      (B).   Comments:  Several commenters agreed that an adjustment factor
could be used but disagreed with the choice of the factor proposed and  the
cutoff  (78 and 90 weeks)  chosen in the proposed Guidance.  These commenters
suggested that the adjustment factor should be applied on a case-by-case
basis, considering such factors as mechanism of action, and type of tumor and
the organ affected.

      EPA agrees with the commenters that the best way to evaluate this
adjustment factor  is on a case-by-case basis since not all carcinogens  behave
mechanistically the same way  (as pointed out by the 1986 Guidelines for
Carcinogen Risk Assessment).  For some carcinogens, exposure early on (during
childhood or adolescence, or the equivalent in test animals) may be critical
to tumor expression.  For others, exposure of a test animal over a year may
result in no more  tumors than exposure over a lifetime  (two years).

      (C).  Final Guidance:  To allow flexibility in making these judgments,
EPA has changed  the final Guidance to allow for less than lifetime adjustments
on a case-by-case basis.  Less than lifetime adjustments to the  cancer  slope
factor are now optipnal, not required, and should be made on the basis  of
existing mechanistic data.  However, in the absence of data on mechanisms and
time to tumor, States and Tribes may continue  to use the adjustment factor and
the duration cutoffs prescribed  in the proposal as a default approach.

      iii.  Species Scaling Factor.

      (A).  Proposal:  In the proposed Guidance,  EPA would have required the
use of a "surface  area species scaling factor" in deriving a dose  response
relationship for humans that  is based on animal data. The proposed Guidance
assumed that exposures in milligrams per kilogram of body weight per  day
raised to the  2/3  power  (also referred to as the two-thirds exponent) would
yield equivalent cancer responses in test animals and humans.   EPA requested
comment on  the proposed use of a  two-thirds  exponent for  surface area,  and the
possible use of  a  three-fourths  exponent for body weight, for  scaling animal
doses to equivalent human doses.  EPA also  asked for comment on whether use  of
some other  scaling factor  should be used in the final Guidance.

       (B).  Comments:  Several  commenters were in favor of  the  3/4 body weight
scaling factor while  others  were in favor  of  the 2/3  surface  area scaling
factor.  Those  in  favor of the  3/4 body weight scaling  factor  cited the draft
consensus by the Inter-Agency group  (U.S. EPA,  1992) and  empirical  data which
supports the scaling  on the  basis of body weight  (Allen et  al.,  1987) .   Others
suggested flexibility and  allowing adjustments on a case-by-case basis,
determined  by  the  pharmacokinetics of the chemical.

      As cited in  the proposed  Guidance, an Inter-Agency  group comprised of
EPA, U.S. Food and Drug Administration  and  CPSC has  deliberated on the  issue
of  appropriate,  consistent scaling factors  for use  by  all agencies in
developing  risk assessments.  While  the  draft  recommendation from this group
was  to  apply a 3/4 exponent  scaling  factor,  this  undertaking has not been

-------
                              Section V: Human Health                           155
finalized nor adopted as EPA policy to date.  The current policy is stated in
the 1986 Guidelines for Carcinogen Risk Assessment: In the absence of
comparative toxicological, physiological, metabolic and pharmacokinetics data
for a given suspect carcinogen, EPA takes the position that extrapolation on
the basis of surface area is considered to be appropriate because certain
pharmacological effects commonly scale according to surface area  (Dedrick
1973, Freireich, et al., 1966; Pinkel;, 1958).  As discussed in the preamble
to the proposal, there are divergent views among the scientific community
regarding whether a body weight or surface area scaling factor provides the
best means of extrapolating results from animal studies to humans, and as
reflected by the work of the Inter-Agency Pharmacokinetics Group, a consensus
among relevant governmental agencies may be reached in the future that the
body weight approach is preferable.  However, no final consensus has yet been
reached.  Because EPA's current, longstanding policy calls for use of the
surface area scaling factor, EPA is hesitant to change its approach in the
absence of a final determination that adoption of the body weight scaling
factor is appropriate.

      (C).  Final Guidance:  Consistent with the existing EPA policy, the
final Guidance calls for the use of the 2/3 exponent scaling factor  (surface
area) , in the development of cancer slope factor development.

      b.    APE.

      For non-carcinogens, the proposed Guidance called for a data hierarchy
for calculating the ADE.  The process outlined is the same one used by EPA's
RfD development process but differs in the amount of data required to develop
a criterion or value, especially with regard to Tier II ADEs.  In some cases,
an ADE may be identical to an EPA RfD if the same data and judgments are used.
However, since States and Tribes may derive values which may vary based on
different interpretations of the supporting data and/or the existence of new
data, EPA proposed the term ADE to distinguish the RfD from the ADE.

      The proposed Guidance indicated that calculating an ADE for a chemical
involves the following steps: 1) determining whether there is evidence from
epidemiologic or animal studies that exposure to a chemical may result in
adverse noncancer health effects; 2)  using available data to determine a
threshold dose value that is likely to be without appreciable risk of adverse
effect;  and 3)  reducing this threshold dose value to account for
uncertainties inherent in the risk assessment to yield an ADE for humans.

i.    List of Deleterious Effects.

      (A).  Proposal:  The proposed Guidance presented a list of adverse
effects which the noncancer criteria are protective of, including adverse
acute, subchronic and chronic effects and reproductive and developmental
effects.  EPA solicited comment on whether it should specify in the
methodology a longer list of deleterious effects against which noncancer
criteria should protect.

      (B).  Comments:  Several commenters agreed with the list of deleterious
effects while others requested that EPA expand the list to include such
effects as immunotoxicity.  Other commenters believed a list was not entirely
necessary since the assessment of adverse effects was mostly based on
professional judgment.

      (C).  Final Guidance:  EPA believes the list as stated encompasses any
effect which can be deemed adverse including immunotoxicity, based on the
State or Tribe's professional judgment.  Thus, EPA is retaining the proposed
list in the final Guidance.

-------
156    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

ii.   Uncertainty Factors.

       (A).  Proposal:  The proposal provided guidance on the use of
uncertainty factors to account for uncertainties in predicting acceptable
exposure levels for the general human population.  The proposal  indicated  that
the size of the uncertainty factor varies depending on the data  available  for
ADE calculation, including whether the data are from a study on  humans  or  test
animals, and on whether the study demonstrates a no observed adverse effect
level  (NOAEL) or a lowest observed adverse effect level  (LOAEL).  Under the
proposed Guidance a composite uncertainty factor of 30,000 was described as
the maximum uncertainty allowed when deriving a Tier I criterion or Tier II
value.  EPA solicited comments on the uncertainty factors included in the
proposal, and particularly on whether other uncertainty factors might offer
better assessment of risk, and be more appropriate in deriving water quality
values and criteria.

       (B).  Comments:  Many commenters stated that the use of default
uncertainty factors of 10  (each) to account for intra-, interspecies
variability, and subchronic to chronic adjustments are too rigidly applied.
Commenters felt uncertainty factors of less than 10 should be allowed if there
is a scientifically sound rationale.  Other commenters suggested that EPA  add
an additional uncertainty factor to the calculation of an ADE to account for
the sensitivity of children or other sensitive subgroups.

      With regard to the use of uncertainty factors less than 10, EPA agrees
that uncertainty factors of less than 10 can be used to account  for intra-,
interspecies variability, subchronic to chronic adjustments and  LOAEL to NOAEL
adjustments as long as data exists to justify a lower uncertainty factor.  As
stated  in the proposed Guidance, EPA generally applies uncertainty factors of
10, but is not advocating the rigid application of uncertainty factors  of  ten
in all  instances.  This is only meant to be a default position when data is
lacking to justify a lower uncertainty factor.  For example, if
pharmacokinetics data shows that a test animal metabolizes a chemical in a
fashion identical to humans, a lower  interspecies uncertainty factor is
probably justified.  For information on selecting uncertainty factors less
than 10, see appendix A of the Human Health TSD.

      With regard to the comment suggesting EPA apply an additional
uncertainty factor to protect children and sensitive subgroups,  EPA believes
an RfD  (ADE) should,be protective  (by definition) of children and sensitive
subgroups.  In assessing the data base, consideration should be  made of these
segments of the population.  If the overall data base does not yield
information on the toxicity of test organisms during all segments of their
lifetime  (such as infancy, childhood, adolescence), or does not  consider
sensitive individuals,  an additional uncertainty factor may be applied  to
account for such uncertainty in an incomplete data base.  It is  important  to
note that the intraspecies uncertainty factor of 10, which is commonly
applied, is designed to account for sensitive individuals in the population.
Therefore,  the application of another uncertainty factor would be redundant.

       (C).  Final Guidance:  As stated above, the proposed Guidance  stated
that a maximum uncertainty factor of  30,000  could be applied in  deriving a
Tier I  criterion or  Tier  II value.  In the  final Guidance, EPA is clarifying
the use of  uncertainty  factors  to include a maximum of 10,000 for Tier  I
criteria and a maximum  of  30,000  for  Tier II value development.  When deriving
a Tier I criterion,  the likely maximum composite uncertainty factor  applied to
a 90-day NOAEL may be 3000.  The  total of 3000  is based  on four  separate
uncertainty factors: 10 to account for intraspecies variability; a factor  of
10 to  account for the uncertainty in  extrapolating animal data to the  case of
humans;  a factor of  10  to account for subchronic to chronic variability;  and a
factor of three  to  account for  an incomplete data base.  However,  in rare
cases  where an  extra uncertainty  factor  is  required' to compensate for an
incomplete  data  set  or  to compensate  for  the uncertainty associated  with  a
LOAEL  (rather than  a NOAEL), a  composite  uncertainty  factor  of  10,000 may be

-------
                              Section V: Human Health                           157
used.  Under this assumption an uncertainty factor of 10 would be applied to
account for an incomplete data base or the use of a LOAEL in place of a NOAEL
instead of using an uncertainty factor of three.  This decision requires the
use of professional judgment in determining the adequacy of the data base.
For Tier II values, the maximum allowable composite uncertainty factor may be
30,000 which would be applied to a studies with durations greater than a 28-
day minimal LOAEL  (e.g., 30 days) .

      The total of 30,000 is based on the three factors of generally 10 each
for intraspecies variability, interspecies extrapolation, subchronic to
chronic extrapolation, a combined factor of 10 to account for incompleteness
of data bases and the difference between a minimal LOAEL and NOAEL (which
together result in an overall uncertainty factor of 10,000), and an additional
uncertainty factor of 3 to account for the uncertainty in extrapolating from a
study greater than 28 days, but sufficiently less than 90 days.
      i .    Proposal :  The proposed Guidance indicated that EPA has a process
to develop consensus on cancer slope factors (ql*s) and RfDs  (referred to as
ADEs for the Great Lakes) .  These values are derived by two EPA work groups
called the RfD/RfC and CRAVE work groups and made available as guidance to EPA
program offices and' the public via a data base called the Integrated Risk
Information System  (IRIS) which is accessible through the National Library of
Medicine's Hazardous Substance Data Base (HSDB) .  The values  (RfDs and ql*s)
listed on IRIS are guidance; they are not regulatory in nature.

      The proposed Guidance recommended that verified IRIS values  (RfDs and
cancer slope factors) be considered as a first step in deriving the Great
Lakes Human Health criteria.  It is also important to note that many chemicals
have been reviewed by the RfD or CRAVE Work Groups, have been verified, but
have not yet been entered  into the IRIS system.  A verified RfD or ql* is a
value which the RfD. or CRAVE Work group has reviewed and officially verified
during an RfD or CRAVE meeting.  EPA encourages States to call EPA's RfD and
CRAVE Work Groups if they are not sure of the current status of a specific
chemical.  A verified RfD or ql* can be utilized by States and Tribes  even
though it has not been officially placed into the IRIS system.  The proposal
also allowed deviation from these values when new data are available or if a
different interpretation of the data is made .  EPA requested comments on
deviating from IRIS values in deriving Great Lakes criteria and values for the
reasons highlighted above.

      ii.   Comments :  Many commenters wanted the flexibility to deviate from
IRIS while others argued that deviation from IRIS values should not be
allowed.

      EPA has decided that deviations from IRIS should be allowed since IRIS
is only guidance and that other interpretations of the data may be valid.
However, to foster consistency between EPA and the States, EPA strongly urges
the States/Tribes to communicate any anticipated or necessary deviation from
IRIS to the EPA RfD and/or the CRAVE work groups as soon as possible .
Following this recommended course of action will allow EPA to discuss the
potential deviations with the State or Tribe and could lead to an expedited
review of the chemical and data by the EPA work group of concern.  In
addition, as noted earlier, States may use verified RfDs or cancer potency
factors, which have not been entered onto IRIS, for development of Tier I
criteria.  Finally, when deviating from IRIS, States/Tribes are encouraged to
work with the Clearinghouse described in section II, to ensure other
States/Tribes are aware of the deviations .

      iii.  Final Guidance :  For the reasons stated above, States and Tribes
may deviate from IRIS- provided the approach is scientifically defensible.

-------
158    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


S.     Exposure Assumptions

      The proposed Guidance identified seven factors which affect an
individual's oral exposure to a chemical: body weight; duration  of exposure;
recreational exposure; drinking water consumption; fish consumption;
bioaccumulation factor and relative source contribution.

      a.    Body Weight.

      i.    Proposal:  The proposed criteria methodology assumed a mean adult
human body weight of 70 kg.  This value is consistent with that  recommended in
the EPA's Exposure Factors Handbook (EPA 600/8-89/043, July 1989).  EPA
requested comments on the use of the  70 kg body weight assumption and also
asked for comments on the issue of using body weights of sensitive
subpopulations  (such as children) when a chemical's toxicity indicates  a
specific subpopulation is most sensitive to exposures.

      ii.   Comments:  Several commenters advocated the continued use of  the
70 kg body weight assumption.  Several other commenters stated that the body
weight used in calculating criteria and values should be lower than 70  kg to
protect women of childbearing age and children and fetuses.  Several of these
commenters recommended 55 kg as the body weight.  No commenters  suggested an
actual child weight, which should be used nor did any commenter advocate a
higher body weight than 70 kg.

      EPA believes 70 kg is an appropriate body weight because it represents a
reasonable measurement for the entire population.  As stated in  the EPA
Exposure Factors Handbook  (1989), the mean body weights for women and men
nationally are 65 kg and 78 kg, respectively, with an overall mean adult  body
weight of 71.8 kg.  The Handbook also indicates that the mean body weight for
child bearing women ages 18-45 is 63  kg.)  If a State believes that use of a
lower body weight is appropriate  (which yields a more stringent  criterion),
the State or Tribe may adopt such an  assumption in calculating their criteria
and values under their authority to establish more stringent requirements
pursuant to section 510 of the Act.

      As to whether lower body weights should be used to protect women  of
childbearing age, children and fetuses, EPA believes that categorically
adopting more conservative body weight assumptions may not be appropriate.
Each chemical must be addressed  separately since some chemicals  may be
generically toxic to both  sexes, while others may be  specifically toxic to one
sex more than the other.   It therefore would not be appropriate  to require
generally that all criteria be based  on conservative body weight assumptions.
In the  case of mercury, however, a fetotoxic chemical, to be protective of
women of child bearing age, EPA has assumed a body weight of 65  kg  (as  opposed
to 70 kg) which  results in a Tier I mercury criterion of 1.8 ng/L, which  is
slightly less than the proposed  criterion of 2 ng/L.  EPA has set a. final Tier
I criterion for  mercury at 1.8 ng/L.

      iii.  Final Guidance:  Based on the reasons above, the final Guidance
retains the 70  kg body weight assumption.   States or Tribes may  use the female
body weight  (55-65 kg) or  a  child's body weight  (10-30 kg) or any other more
 "stringent" assumption on  a  chemical-by-chemical basis as deemed appropriate
based on the properties of a particular chemical.  If a  child's  body weight  is
assumed, the water  consumption  rate  should  also be adjusted for  a child to  l
liter/day.

      b.    Duration  of Exposure.

       i.    Proposal:  The proposed  Guidance assumed  that oral exposure
 remains constant for  a  lifetime  and  assumed a  70 year exposure period in
 developing  criteria  and values.   EPA requested comments  on  the use  of  longer
 lifetime exposure periods,  such as 75 years instead of  the  currently proposed
 70 years,  since recent  census  data indicates that  the average  lifespan for

-------
                              Section V: Human Health                           159
Americans is 74.7 years.  EPA also requested comments on whether the use of
shorter exposure periods  (i.e., less than 70 years) would be more appropriate
to account for mobility of individuals in and out of the Great Lakes basin.

      ii.   Comments:   Several commenters suggested that EPA increase the
lifespan to 75 years or longer to account for women and people who live past
70 years.  Other commenters suggested  retaining the proposed 70 year value.
Other commenters cited the EPA Exposure Factors handbook which identifies nine
and 30 years as representative of the average and reasonable upper bound,
respectively, of length of time residing in the same house.  Commenters also
suggested that mobility in the area is high and that individuals in the Great
Lakes will not be exposed.to contaminants uniformly over their lifetimes.

      EPA believes the 70 year lifespan assumption is appropriate because it
conservatively assumes a uniform exposure to contaminants over a lifetime.
Since consumption patterns may change from infancy to death, EPA is seeking to
protect those individuals who may live their entire lives in the Region and
those who may encounter similar exposures while outside the region.  Without
exhaustive analysis, it would be difficult to determine that people who
migrate in and out of the region are actually experiencing different exposure
patterns.  With regard to increasing the lifetime exposure to 75 years or
more, EPA believes 70 years is an appropriate assumption as it is still a
generally accepted value used by most risk assessors and regulatory agencies.

      iii.  Final Guidance:  For the reasons cited above, the final Guidance
retains the 70-year exposure period for developing criteria and values.  A
State or Tribe may chose to be more stringent and adopt a longer exposure
period in its criteria derivation.  It should be noted that this assumption
does not influence cancer or noncancer criteria and values which are based on
animal data since all lifetime animal data is assumed equivalent to a human
lifetime, but may have a bearing on criteria and values based on epidemiology
studies, depending on other numerical considerations in the calculation.  If
an epidemiology study is less than lifetime in duration, the dose must be
translated into a lifetime human dose based on a specific lifetime duration
assumption, either numerically or through the use of an uncertainty factor.
In such a case,  the assumption regarding the duration of a lifetime would be
relevant to determining the human health criterion.  In fact, it is relatively
rare that epidemiological data is used to derive human health criteria due to
limitations in the study which preclude a conclusion of a cause and effect.
In the final Guidance there are two cases where epidemiological data were
used; in deriving the HCV for benzene and the human noncancer value (HNV) for
mercury.

      c.    Incidental Exposure.

      i.    Proposal:   The proposed Guidance included an incidental ingestion
exposure factor which accounted for oral exposures which might occur through
recreational activities in or on the water.  The factor  (0.01 liters/day) was
included in the derivation of the GLWQI criteria for those surface waters that
are not used as a drinking water source (Nondrinking HNVs and HCVs).  This
factor was not used in those cases where the waterbody was designated as a
drinking water source.  EPA requested comments on whether the factor was
justified and also requested comments on whether a factor should be included
for incidental dermal exposure which occurs through recreational activities.
EPA also requested submission of any data that could be used to derive such a
factor.

      ii.   Comments:   EPA received many comments stating that the extra
factor for incidental exposure was unnecessary since the drinking water
assumption of 2 L/day appears overly conservative.  EPA also received comment
that attempting to account for dermal uptake would be very difficult,  if not
impossible.  No data were received on dermal uptake.

-------
160    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


      EPA agrees that the extra factor for incidental exposure is not
necessary to include in the calculation of criteria and values which are to be
used in protecting waters designated as drinking water supplies  because such
an exposure assumption is of negligible significance when viewed in light of
the 2 L/day assumption used in deriving criteria for these waters.
(Numerically, the 2.0 liter value in the numerator is so large in comparison
to the 0.01 liter value,  that it has no numerical effect on the  calculated
criterion.)  However, EPA continues to believe that the inclusion of the
incidental exposure factor is appropriate in developing criteria and values
for waters designated as non-drinking water  (i.e., recreational  only).

      The effect of using the incidental exposure factor when deriving
criteria and values for nonbioaccumulative chemicals can be numerically
significant.  For example, for nondrinking water sources, when the incidental
exposure factor is included, the HNV for methylene chloride is 90 mg/L.  When
the incidental exposure factor is not included, the HNV is 120 mg/L.

      In addition, EPA also believes it is appropriate to account for
incidental ingestion in a recreational setting since on an individual or
subpopulation basis this exposure may be significant.  For example, those
people who spend more than an average amount of time recreating  in water may
exceed the 0.01 liters/day incidental ingestion rate.   (The 0.01 value  is
based on a number of average assumptions for the Great Lakes basin population
as a whole.)  If States or Tribes want to establish a larger incidental
exposure factor for recreational  subpopulations  (swimmers, waterskiers,
rafters, kayakers), they may do so.  This factor may become especially
important in the future,  when criteria for microbiological agents are
developed, since very minute amounts of microbially contaminated water  can
cause adverse health effects.

      iii.  Final Guidance:  The  final Guidance retains the incidental
exposure factor of 0.01 liters/day for those surface waters that are
designated as nondrinking water sources.

      d.    Drinking Water Consumption.

      i.    Proposal:  EPA proposed that the criteria and values be derived
using an assumption,of 2 L/day drinking water consumption from untreated
surface waters for waters designated as drinking water sources.  This
represents approximately the 90th percentile ingestion value for the  general
population.  The Exposure Factors Handbook,  1989, states:  For the reasonable
worst-case value, the 90th percentile rate  reported by Gillies and Paulin
 (1983), 1.90 L/day, suggests that a rate of  2.0 L/day may be a reasonable
approximation.

      The  90th percentile value suggested by Cantor et al.  (1987) is  also
approximately 2  L/day.  This is also  supported by a more recent  study
specifically characterizing  tap water intake by Ershow and Cantor  (1989) in
which 2 L/day represents approximately the  85th percentile value of drinking
water consumption.  This value, 2 L/day, is  recommended as the reasonable
worst-case  consumption rate.  EPA requested comment on  whether  selection  of
another value such as 1.4  liters  per  day, the national average,  would be more
appropriate.  EPA also noted that, since the 2.0  liters value  is a
conservative assumption  (only 10  percent of  the population drinks two liters
of water a day and considerably less  can be  expected to drink  two liters of
untreated  surface water) and the  0.01 liter associated with incidental
exposure is  relatively insignificant  in comparison, EPA presumed that 2 L/day
 is protective of both drinking water  and incidental ingestion  exposures for
waters  which may be^both a  drinking water source  and used for  recreation.  EPA
requested  comments on whether such an assumption  was justified.  EPA  also
requested  comments on whether surface water criteria for waters  designated for
drinking water uses  should assume consumption of  untreated water, as  was
proposed.

-------
                              Section V: Human Health                           161
      ii.   Comments:  Several commenters suggested 2 L/day was too
conservative.  Others stated that the level of conservatism was appropriate.
Those that felt it was overconservative believed that the national average is
1.4 liters/day and that people do not generally drink untreated surface water
(in any amount).  No commenters suggested that 2 L/day was underprotective.
Other commenters stated that some cultural activities conducted by Native
Americans, involve the direct consumption of untreated surface water, although
the specific nature of these practices was not described.

      EPA believes that the consumption rate of 2 L/day, even though
conservative, is reasonable given that it will provide a high level of
protection for the people in the Great Lakes consistent with the protective
goals of the Act.  EPA, in determining a water consumption rate, considered
adopting the average consumption rate (1.4 liters/day), as discussed in the
preamble to the proposal and suggested by commenters.  However, in choosing
among the available'policy options, EPA chose 2 L/day to provide an extra
degree of protection for those individuals (which may be a substantial number
of persons) who consume more than the population on average.  EPA believes
that this assumption is reasonable in light of the public health protection
goals of the CWA.  EPA also believes that it would not be appropriate to
dismiss drinking water usages, because as asserted by some commenters, there
are expressed cultural practices which include consumption of untreated
drinking water.  EPA also believes that it would not be appropriate to assume
that untreated water is never consumed by users since the CWA requires
protection of designated uses.  To protect waters designated as drinking water
sources,  it is impprtant that public drinking water systems not be the only
point of clean up and that surface water discharges also share the burden of
maintaining the use designation of the water.

      iii.  Final Guidance:  The final Guidance retains the assumption that
individuals consume 2 L/day of untreated surface waters when developing
criteria/values for waters designated as drinking water sources.

      e.    Fish Consumption.

      i.    Proposal:  The proposed Guidance included a fish consumption rate
of 15 grams per day-,  which represents the mean exposure level for regionally
caught fish for the regional sportfishing population.  This value also
approximates the 90th percentile for the entire regional population.  Thus, a
more conservative target population was chosen than is used in the National
criteria methodology and the proposed fish consumption value was based on
Great Lakes-specific statistical data.  The actual value of 15 grams per day
was derived from review of several regional studies in Michigan (West, et al.,
1989),  Wisconsin (Fiore et al.,  1989) and New York (Connelly, et al., 1990).
Since the proposal, several new pieces of information  have emerged that were
used in EPA's consideration of the appropriate fish consumption rate: First,
an Executive Order on environmental justice was issued February 11, 1994.  It
requires Federal agencies (to the greatest extent practicable and permitted by
law)  to "identify and address, as appropriate, disproportionately high and
adverse human health or environmental effects of its programs, policies, and
activities on minority population and low income populations."  Federal
agencies are also required,  whenever practical and appropriate to "collect,
maintain, and analyze information on the consumption patterns of populations
who principally rely on fish and/or wildlife for subsistence."

      Second, a study conducted by West et al. (1993) for the State of
Michigan was provided by the author to EPA during the public comment period.
This study is a full year (February 1991 to February 1992)  fish consumption
survey of 7000 licensed Michigan anglers.  The survey found that the average
sport fish consumption rate, adjusted for non-response bias, was 14.5
grams/day.  The average total fish (all fish, not just Great Lakes sport fish)
consumption rate, adjusted for non-response bias, was 24.4 grams/day.  This
study indicated that fish consumption rates may differ according to race and
income  level.  The lowest income group (< $i4,999/year)  averaged 21 grams/day

-------
162    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


sport fish consumption as compared to 14.7 grams/day for those making  $40,000
or more/year.  The average sport fish consumption rate for minorities  was  23.2
grams/day as compared with 16.3 grams/day for non-minority individuals.  Lower
income ($24,999 or less) minorities averaged the highest consumption rate  of
all groups in the survey: 43.1 grams/day sport caught fish and 57.9 grams/day
total fish;  Non-minority individuals of lower income averaged 18.6 grams/day
sport fish and. 25.8 grams/day total fish.  The study also indicated that
minorities eat less.fish from the Great Lakes and more fish from the inland
tributaries than non-minority individuals.

      In the proposal, EPA also requested comments how to protect against
acute and subchronic effects, the possibility of adverse effects due to binge
eating.  Specifically, EPA invited comment on the use of 454 grams  (one pound)
as a reasonable worst-case, one-day fish consumption estimate and 2,240 grams
as a reasonable worst case, 10-day fish consumption estimate based on  10
consecutive days of consumption of approximately one-half pounds of fish per
day.  These values could be used in deriving one-day and 10-day
criteria/values protective of acute and subchronic effects.  In a Notice of
Availability dated August 30, 1994  (U.S. EPA, 1994a)  EPA asked for comments
on the fish consumption rates identified in the West  et al.  (1993) study  and
on the quality of the study itself.

      ii.   Comments:  Many commenters believed EPA's proposed rate of 15
grams/day would not be protective of recreational and subsistence anglers  such
as the Native American anglers and minority anglers, or women of childbearing
age and children.  An alternative rate of 50 grams/day was suggested.   Others
believed the rate of 15 grams/day was too conservative since it is protective
of at least 90 percent of the overall population and these commenters
suggested a lower rate.  With regard to the use of one-day and ten-day worst
case rates, comments suggested that this is a worthwhile attempt at dealing
with a potential short term risk, but it was not clear how this could  be
implemented in a water quality standards context.  Other commenters stated
that one or ten day exposures are not long enough for biouptake and
distribution to reach a steady state of chemical concentration in a human  high
enough to result in a toxic effect.

      With regard to comments on the August 30, 1994 Notice of Data
Availability  (U.S. EPA, 1994a), many commenters were in favor of retaining the
proposed fish consumption rate of 15 grams/day, citing the relative
conservatism of the-entire criterion methodology.  Many of these same
commenters questioned the validity of the West et al.  (1993) study findings.
Specifically, they questioned the lower income minority average of 43.1
grams/day since this number was based on a very small sample size and  was
associated with a large standard deviation.  These same commenters
acknowledged that there are subpopulations who consume more than 15 grams/day
and that the best way to protect these people is to set site-specific  fish
consumption rates.  Several other commenters were in favor of a higher fish
consumption rate ranging from 30-60 grams/day.  These  same commenters
supported the findings of the West et al.  (1993) study, claiming that  the
study does show that lower income and minority subpopulations eat more sport
caught fish on average, but also pointed out that while the data was presented
correctly, many of the opinions presented in the study appeared overly
subjective.

      One group of commenters suggested that EPA establish in the final
Guidance a new approach to establishing fish consumption rates based on
default values derived  from national data.  These commenters argued that
regional fish consumption  survey data are time-consuming and costly to
collect, and  consequently are not often available.  The commenters recommended
that States and Tribes  instead utilize default fish consumption  rates  based
upon the particular-fish consumption use of the waterbody  (i.e., 20 grams/day
for average fishing, 45-140 grams/day for sport fishing and 90-165 grams/day
for subsistence  fishing, with special consideration for Native American
subsistence uses).  These  commenters  also argued that use of the proposed fish

-------
                              Section V: Human Health                           163
consumption rate would not be in accordance with Executive Order 12898
addressing Environmental Justice, and would be discriminatory under the Civil
Rights laws.

      After consideration of the public comments, EPA has decided to retain
the 15 grams/day fish consumption rate.  EPA does not agree with commenters
who argued that this assumption would not provide adequate protection of '
public health.  As stated in the proposal, this value approximates the 90th
percentile for the overall population in the Great Lakes basin and the mean
value for sport anglers in the basin.  EPA believes the results from the West
et al. (1993) study  (a mean consumption rate of 14.5 grams/day for Michigan
sport anglers), along with the studies cited in the proposal  (see 58 Fed. Reg.
20870)  supports the reasonableness of the fish consumption rate used in the
final Guidance.

      EPA also believes that attainment of criteria in ambient waters that are
derived using the assumed fish consumption rate will result in adequate
protection for highly exposed populations.  In carrying out regulatory actions
under its statutory authorities, including the CWA, EPA generally views an
upper bound incremental cancer risk in the range of 10"4 to 10"* as adequately
protective of public health.  As discussed above, the human health criteria
methodology is based on a target risk level of 10"5.  If fish are contaminated
at the level permitted by criteria derived under the final Guidance,
individuals eating up to ten times the assumed fish consumption rate  (i.e.,
150 grams/day) would still be protected at a 10"4 incremental risk level,
within the range that EPA believes is protective of public health.  Thus,
while EPA acknowledges that some portion of the population will consume
greater than 15 grams/day, the fish consumption rate and the target risk level
chosen in the final Guidance combine to ensure that the population as a whole
will be adequately protected by human health criteria when the criteria are
met in the ambient water.

      Available data support the protectiveness of this approach.  An analysis
of the West study indicates that approximately 99 percent of all persons
surveyed in that study consumed less than 150 grams/day and that, even among
low-income minorities who as a group consume more fish than the population on
average,  approximately 95 percent consumed less than this amount.  Thus, EPA
concludes that full implementation of the final Guidance will result in a high
degree of protection for the population as a whole, including those
populations that consume greater amounts of fish than the population on
average.

      The final Guidance requires, moreover, that States and Tribes adopt an
additional degree of protection where it is determined to be appropriate for
highly exposed populations.  Section A.4.a. of procedure 1 in appendix F to
part 132 states that "Human health criteria or values shall be modified on a
site-specific basis to provide additional protection appropriate for highly
exposed populations."  Thus, if a State or Tribe determines that a highly
exposed subpopulation (e.g., a Tribe of subsistence fishers) would not be
adequately protected by criteria or values derived using the final Guidance
methodology, the State or Tribe would be required to adopt a more stringent
site-specific modification to criteria to provide appropriate additional
protection (e.g., to ensure protection at least at the 10"4 risk level, or
other more stringent degree of protection determined by the State or Tribe to
be appropriate).  Finally, the final Guidance requires States and Tribes to
adopt provisions to ensure that human health is protected from the adverse
effects of mixtures of pollutants in the Great Lakes System, including
mixtures of carcinogens.  See procedure 4 of appendix F to part 132; section
VIII.D of this document.  Thus, taking into account all relevant portions of
the final Guidance - - e.g., the target risk level, site-specific modification
procedures,  additivity provisions, as well as the many conservative
assumptions in the human health methodology discussed elsewhere in this
document --  EPA believes that waterbodies meeting criteria developed in

-------
164    Water Quality Guidance for the Great Lakes System -- Supplementary Information Document


accordance with the final Guidance will be protective of all segments of the
public, including any more highly exposed populations.

      EPA acknowledges the problem identified by commenters that high quality
survey data describing fish consumption patterns are often not available.  EPA
believes, however, that States and Tribes retain substantial flexibility in
determining how to address the fish consumption issue should the State  or
Tribe decide that a more stringent approach than EPA's is appropriate.  Where
there is an absence of survey data for an area, but evidence of uses that the
State or Tribe believe should be specifically reflected in a more stringent
fish consumption rate, the State and Tribe could reasonably decide to adopt
"default" values to provide a level of protection determined by the State or
Tribe to be appropriate for the particular fishing use made of the waterbody.
(EPA does not express a view as to the appropriateness of the specific  default
values suggested by commenters and noted above; such a determination can
appropriately be made by the State or Tribe based on its best professional
judgment.).

      Finally, EPA believes that its action here is fully consistent with the
Executive Order on Environmental Justice and civil rights laws because  the
final Guidance ensures that waterbodies meeting criteria developed using
methodologies consistent with the final Guidance will protect the health of
all of the public, including more highly exposed populations.  Section  1-101
of Executive Order 12898 provides that "[t]o the greatest extent practicable
and permitted by law  . . . each federal agency shall make achieving
environmental justice part of its mission by identifying and addressing, as
appropriate, disproportionately high and adverse human health or environmental
effects of its programs, policies, and activities on minority populations and
low-income populations . . . ."  Section 2-2 directs federal agencies to
conduct their activities in a manner that ensures that the "activities  do not
have the effect of ... subjecting persons  (including populations) to
discrimination under such programs, policies and activities because of  their
race, color, or national origin."  Finally, section 4-4 of the Order
specifically discusses subsistence consumption of fish and wildlife, and
provides that "[f]ederal agencies, whenever practicable and appropriate, shall
collect, maintain and analyze information on the consumption patterns of
populations who principally rely on fish and/or wildlife for subsistence," and
that "[f]ederal agencies shall communicate to the public the risks of those
consumption patterns."

      EPA's action here is consistent with the terms of this Executive  Order.
In developing the final Guidance, the EPA made a special effort to identify
and address the potential risks to highly exposed low-income and minority
populations, to maintain and analyze information on these risks, and to
communicate risks to the public, as reflected in the August 30, 1994, Notice
of Data Availability soliciting public comment on the fish consumption  issue
and the provisions in the final Guidance specifically targeted to protecting
highly exposed populations discussed above.  EPA also flatly rejects the
notion, advanced by-some commenters, that the final Guidance is in any  respect
discriminatory against persons or populations because of their race, color, or
national origin.  The final Guidance establishes criteria methodologies and
implementation procedures that are expressly designed to ensure full
protection of the public, including highly exposed populations.  While  some
groups and individuals, including some low income and minority persons  and
populations, may face a greater risk of adverse health effects than the
general population due to their particular fish consumption patterns, EPA has
included in the final Guidance mechanisms that, upon full implementation by
States and Tribes, will ensure that these groups will nonetheless receive a
level of public health protection within the range that the EPA has long
considered to be appropriate in its environmental programs  (i.e., 10"4 to 10"6
incremental cancer risk).  Obviously, as long as there is variability in fish
consumption patterns  among various segments of the population, it would be
impossible for EPA to ensure that all groups would face identical risk  from
consuming fish.  Therefore, EPA has sought to ensure that, after attainment of

-------
                              Section V: Human Health                           165
water quality criteria in ambient waters, no group is subject to increased
cancer risks greater than the risk range that the EPA has long considered
protective.

      As noted above, EPA solicited comment on whether EPA should develop
short-term criteria to protect against possible effects of worst-case short
term exposures.  No commenter supported the development of such criteria.
Commenters generally did not understand how such short-term criteria could be
implemented in the water quality criterion program as now constituted.
Commenters did generally support further research into the unanswered
questions associated with high, short-term exposure effects, including
fetotoxic effects.  Several commenters felt that fish advisories were the best
way to deal with this issue.

      EPA agrees with the commenters that this is an area that needs more
research before short-term water quality standards could be set.  Research in
the area of bio-uptake and whether a steady state can be achieved in a short
time period is needed before EPA could determine whether to develop short-term
criteria and values that would be sufficiently supported by the available
science to serve as sources of regulatory controls.  EPA acknowledges that
fish intake rates d& vary over the course of a lifetime and while EPA believes
the 15 grams/day assumption is adequately protective of the Great Lakes
population over a lifetime, States have the flexibility to establish criteria
specifically targeted to provide additional protection to sensitive
subpopulations (e.g., pregnant/nursing women, infants, children) or highly
exposed subpopulation (e.g.. Native Americans) using adjusted values for
exposure parameters for fish consumption, body weight, and duration of
exposure.  EPA believes the values (short-term, worst-case fish consumption
exposure assumptions of 448 grams as a worst-case, one-day fish consumption
estimate and 2,240 grams as a worst-case, 10-day fish consumption estimate)
provided may be used in setting fish advisories, especially for those
chemicals for which acute noncancer effects are most notable.

      iii.  Final Guidance: For the reasons cited above, the final Guidance
uses a fish consumption rate of 15 grams per day when deriving human health
criteria.  The 15 grams per day is divided into the grams of trophic level 3
fish consumed (3.6 grams) and the grams of trophic level 4 fish consumed  (11.4
grams).  The grams of fish consumed at each trophic level were estimated using
the data in the West et al. (1993) survey.  The results of this analysis are
discussed in section IV.B.3 of this document.  EPA believes separating the
quantity of trophic level 3 and trophic level 4 fish consumed better
represents the potential exposure to consumers.

      f.     BAFs.

      The proposal included a methodology for deriving BAFs, and technical
support documents describing the methodology and BAF derivation for those
chemicals for which human health and wildlife criteria are proposed.  The BAF
methodology in the final Guidance has been modified from the proposal.
However,  BAFs are still required to be utilized in the derivation of human
health criteria.

      The new minimum BAF data required to derive Tier 1 human health criteria
for organic chemicals include either: (a) a field-measured BAF; (b) a BAF
derived from the BSAF methodology, or (c) a chemical with a BAF less than 125
regardless of how the BAF was derived.  For all inorganic chemicals, including
organometals such as mercury,  the minimum BAF data required to derive a Tier I
human health and wildlife criteria include either: (a) a field-measured BAF or
(b) a laboratory-measured BCF.  For the majority of inorganic chemicals, the
BAF is equal to the BCF  (i.e., FCM = 1)  because there is no apparent
biomagnification or metabolism.  The basis for these new requirements are
discussed in section IV of this document.

-------
166    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      g.    Relative Source Contribution.

      i.    Proposal:   In the proposed Guidance, EPA assumed an  80 percent
relative source contribution (RSC) from surface water pathways  (water and
fish) for BCCs, and 100 percent RSC for non-BCCs, in deriving noncancer
criteria/values.  A 100 percent RSC was assumed for all chemicals in deriving
cancer criteria/values.  The proposal explained the reasoning of the
Initiative Committees for choosing each RSC default percentage  (see 58 FR
20870).

      EPA requested comments on the proposed RSCs and on possible alternatives
for derivation of noncancer criteria and values  (see 58 FR 20870).  EPA also
requested public comment on whether any of the options described in the
proposal for use of an RSC in deriving noncancer criteria and values should be
considered in calculating Great Lakes cancer criteria and values (HCVs).

      ii.   Comments:   Many commenters suggested that EPA not use a default of
100 or 80 percent, but to only derive an RSC when exposure data exists.  Other
commenters stated that the reason for a different RSC for BCCs and non-BCCs
for noncarcinogens was not clear.  One commenter suggested we use an RSC of 40
percent for all chemicals, but also acknowledged that this value was chosen
arbitrarily, not based on any specific data set.  Many other commenters
believed that no RSC should be derived at all; that the RSC should be 100
percent for BCCs and non-BCCs.  Several commenters stated that EPA should not
apply an RSC to the development of cancer criteria and values but did not
provide reasons for this opinion.

      EPA agrees with the commenters who suggested that actual data should be
used in developing an RSC when available.  As stated in the 1980 National
Guidelines, to account for exposures from other sources, actual  exposure data
can be subtracted from the RfD  (ADI, as it was called in 1980) to account for
contributions of the pollutant from diet and air  (ADI -  (DT + IN) where DT is
the estimated non-fish dietary intake and IN is the estimated daily intake by
inhalation  (U.S. EPA,   1980).  Therefore, where data are available, if States
or Tribes want to use actual data in developing their RSC, they  may do  so,
following the procedure outlined  in the 1980 National Guidelines.       It is
important to note, however, that EPA's policy on how to use exposure data in
developing an RSC is now under review.  Once EPA has finalized its policy
review on the RSC, EPA will address the application of the RSC during the
triennial review of Water Quality Standards under section 303 of the CWA.
Until such time, the EPA has decided  to apply an RSC of 80 percent to all
noncarcinogenic chemicals  (both BCCs  and non-BCCs).

      After further consideration and review of public comments, EPA does not
support the reasoning  of the Initiative Committees that there is a clear
difference in RSC development for BCCs as opposed to non-BCCs.   While it may
be true that surface water may be the major route of exposure for
bioaccumulatives  (through fish consumption),  even though a pollutant is not
bioaccumulative, it does not preclude the possibility that there may be other
significant sources of exposure.

      With regard to the use of a 80  percent default value, EPA believes that
the assumption helps to provide some  measure of protection against the
possibility that exposures from other sources may contribute to  the overall
exposure of the public to a particular contaminant.  Available  data for
indicate that non-water sources contribute  varying amounts to overall exposure
to a particular  chemical  (U.S. EPA  1982; U.S. EPA 1983).  Such  exposures can
occur  through air and  the  diet.   Since available  data indicate  that such
exposures can and do occur,  but these data  are  often limited  in their ability
to predict with  precision  the relative source contribution, EPA believes it is
prudent not to assume  that all exposure  to  a. pollutant  occurs from one  medium.
The  80 percent default was chosen because  it 'reflects the approximate
contribution from surface water pathways  (fish  consumption) to  the overall

-------
                              Section V: Human Health                           167
exposure to BCCs such as PCBs in the basin  (see section III., Relative Source
Contribution, in the Human Health TSD).

      For nonbioaccumulatives, 80 percent was also chosen as a default value
to account for the other possible non-water sources which may contribute to
the overall exposure of the chemical.  However, actual exposure data may also
be used in the. final Guidance by States and Tribes to calculate a relative
source contribution.  EPA recognizes that the choice of a default value of 80
percent in these cases is fundamentally a policy judgment that criteria
development should reflect the fact that exposures to a pollutant occur
through other media', rather than an empirically-based calculation of the
precise proportion of exposure via water versus non-water sources, since such
values vary on a case-by-case basis.  EPA also acknowledges that use of a 80
percent default for non-BCCs is a conservative measure, however, if other
significant exposures are not accounted for, the criteria could underestimate
overall exposure to the chemical and thus could underestimate the risk of
adverse health effects.  In addition, in the absence of data, it is prudent
and consistent with the health protection goals of the CWA to include a margin
of safety in the event that there are exposures from other sources.  The
important fact, EPA believes, is to take some accounting of other possible
exposure pathways. •

      With regard to the concern raised by some commenters that point sources
should not be expected to compensate for the failure to address other
pollutant sources, EPA does not believe that the relative source contribution
factor in the final methodology unduly burdens point source dischargers.  It
is common practice in EPA programs  (e.g., in establishing maximum contaminant
level goals under the Safe Drinking Water Act) to take into account other
routes of exposure to a chemical when establishing health-based standards for
a particular route of exposure.  If this step is not taken, and EPA were
always to assume that no exposures occurred through other media (in spite of
evidence to the contrary), then the totality of exposures could obviously
result in adverse health effects, contrary to EPA's goal of establishing
standards that insure that such effects do not occur.  EPA agrees, however,   "
that it is important to take steps to address all routes of exposure to
pollutants in order to achieve the greatest overall public health protection
at the least cost.

      iii.  Final Guidance:  The final Guidance provides for the application
of a 80 percent RSC for all noncarcinogenic chemicals.  This is a change from
the proposal, which provided for the application of the 80 percent RSC to
noncarcinogenic BCCs only.  The rationale for the change is explained above.

6.    Minimum Data Requirements/Tier I and Tier II

      In the proposal, EPA established a two tier system, each tier with a
different set of minimum data requirements.  This was established to
facilitate interpretation of State narrative standards and to lead to
development of as many criteria and values as possible.  EPA stated that the
data base necessary to derive Tier I criteria was fairly extensive and could
preclude the development of a permit limit for a chemical of concern, if the
minimum data was not available. In the absence of Tier I criterion, the
permitting authorities needed some mechanism with which to interpret and
ensure that the narrative prohibition against the discharge of toxic
substances in toxic amounts is reflected in water quality-based effluent
limitations.  To address this issue, a Tier II methodology, which requires a
less extensive data base (similar to the Tier II methodology for the
development of aquatic life values), was proposed for the development of human
health values.

      In the proposal, Tier I criteria and Tier II values for human health
were differentiated based on the quantity and quality of toxicological data
only.  There was no. differentiation based on the quantity or quality of BAFs.
After reconsideration, EPA has decided to differentiate the Tier I criteria

-------
168    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

and Tier II values for humans based on both the quantity and quality of
toxicological and bioaccumulation data."  The minimum BAF data required  to
derive a Tier I human health criterion for non-polar organics include either a
field-measured BAF, a BAF predicted from a field-measured BSAF, or a chemical
with a BAF less than 125 regardless of how the BAF was derived.  For inorganic
chemicals including mercury, BAFs derived based on field-measured BAFs  or
laboratory-measured BCFs are considered adequate for derivation of Tier I
criteria because there is little concern with metabolism for the majority of
inorganics.  The rationale for establishing the minimum data requirements for
BAFs is included in section IV.2.b of this document.  Based on this change,
the minimum data to derive Tier I criteria must include both the minimum
toxicological data, discussed below, and the minimum BAF data.  If these data
are not available, then a Tier II human health value must be developed.

      a.    Carcinogens.

      i.    Proposal:  Under the proposal, the methodology for deriving Tier I
criteria and Tier II values for carcinogens  {the human cancer values or HCVs)
was identical.  The only difference between Tier I and II was the weight of
evidence, and the amount and quality of data that was required for use  in
deriving the criteria or values.  The major issue regarding the two Tiers was
how to address Group C chemicals.  The proposal would have provided that Tier
I criteria be set for those types of Group C chemicals which are well
characterized and supported by a well-conducted study.  For those Group C
chemicals in which the cancer study  (or studies) indicate(s) a significant
increase of cancer in test animals but are limited by either: a marginal
statistical correlation between chemical and tumors due to high control tumor
incidence; a weak dose-response relationship; or an incidence of benign tumors
rather than malignant tumors, Tier II cancer values were to be derived  (if
enough data is available to conduct a quantification).  If a cancer
quantification could not be conducted due to lack of data  (number of test
animals, and/or only one dose group of animals has responded, making it
impossible to determine a slope factor) then the chemical was to be assessed
on a noncancer basis and a Tier I or Tier II human noncancer criteria or value
 (HNV) should be developed, if available data exists.

      EPA requested comments on: the procedures proposed for derivation of
Tier I criteria and Tier II values for possible carcinogens  (Group C);  the
alternative of using an additional uncertainty factor  (up to 10) on a
noncancer endpoint for Group C chemicals to provide protection from possible
carcinogenicity; and the alternative of deriving criteria and values for Group
C only through noncancer assessments without an added uncertainty factor for
possible carcinogenicity.

      Another option, which EPA requested comments on, was whether a Tier II
value could be set for an unquantifiable Group C chemical, whether endpoints
other than malignant tumors such as benign tumors or other precursors to
malignant tumors  (such as hyperplastic nodules  or peroxisome proliferation)
could be used to set a value.

      ii.   Comments:  With regard to how Group C chemicals should be
addressed, several commenters believed EPA should set criteria and/or values
for Group C chemicals on a case-by-case basis since the variability in  quality
of the data base which is used in establishing Group C status can be so wide.
Others were split on whether Group C  chemicals should be regulated at all
under Tier II.   Several commenters believed  the data supporting  Group C
chemicals was much too minimal to support Tier II value development.  Other
commenters supported the development  of Tier II values for Group C chemicals
citing the need  to protect the public from potential carcinogens even when the
supporting data  base is minimal.

      With regard  to the use of uncertainty  factors to account for potential
carcinogenesis,  commenters were split on  this issue.  Many found the practice
of using an extra uncertainty factor  indefensible from a  scientific viewpoint;

-------
                              Section V: Human Health                           169
others felt it was a sound policy which should be continued in the Drinking
Water Program and also used in the final Guidance.

      In response to the use of precursors, such as hyperplastic nodules or
peroxisome proliferation, to set a value, some commenters stated that their
use in setting criteria or values was controversial and could be difficult to
defend.  They cited examples where precursors such as hyperplastic nodules or
peroxisome proliferation were not associated with the onset of cancer.

      With regard to how Group C chemicals should be addressed, EPA agrees
with commenters that Group C chemicals should be dealt with on a case-by-case
basis and has changed the final Guidance to reflect this.  As the final
Guidance is written, States and Tribes have the discretion to develop criteria
or values for Group C chemicals based on the overall toxicological data base.
The final Guidance directs that this case-by-case determination be made taking
into account information including data on mutagenicity, genotoxicity,
structure activity, and mode of action.  EPA believes that those Group C
chemicals which act via a genotoxic mechanism (that is through direct
interaction with DNA),  may be most appropriately dealt with through use of a
linearized multistage model (IiMS) or other models which appropriately reflect
this type of mechanism of action (nonthreshold).  If the chemicals does not
interact with DNA, it may be best dealt with as a noncarcinogen and an RfD
should be developed.  (See the updated Human Health TSD, section II - Tier
designations - for guidance on determining whether an agent interacts with DNA
directly.  Several assays which are key to making such a determination are
listed.)

      With regard to the use of uncertainty factors to account for potential
carcinogenesis,  EPA believes the use of an extra uncertainty factor of up to
10 can be justified if there is concern of potential carcinogenesis (i.e.,
equivocal bioassay and genotoxicity results) and that the State or Tribe
should make this determination on a case-by-case basis.  However, as stated
above, a clear determination should be made whether the chemical interacts
directly with DNA.  If this is a clear cut decision (i.e., the genotoxicity
data is not equivocal),  then the use of an extra uncertainty factor may not be
necessary: either the chemical can be addressed as a carcinogen and quantified
using an LMS or other appropriate model or it can be addressed as a noncancer
agent and an ADE is set.  The determination whether to use an extra
uncertainty factor can also be based on which data set is more reliable or
convincing.  If the cancer data is marginal in terms of testing protocol and
statistics, but the noncancer data is well-conducted and unambiguous, it may
be preferable to use the noncancer data in setting a criterion with an extra
uncertainty factor of up to 10 to account for possible carcinogenicity.  EPA
stresses that the entire data base should be used in developing an overall
weight of evidence for human carcinogenicity before choosing a course of
action with regard to selecting a Tier or a risk assessment approach  (cancer
or noncancer).

      With regard to the use of precursors in developing Tier II values, EPA
agrees that the use of precursors as an endpoint may be controversial and
difficult to defend unless a clear determination of mechanism of action is
made implicating a precursor to an eventual tumor incidence.  However, if a
chemical is well studied and the mechanism of carcinogenesis is well
established indicating a clear procession from precursor to malignant tumor,
such endpoints can be used by States/Tribes in developing a Tier I criterion
or Tier II value.

      iii.  Final Guidance:  The final Guidance provides for a case-by-case
determination of carcinogenicity for Group C chemicals with an emphasis on
evaluating the entire data base.

-------
170    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      b.    Non-carcinogens.

      i.    Proposal:  The proposed Guidance also made a clear distinction
between Tier I and Tier II for noncarcinogens. Human Noncancer Values  (HNVs)
for the Tiers I and II are again distinguished on the basis of the available
data base.   The minimum acceptable data base for derivation of a Tier  I
criterion is at least one well-conducted subchronic mammalian study.   The
duration of the study must be at least 90 days in rodents, or 10 percent of
the lifespan of other appropriate species.  The preferred,minimum data point
for decision making is a NOAEL; however, the proposal also allowed the use of
a LOAEL involving mild, reversible effects, which may be considered acceptable
from longer term studies where a NOAEL may not be available. In developing
Tier II values, the proposal stated that the minimum acceptable data base was
a well conducted repeated dose mammalian study of at least 28 days.  However,
the proposal also maintained that the minimum acceptable data point for
decision making on such short term exposure data was a NOAEL.  In addition,
the proposal stated that the study ideally should be designed to observe all
possible systemic effects and include examinations for histopathology.  Data
from studies of longer duration (greater than 28 days) and LOAELs from such
studies were also allowed in some cases for derivation of Tier II values.  EPA
did not want to preclude the use of LOAELs from studies slightly longer than
the required 28-day studies  (such as 30 day tests) if the LOAEL from such a
study represents an effect which is mild, reversible, close to a probable or
actual NOAEL, and is  representative of effects observed over chronic
exposures.

      EPA also recognized in the proposal that when the Tier II methodology is
used to derive a HNV, an additional uncertainty factor of up to 10 may be
applied in deriving the ADE to account for the difficulties in extrapolating
from a short term NOAEL to a long-term NOAEL.  Structure activity
relationships  (SARs), and all other available data on the chemical should be
used to determine the appropriate additional uncertainty factor.

      EPA requested comments on several issues including: the use of a 28-day,
or other subacute study with the use of additional uncertainty factors;
whether the use of the Tier II human health methodology was appropriate;
whether or not even shorter term studies, such as 14 day studies, might
effectively be used in a Tier II HNV approach; and the appropriateness of
using surrogate chemicals employing an SAR evaluation to develop Tier  II
values and the use of SAR as a screen for requiring additional toxicity
information.

      ii.   Comments:  The majority of commenters opposed the use of 14 day
studies, while they were divided on the use of 28-day data for developing Tier
II values.   Many commenters stated the 28-day minimum was scientifically
indefensible since a 28-day study is not designed to characterize subchronic
or chronic risk  (it is less than 10 percent of most test mammals' lifespan)
and that EPA should only develop Tier I criteria using the data base specified
for Tier I.  Other commenters believed the 28-day study was the minimally
acceptable test but they also cautioned that the values derived with 28-day
study data be used only on an interim basis in order to encourage the
development of longer term data.  Other commenters stated that the use of a
LOAEL should not be allowed.  With regard to comments on the overall Tier II
methodology, several commenters supported the Tier II process but indicated
that resulting values should only serve as an interim risk assessment  on a
chemical.  Other commenters opposed the Tier II process stating that if good
data does not exist for a chemical, it is unreasonable to regulate it  on such
a basis.

      With regard to the use of SAR in the development of Tier II values, many
commenters were  in favor of using SAR to evaluate Tier II chemicals but not to
develop  surrogate chemical values.  Commenters believed that SARs are  very
theoretical in nature, without solid data foundation, and could not form the
basis for a permit limit.

-------
                              Section V: Human Health                           171
      EPA believes the use of 28-day data is supportable in the context of
Tier II development.  While such short 'term studies typically do not reveal
evidence of possible adverse effects resulting from longer term exposures,
there is some evidence that  the data can be correlated to longer term studies
(Wiel and McCollister, 1963) .

      EPA acknowledges that the certainty in the overall risk assessment is
not as high as- with 90-day study results since it is necessary to extrapolate
from a short term study to lifetime exposures, and the correlatability of
short term study results to long term study results is not definitive and
quantitative.  However, EPA believes that the number derived using the 28-day
data and an extra uncertainty factor will be adequately protective  (since the
uncertainty factor will further reduce the final value) and will serve as
motivation to obtain longer term data to develop Tier I criteria.  For these
same reasons, EPA believes the Tier II process is a sound approach to
developing interim regulatory values for pollutants with minimal data bases.
      In addition, EPA believes that for the majority of chemicals that will
be found in discharge effluents, there is already adequate data to develop
Tier I criteria.  This is based on a. review of the toxicological data bases
for the 138 pollutants of initial focus.  These are the chemicals identified
by the Steering Committee to be known or suspected of being of primary concern
in the Great Lakes basin (see 58 FR 20843).  Of the 138, EPA estimates that
about 120 have enough data to derive Tier I criteria.  This leaves about 20
chemicals with either insufficient data to calculate a Tier I criterion or no
data to calculate either a Tier I criterion or Tier II value.  Thus, EPA
believes the number of chemicals with Tier II values will be minimal.  (See
section II.C.5 for a more complete discussion.)

      With regard to the use of SAR, EPA agrees that SAR may be useful in
assessing the potential effects of a chemical and may be valuable in selecting
the uncertainty factor for a Tier II value.  However, it is a process which
requires a great deal of scientific judgement and can be open to differing
interpretations.  Because of these concerns, EPA has decided to not require
the use of SAR as the basis for II values.   EPA believes the expertise or the
resources may not exist in many States to use SAR for routine regulatory
purposes.  In addition, the interpretation of SAR data may lead to
inconsistent values, among the Great Lake States. With further research and a
greater confidence in the process, SAR may be used in the future to derive
surrogate chemical Tier II values.

      iii.  Final Guidance:   For the reasons stated above, EPA is maintaining
the methodology for developing Tier II values, which relies upon the use of
28-day study results as the'basis of value development and the use of SAR for
the selection of uncertainty factors.

D.    Criteria Derivation

      The final Tier I human cancer criteria or Tier II value are calculated
as follows:

      HCV  =      RAD x BW	
                  WC + [(FCnj x  BAFTU)  +  (PC™ x BAFTL4) ]

      Where:

      HCV = Human Cancer Value in micrograms per liter  (/jg/L) .

      RAD = RAD in milligrams toxicant per kilogram body weight per day
(mg/kg/day)  that is associated with a lifetime incremental cancer risk equal
to 1 in 100,000.

      BW = Body weight of an average human (BW = 70kg).

-------
172    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      WC = average per capita water consumption  (both drinking and incidental
exposure) for surface waters classified as public water supplies  (WCd = 2
L/day) and average per capita incidental daily water exposure for  surface
waters not used as public water supplies  (WCr =  0.01 liters/day)

      FCnj = mean consumption of trophic level 3 fish by regional  sport
fishers = 0.0036 kg/day

      FCtL4 = mean consumption of trophic level 4 fish by regional  sport
fishers = 0.0114 kg/day

      BAFTL3 = BAF for trophic level 3 fish

      BAFTL4 = BAF for trophic level 4 fish

      The Tier I human noncancer criteria or Tier II value  is calculated as
follows:

     HNV  =  APE x BW x RSC	
             WC +  [(FCTL3 x  BAFTL3)  + (FCru x BAFTL4) ]

      Where:

      HNV = HNV in micrograms per  liter  (jjg/L) .

      ADE = ADE in milligrams toxicant per kilogram  body weight  per day
 (mg/kg/day).

      RSC = RCS factor of 0.8 for  all chemicals  of concern.   This is used to
allow for potential exposure via sources  other than  consumption  of
contaminated water and fish recreational  exposure.   States may develop an RSC
using actual exposure data following the  procedures  specified in the 1980
National Guidelines.

1.    Proposed Criteria and Values

      40 CFR part  132, Table 3, sets forth HCVs  and  HNVs for 18  chemicals
which have been derived using the  final human health methodology.  Note that
for each HCV and HNV, two criteria are provided.   The first  is that which
applies when exposure is from recreational activities and consumption of
aquatic organisms. .The second  is  that which applies when exposure is from
consumption of aquatic organisms,  drinking water and recreational activities.
EPA requested comments on the proposed HCVs and  HNCs in Table 3  of proposed 40
CFR part 132.  The docket for the  final rulemaking contains  technical support
documents describing the details of derivation of each criterion and value.

      The criteria for the chemicals in Table 3  of part 132  have been modified
from the proposal  to reflect changes in the BAFs and the use of  a default RSC
of 80 percent for  all noncarcinogens not  just those  noncarcinogens that are
bioaccumulative chemicals of concern  (BCC).  The rationale  for the changes in
the BAFs are discussed in section  IV of this document.  The  change in the RSC
is discussed in section V.C.S.g above.

      The criterion for heptachlor and pentachlorophenol are not included in
Table 3 of part 132 because  the BAFs for  the chemicals were  estimated using a
laboratory predicted BCF multiplied by a  food chain  multiplier and their BAFs
were greater than  125.  Consequently, they did not meet the  minimum BAF data
requirements for deriving a  Tier  I human  health  criterion as discussed in
 section V.C.6 above.

      The noncancer criterion  for  trichloroethylene  and toxaphene are not
 included in Table  3 of part  132 because  the RfDs needed to derive a human
noncancer  criterion'are currently under  review by EPA.  In the proposal, the
noncancer  criterion were derived  using RfDs that had not been verified by EPA.

-------
                              Section V: Human Health                           173
EPA has decided that it would not be appropriate to include criterion  in Table
of part 132 that do not have a RfD verified by EPA.  For this reason,  the
final Guidance does not include a human noncancer criterion for  toxaphene or
trichloroethylene.  However, even if the noncancer criterion used in the
proposal were retained, the cancer criterion for the two chemicals would be
lower  (toxaphene - cancer criterion is 6.8 x 10"5 A*g/L and noncancer criterion
would be 2.1 x 10~2 pig/L; trichloroethylene - cancer criterion is 29 /xg/L and
noncancer criterion would be 470 /KJ/L) .

      The ADE  (RfD) for chlordane was modified to correct an error in  the
uncertainty factors used in the proposal.  The ADE in the proposal was 5.50 x
10"4 mg/kg/day.  This was based on a NOAEL of 5.5 x 10'2 mg/kg/day and an
uncertainty factor of 100 to account for interspecies variability and
intraspecies variability.  An additional uncertainty factor of 10 should have
been included to account for the lack of an adequate reproduction study and
adequate chronic study in a second mammalian species and the generally
inadequate sensitive endpoints studied in existing studies.  The resulting
uncertainty factor of 1000 is recommended by EPA in its IRIS data base.  The
resulting ADE is 5.50 x 10"5 mg/kg/day and the noncancer criterion for
chlordane is 1.4 x 10"3 /ig/L.  The cancer criterion for chlordane is 2.5 x 10"*
      The NOAEL for hexachloroethane in the proposal was adjusted from 1.0
mg/kg/day to 0.71 mg/kg/day to account for the fact that the test animals were
fed only 5 days a week.  Upon further review of the study used to estimate the
NOAEL, it was determined that the test animals were fed every day of the week
and not 5 days per week as assumed in the proposal.  An adjustment of the
NOAEL in the proposal to account for the 5 day feeding week was therefore not
justified.  Thus, the final Guidance uses a NOAEL of 1.0 mg/kg/day.

      The NOAEL used for deriving noncancer criterion for methylene chloride
in the proposal was 6.47 mg/kg/day.  This was based on a study in which the
NOAELs were 5.85 and 6.47 mg/kg/day for male and female rats, respectively.
The NOAEL for the female rat was used in the proposal as the basis for
deriving a noncancer criterion.  Based on further review, EPA believes that
the NOAEL of the male rats should be used in deriving a noncancer criterion.
EPA decided to use the male NOAEL because it will provide slightly greater
protection and is consistent with the NOAEL used in IRIS.  In either case, the
cancer criterion for methylene chloride is more stringent than the resulting
noncancer criterion,.  The noncancer criterion using the female rat study would
be 1.8 x 103 £jg/L,  the  criterion using the  male  rat study would be 1.6  x 103,
and the cancer criterion is 47
      a.    PCBs  (Human Cancer Value)

      i -    Proposal :  The proposed criterion document for PCBs set a HCV of 3
pg/L for both drinking water sources and non-drinking water sources.  These
criteria were based on a slope factor of 7.7  (mg/kg/day) -1 derived from the
rat bioassay of Norback and Weltman  (1985) .  This study utilized large numbers
of Sprague Dawley rats  (70/sex/dose) in a  chronic Aroclor 1260 feeding study.
Animals were administered 100 ppm for 16 months, followed by a 50 ppm diet for
an additional eight months, then a basal diet for five months. Among the test
animals that survived for at least 18 months, females exhibited a 91 percent
incidence of malignant liver tumors.  In males, corresponding incidences were
four percent for liver tumor and 11 percent for neoplastic nodules.

      ii .   Comments:  Several commenters  were critical of the PCB criterion
and made the following comments : there is  no evidence that congeners other
than those found in Aroclor 1260 are carcinogenic; EPA should pool the data
from all Aroclor studies and develop a geometric mean from these studies; the
re-evaluation of tumors in the Norback and Wellman (1985) study by IEHR
results in a substantially lower cancer potency; and the epidemiology data on
PCBs indicates that the current animal based cancer potency is overly
conservative .

-------
174    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


      With regard to the comment that congeners other than those found  in
Aroclor 1260 are not carcinogenic, as stated on IRIS:  "Although animal
feeding studies demonstrate the carcinogenicity of commercial PCB
preparations, it is not known which of the PCB congeners in such preparations
are responsible for these effects, or if decomposition products, contaminants
or metabolites are involved in the toxic response."  Nominally, it appears
that animal studies with Aroclor 1260 are the only studies indicating
carcinogenic response.  However, there are indications that Aroclor 1254 may
also be carcinogenic.  A study by NCI (1978) reported carcinomas of the
gastrointestinal tract in Fischer rats treated with Aroclor 1254, however the
incidence was not statistically significant.  While it is not statistically
significant, the presence of such a response still raises the concern
regarding the carcinogenicity of PCB mixtures other than Aroclor 1260.   The
EPA believes it is not reasonable to develop a criterion for each PCB Aroclor
mixture.  PCBs are mixtures of chlorinated biphenyls.  Each mixture may
contain up to 209 possible individual compounds.  Each of the mixtures would
be expected to contain all combinations of chlorinated compounds even though
some of them only in small or trace amounts.  Since it is not known which
compound of the 209 is clearly responsible for eliciting a cancer response, to
conservatively protect against the potential for carcinogenicity associated
with all PCB mixtures, EPA has chosen to base criteria on study results using
Aroclor 1260.

      EPA disagrees*with the commenters who suggest EPA should pool all the
cancer data from all the available congener studies in developing a criterion
for PCBs for following reasons:  The Norback and Wellman  (1985) study is
judged by EPA as acceptable in design and conduct compared to other available
studies.  As stated on IRIS, "The estimate  (slope factor) based on the data of
Norback and Weltman  (1985) is preferred because Sprague Dawley rats are known
to have low incidence of spontaneous hepatocellular neoplasms.  Moreover, the
latter study spanned the natural life of the animal, and concurrent
morphological liver studies showed the sequential progression of liver  lesions
to hepatocellular carcinomas."   Normally EPA will pool study data only if
available studies are considered of marginal quality.  The Norback and Wellman
study is considered a good study and therefore the pooling of studies is not
needed.

      In response to the reevaluation of the tumor data by IEHR, the IEHR
potency factor was 5.1  (mg/kg/day)-1 as compared to the EPA derived cancer
potency factor of 7.7  (mg/kg/day)-1.  The IEHR reevaluation was a reread of
the histopathology slides from Norback and Wellman study, in which several
lesions which were originally interpreted as malignant lesions were
reinterpreted as non-malignant lesions  (neoplastic nodules).  The EPA is
currently reviewing the IEHR data for PCBs.  Until the EPA can fully assess
the validity of the'conclusions drawn in the IEHR reevaluation, it would be
premature to make the suggested changes in the criterion.

      In response to the epidemiology data indicating that humans are less
sensitive than test species to PCBs,  EPA does not believe the epidemiological
results are as conclusive as the animal results from the Norback and Wellman
study.  As presented on IRIS, the human carcinogenicity data is considered
inadequate.  Although there are many epidemiological studies, the data  are
inadequate due to confounding factors.  The factors noted in IRIS are:
population differences  in alcohol consumption, dietary habits, ethnic
composition, contamination of PCBs by dibenzofurans, and exposure of workers
to other known carcinogens.  It is EPA's longstanding practice to rely  upon
studies in animals for  risk assessment in the absence of adequate human data.

      iii.  Final Guidance:  For the  reasons stated above, EPA continues to
believe that the study  by Norback and Weltman is the best available study to
use for deriving a human health criterion for PCBs.

-------
                              Section V: Human Health                           175
      b.    TCDD.

      In the proposal, EPA stated that it is currently conducting a major new
dioxin research and analysis effort, the results of which could not be
reflected in the proposed Guidance.   EPA also stated that if the results of
this study became available prior to the finalization of the rule, EPA
expected to publish-a notice of availability and solicit comment on whether
the rule should be modified to reflect this new information.   EPA invited
comment on the approach it should take to establishing dioxin criteria pending
completion of its ongoing dioxin studies.

      Many commenters stated that EPA should not develop criteria for dioxin
until the dioxin reassessment is final.  Other commenters stated that EPA
should not delay the publication of dioxin criteria until the reassessment is
finalized.

      EPA agrees with commenters that it is important that EPA develop a
dioxin criterion regardless of the status of the dioxin reassessment.  Dioxin,
from all indications, is one of the most potent carcinogens and must be
regulated with the most recent available data at hand.  Once the dioxin
reassessment is final, EPA will revisit the dioxin criterion, and make changes
if needed.  EPA's proposed dioxin reassessment was made public on September
13, 1994.  The final dioxin reassessment is anticipated sometime in late 1995.

      i.    TCDD - Noncancer Criterion

       (A).  Proposal:  In the proposal, EPA developed a Tier I criterion of
0.1 pg/L for drinking water and nondrinking water sources based on a
reproductive study by Bowman et al.  (1989)  on rhesus monkeys which indicates a
LOAEL based on behavioral effects at 25 ppt (0.67 ng/kg/day) and a NOAEL at 5
ppt (0.13 ng/kg/day).  The ADE was developed by dividing the NOAEL of 0.13
ng/kg/day by an uncertainty factor of 100 (10X for intraspecies variability
and 10X for interspecies extrapolation).

       (B).  Comments:  Several commenters believed the intra- and interspecies
adjustment of 10 was unjustified since the rhesus monkey data was used.  They
argued that the rhesus monkey is very close to humans in terms of
pharmacokinetics and therefore an interspecies uncertainty factor of three is
warranted instead of the usual uncertainty factor of 10 which is applied for
interspecies differences between humans and less similar mammals such as
rodents.

       (C).  Final Guidance:   With regard to the noncancer criterion, it is
EPA's judgement to apply an intraspecies uncertainty factor of 10 to account
for variability and sensitivity within the human population.   EPA also
believes that the interspecies uncertainty factor of 10 is justified because
the study groups were very limited in size and the statistical and biological
significance of the findings are unclear.   In addition, metabolic and
pharmacokinetics parameters for humans and rhesus monkeys may be sufficiently
different.  For these reasons, EPA believes an interspecies uncertainty factor
of 10 is justified.

      ii.   TCDD - Cancer Criterion

       (A)   Proposal:  The proposed criterion document for dioxin presented a
human cancer value of 0.01 pg/L for drinking and nondrinking water sources.
These criteria were based on a slope factor of 7.5 x 104  (mg/kg/day)-1  based
on the pooled significant tumors in female rats of Kociba, et al. (1978) with
the liver tumor reevaluation of the Pathology Working Group  (Sauer,  1990) .

       (B).  Comments:  Commenters stated that the dioxin criteria should
reflect the 1986 cancer guidelines  (U.S. EPA,  1986), not the draft EPA cancer
guidelines which EPA is in the process of revising.

-------
176    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      (C).   Final Guidance:  EPA agrees that the 1986 cancer guidelines are
appropriate for use in developing the dioxin criterion, and believes this is
appropriate until the revised cancer guidelines are peer reviewed, publicly
reviewed and finalized.

      c.    Mercury.

      i.    Proposal:  EPA proposed a criterion of 2 ng/L for mercury for both
drinking water and nondrinking water sources.   This criterion was based on a
LOAEL of 3 ug/kg/d and an uncertainty factor of 50.  The LOAEL was based on
several studies that have shown neurological symptoms of mercury toxicity in
adults at blood levels of mercury in the range of 200 to 500 ng/ml.  The 200
ng/ml mercury levels in blood have been associated with an oral intake  in
adults of 3 ug/kg/d of mercury and with adult hair concentrations of 50 ug/g.
The uncertainty factor of 50 is composed of a 10-fold factor to adjust  the
LOAEL to a NOAEL and an additional 5-fold factor to ensure the criterion will
provide protection from the potential fetal effects of mercury exposure via
maternal ingestion of mercury contaminated fish.

      ii.   Comments:  Several commenters agreed with the EPA criterion and
the choice of the study and uncertainty factors. Other commenters agreed with
the mercury criterion but disagreed with the basis for developing the
criterion.   These commenters stated that the LOAEL should be based on studies
indicating fetal effects on the central nervous system occur at a LOAEL of 10
ug/g in maternal hair.  They argue that if the 10 ug/g LOAEL is used, the 5-
fold uncertainty factor to ensure the criterion will provide protection from
potential fetal effects would not be needed.

      iii.  Final Guidance:  EPA continues to believe the adult LOAEL of 3
ug/kg/d should be used in the derivation of criterion instead of the 10 ug/g
maternal hair concentrations suggested by the commenter.  EPA believes  the
adult effects are more clearly delineated from the available data than  the
fetal effects and thus the use of a LOAEL of 3 ug/kg/d  (50 ug/g adult hair
concentrations) is appropriate.  The LOAEL of 10 ug/g maternal hair
concentrations is predicted and therefore can be viewed as a somewhat less
reliable endpoint upon which to base a criterion than the adult endpoints.
EPA continues to believe the 5-fold uncertainty factor is justified to protect
central nervous systems development during the sensitive fetal life stages.
In addition, as discussed in section 5.a. above, EPA has assumed a body weight
of 65 kg (as opposed to 70 kg) for mercury.  The resulting Tier I mercury
criterion is therefore 1.9 ng/L, which is slightly less than the proposed
criterion of 2 ng/L.

      Since the proposal, EPA's Rf>  work  group has  recently revised the  RfD,
using an effect level of 1 ug/kg/d and using an uncertainty factor of 10 to
account for within-human variability and for an insufficient data base.  The
resulting RfD is 0.1 ug/kg/d which is higher than the proposed RfD  (ADE) of
0.06 ug/kg/d.  However, because the new RfD of 0.1 ug/kg/d was not verified
until early February 1995, it was not possible to publish the data, request
comment, and revise the final Guidance, if needed, prior to promulgation of
the final Guidance. Consequently, EPA plans to publish a Notice of Data
Availability after the publication of the final Guidance with the new mercury
assessment for human health and will change  the final mercury criteria  for
human health if appropriate.

E.    Relationship of the  Great Lakes Initiative Guidelines to National
      Guidelines Revisions

      1.    Proposal:  As  stated  in the proposal, much of the Great Lakes
methodology for deriving human health criteria was based on the  1980
methodology and advances in  the science since 1980.  Concurrent with the
development of the fi-nal Guidance, EPA is also in  the process of reviewing and
revising the 1980 National Guidelines which would  apply to development  of EPA
National water quality criteria under section 304(a) of the CWA.    It is

-------
                              Section V: Human Health                           177
expected that a proposed revision to the National guidelines will be published
in the Federal Register in  1995 and that there will be a separate opportunity
for public comment on that proposal.  In the proposal, EPA discussed the
possibility of making no change from the current 1980 methodology.
Accordingly, EPA requested comments on the possibility of retaining the
approach set forth in the 1980 National Guidelines with respect to each
individual component of the proposal that differs from the current Natiorial
guidelines.

      2.    Comments:  A few commenters stated that EPA should retain the 1980
methodology but these commenters did not  discuss why EPA should take such an
approach.  The majority of commenters were in favor of EPA using the best
science in conducting any risk assessment which becomes the basis for water
quality criteria.  Other commenters stressed that EPA must strive for
consistency between the methodology presented in the proposal and the revised
national methodology.

      3.    Final Guidance:  EPA believes it has presented a methodology which
reflects the best science to date.  EPA  also believes it is essential to
revise the 1980 National Methodology to reflect the latest science and policy
of the EPA and the scientific community.  Prior to revising the Guidelines EPA
will request comments on the latest developments in draft EPA policy such as
the draft revised Cancer Guidelines, the proposed policy on body weight to
surface area scaling, and the proposed RSC policy.   Dntil EPA's revised
policies are finally adopted, after consideration of public comments, EPA
believes it would be premature to adopt such revisions as part of the final
Guidance.

      With regard to fish consumption, EPA has not yet revised its national
policy on developing a fish consumption rates  However, in the final Guidance,
EPA believed it was appropriate to develop a regional fish consumption rate
for the protection of the population of the Great Lakes basin and therefore
adjusted the national fish consumption rate accordingly with regional
consideration in mind.

F.    Comparison with the CWA and Great Lakes Water Quality Agreement

      The Great Lakes Critical Programs Act of 1990 (CPA) states that the
proposed Guidance shall be no less restrictive than the provisions of the CWA
and national water quality criteria and guidance.  The CPA also specifies that
the final Guidance is to conform with the objectives and provisions of the
Great Lakes Water Quality Agreement (GLWQA).  The discussion below addresses
conformance of the final human health methodologies and criteria with these
requirements.

      1.    Tier I Criteria/Methodology

      a.    Comparison with the CWA.

      Under the authority of section 304(a)(1) of the CWA, EPA established the
1980 National Guidelines, to be used in deriving National human health
criteria.  EPA believes that although the final Tier I human health criteria
methodology and the criteria are not identical to the 1980 National Guidelines
and individual National criteria in all details, they are generally no less
restrictive.

      First, as discussed above in this section of the document, EPA finalized
Tier I human health criteria for 18 pollutants for which National criteria
exist.  These pollutants include a broad selection of pollutants of initial
focus proposed by the Initiative Committees to test the proposed methodology.
Although the final Guidance includes only these 18 pollutants while National
human health criteria are currently available for 91 pollutants, EPA believes
that this approach will not result in less stringent levels of control.  This
is because under the implementation scheme presented today, Great Lakes States

-------
178    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


would be required to derive criteria and values for these pollutants and  for
all other pollutants except those listed in Table 5 of part 132  whenever
sufficient data exist to meet Tier I or Tier II minimum data requirements and
the State determines that it is necessary to control these pollutants.  Thus,
the scope of the final Guidance in terms of pollutants covered is actually
broader than the current National guidance.

      Furthermore, because the Tier I criteria for human health  assume  a
higher fish consumption rate than the national criteria, use BAFs rather  than
BCFs to calculate fish tissue residues, and include relative source
contributions for all noncarcinogens the final numeric criteria  are equivalent
to, or more restrictive than, the current national criteria.

      b.    Conformance with the GLWOA.  For the reasons stated  in section
III.D  (Aquatic Life) of this document, EPA believes that the final Guidance
conforms to the General Objectives of the Agreement regarding the elimination
or reduction of discharges into the Great Lakes System.  For the 18 pollutants
for which Tier I human health criteria have been derived, the final Guidance
criteria are more stringent than the water quality criteria presented in  the
Agreement, except for lindane.  The GLWQI criterion for lindane  is based  on
noncancer effects. EPA is currently reviewing the carcinogenicity of lindane.
EPA believes that a.cancer criterion for lindane would be lower  than the  GLWQA
for lindane.

2.    Tier II Criteria/Methodology

      a.    Comparison with the CWA.  EPA's current guidance and regulations
for water quality standards contain nothing directly analogous to the two-tier
approach proposed today for human health.  States currently have very broad
discretion when regulating pollutants that are subject only to narrative
criteria.  EPA believes that the final Guidance is more rigorous than the
current National requirements in this area because the Tier II methodology
derives generally more conservative values for non-cancer criteria to
compensate for greater uncertainty in the data base.   Based on  studies done
to date, EPA expects that Tier II values will be more stringent  than existing
standards for these pollutants in most cases.  Further, this approach imposes
a structure to the process of translating narrative criteria into numeric
values.  Finally, the final Guidance will result in more uniform control  of
pollutants lacking National standards in the Great Lakes States.

      b.    Conformance with the GLWOA.  EPA believes that the Tier II
methodology is consistent with the General Objectives of the Agreement.
Moreover, it serves.as a translator mechanism of the States' narrative  water
quality standards.  The Tier II methodology will enhance regulatory efforts in
the Great Lakes basin, will serve its purpose of promoting consistency  in the
regulation of toxics in the Great Lakes basin, and is therefore  also in
Conformance with the Agreement.

G.    References

      Bowman, R.E., et al. 1989.  Chronic dietary intake of 2,3,7,8-
tetrachlorodibenzo-p-dioxin  (TCDD) at 5 or 25 parts per trillion in the
monkey: TCDD kinetics and dose-effect estimate of reproductive toxicity.
Chemosphere. 18(1-6); 243-252.

      Cantor, K.P., R.  Hoover, P. Hartage.  1987.  Bladder cancer, drinking
water source, and Tap water  consumption: A  case control study.   J. National
Cancer  Institute. 79(6): 1269-1279.

      Connelly,  N.A., T.L. Brown  and  B.A. Smith.  1990.  New York Statewide
Angler  Survey, 1988.  New York State  Department of Envirnomental Conservation,
Albany, NY.

-------
                              Section V: Human Health                           179
      Dedrick, R.L/1973.  Animal Scale Up.  J. Pharamcokin. Biopharm. 1:435-
461.

      Ershow, A.G. and K.P. Cantor, 1989.  Total Water and Tapwater Intake' in
the United States: Population-Based Estimates of Quantities and Sources,
National Cancer Institute, Bethesda, MD.

      Fiore, B.J. et al., 1989.  Sport fish consumption and body burden levels
of chlorinated hydrocarbons: A study of Wisconsin anglers, Archives of
Environmental Health, 44:82-88.

      Freireich, E.J., E.A. Gehan, D.P. Rail, L.H. Schmidt, and H.E. Skipper.
1966.  Quantitative comparison of toxicity of anticancer agents in mouse, rat,
hamster, dog, monkey and man.  Cancer Chemother.  Rep. 50:219-244.

      Interagency Regulatory Liaison Group. 1979.  scientific Basis for
Identification of Potential Carcinogens and Estimate of Risks.  Journal of
National Cancer Institute. Vol. 63. pp. 241-268.

      Kociba, R.J. et al.  1978.  Results of a two-year chronic toxicity and
oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats.  Toxicol.
Applied Pharmacol. '46:279-303.

      NCI  (national Cancer Institute). 1978. bioassay of Aroclor 1254 for
possible carcinogenicity. CAS No. 27323-18-8.  NCI Carcinogenesis Tech. Rep.
Ser. no. 38.

      Norback, D. and R.H. Weltman.  1985.  Polychlorinated biphenyl induction
of hepatocellular carcinomas in the Sprague-Dawley rat.  Environ. Health
Perspectives.  60:97-105.

      Pinkel, D.. 1958.  The use of body surface area as a criterion of drug
dosage in cancer chemotherapy.  Cancer Res. 18:853-856.

      Sauer, R.M. 1990.  Pathology Working Group: 2,3,7,8-Tetrachlorodibenzo-
p-dioxin in Sprague-Dawley rats.  Pathco, Inc.  Submitted to the Maine
Scientific Advisory Panel.

      U.S. Environmental Protection Agency.  1980.  Water Quality criteria
Availability, Appendix C - Guidelines and methodology used in the preparation
of health effects assessment Chapters of the Consent Decree Water Quality
Criteria Documents.  Federal Register. Vol. 45, November 29, 1980, 79347-
79357.

     U.S. Environmental Protection Agency. 1982.  Dichloromethane  (Methylene
Chloride).  Occurrence in Drinking Water, Food and Air.  Prepared by JRB
Associates Under EPA Contract No. 68-01-6388.  Office of Drinking Water,
USEPA.  July 11, 1982.

     U.S. Environmental Protection Agency. 1983.  Benzene.  Occurrence in
Drinking Water, Food and Air.  Prepared by JRB Associates Under EPA Contract
No. 68-01-6388.  Office of Drinking Water, USEPA.  November 23, 1983.

      U.S. Environmental Protection Agency. 1986.  Guidelines for Carcinogenic
Risk Assessment.  Federal Register, vol. 51, No. 185, September 24, 1986,
33992-34002.

      U.S. Environmental Protection Agency. 1989.  Exposure Factors Handbook,
Washington, DC, Office of Health and Environmental Assessment.  EPA/600/8-
89/043.

      U.S. Environmental Protection Agency. 1991.  General Quantitative Risk
Assessment Guidelines for Noncancer Health Effects.  External Review Draft.
EPA Document No. ECAO-CIN-538.  Feb. 1991. pp. 1-3.

-------
180    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


      U.S. Environmental Protection Agency-  1992.  Draft Report:  A Cross-
species 'Scaling Factor  for Carcinogenic  Risk Assessment based on Equivalence
of mg/kg3'4 per  day.   Federal  Register.  June 5, 1992, 57 FR 24152.

      U.S. Environmental Protection Agency.  1994a.  Notice of data
availability and request for  comments. Federal  Register.  Vol. 59,  No. 167,
August 30, 1994.  44678-44684.

      U.S. Environmental Protection Agency.  1995.   Fish Consumption Estimates
Based on  the 1991-1992  Michigan  Sport  Anglers Fish Consumption Survey. Final
Report. EPA Contract No. 68-C4-0046, SAIC Project  No.  01-0813-07-1676-040.
February  21, 1995.

      West, P., M. Fly, R. Marans, F.  Larkin and D.  Rosenblatt.  1993.  1991-
1992 Michigan Sport-Anglers Fish Consumption Study.   Final report to the
Michigan  Great Lakes Protection  Fund,  Michigan  Dept.  of Natural  Resources.
University of Michigan, School of Natural Resources.  Natural Resources
Sociology Research Lab.  Technical Report #6.   May 1993.

      West, P.C. et al. 1989.  Michigan  Sport Angler Fish Consumption Survey:
A Report  to the Michigan Toxic Substance Control Committee,  University of
Michigan  Natural resources Sociology Research Lab.  Technical Report #1, Ann
Arbor, Michigan, MDMB Contract # 87-20141.

-------
                                Section VI: Wildlife                             181
                           VI.   WILDLIFE
A.    Introduction
      For the purposes of the final Guidance,  "wildlife" is defined as all
non-domesticated species in the taxonomic classes Aves and Mammalia (birds and
mammals).   The final Guidance for deriving wildlife criteria is provided in
appendix D to part 132.   The final Great Lakes Water Quality Initiative
Technical Support Document for Wildlife Criteria (EPA 820-B-95-009)  (Wildlife
TSD) and the final Great Lakes Water Quality Initiative Criteria Documents for
the Protection of Wildlife:  DDT and Metabolites; Mercury; 2,3,7,8 TCDD; and
PCBs (EPA 820-B-95-008)  (Wildlife Criteria Documents),  which provide the data
and the derivation of each individual criterion, are available in the docket
for this rulemaking.

      As stated in the proposal and in section I (Background)  of the final
Guidance,  wildlife in the Great Lakes are at risk from contaminants in Great
Lakes water.  To address this problem, the proposal presented a methodology
for deriving criteria and values that would likely be protective of wildlife
within the Great Lakes basin.  EPA requested comments and asked for additional
information on several topics relating to the methodology.  Described below is
a summary of the proposal for that topic, significant comments received (with
EPA's response to each), and the requirements of the final Guidance.
Responses to all comments received are provided in the Response to Comments
Document,  which is included as part of the docket for the final Guidance.

      Currently, there is no National methodology for the development of water
quality criteria for the protection of wildlife comparable to the proposed
Guidance;  however, there is a mechanism for consideration of wildlife impacts
within the 1985 National aquatic life criteria guidelines (Stephan et al.,
1985).   While the aquatic life methodology provides a mechanism to protect
against bioaccumulation of a compound within a food web, it has several
limitations.  The current approach in the national aquatic life guidelines to
provide protection to wildlife is the use of the final residue value (FRV),
which represents the highest prey tissue concentration of a toxicant which
will not produce an adverse effect in the consumer organism.  The FRV is
derived through one of three approaches:  either from a Food and Drug
Administration  (FDA) action level; a long-term wildlife field study; or a
chronic wildlife feeding study.  Because an FDA action level is intended to
protect humans, and not wildlife, an FRV derived from an FDA action level for
a specific contaminant does not ensure protection of wildlife species which
may consume contaminated aquatic organisms as a larger portion of their diet
or exhibit a greater sensitivity than humans.   While an FRV may be more
adequate based on a*long-term wildlife field study or a chronic wildlife
feeding study, there are very few chemicals for which these studies are
available.  In addition, such studies are frequently only conducted after the
chemical has been theorized to be responsible for a problem in the
environment.

      In cases where no FRV is available, biomagnification of a chemical into
the higher trophic levels of a food web, and potential impacts on these
wildlife species, is not considered in the derivation of the aquatic life
criterion.

      EPA has begun a separate effort to derive National wildlife criteria.
Following the release of the 1987 General Accounting Office (GAO) report
entitled "National Refuge Contamination is Difficult to Confirm and Clean Up,"
(GAO,  1987), EPA began to work cooperatively with the U.S. Fish and Wildlife
Service to develop methods for deriving National wildlife criteria.   The
wildlife criteria efforts carried by the Initiative Committees have been

-------
182    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

coordinated with the on-going National efforts. Because the effects on
wildlife may not be accounted for in other methodologies for deriving water
quality criteria, the proposed methodology was developed to address this
deficiency-  The proposed methodology contained both an effect and an exposure
component to account for varying species susceptibility to chemical
intoxication, and dissimilar feeding and drinking behaviors and natural
histories.

B.    Scope of Methodology

      1.    Proposal:  The proposed wildlife methodology was based largely on
the noncancer human health paradigm described in appendix C to part 132.  The
methodology, like the aquatic life and human health methodologies, included  a
two-tiered approach for deriving wildlife criteria and values, as described  in
the proposal  (58 FR 20878).  The main difference between the two tiers  in the
proposed Wildlife methodology was the extent of the data base needed to
generate a final wildlife criterion or value.  In order to generate a Tier I
criterion, data frofh both mammals and birds were required.  Separate values
for mammals and birds were developed with the lower of the two values being
the final Tier I criterion.  In order to generate Tier II values, data  from
either mammals or birds were required.  This value was modified by an
interclass uncertainty factor  (UF) to account for sensitivities between
taxonomic classes, and was then applied as the Tier II wildlife value.
Because the rationale behind Tier II values was that they were to be typically
more stringent than Tier I criteria, due to the smaller data set available,  it
was expected that the magnitude of some of the UFs applied in the derivation
would be larger.

      2.    Discussion of Comments

      Comment:  Some commenters believed that the focus of the wildlife
methodology is misdirected because other factors, such as habitat destruction
and exotic species, are more important than toxic contaminants in affecting
species' viability in the Great Lakes System.

      Response:  EPA disagrees that the wildlife methodology focus is
misdirected.  As discussed in section I of this document, research on wildlife
species resident in the Great Lakes indicates that some wildlife populations
remain threatened in areas of high contamination by toxic chemicals.  In  the
Great Lakes, reproductive  impairment  of numerous wildlife species has been
correlated with  the presence of polychlorinated biphenyls  (PCBs), p,p'-
dichlorodiphenyltrichloroethane  (DDT) and its metabolites, and other
contaminants  (58 FR 20806).

      Because many wildlife  species are at  the top of the aquatic food  web,
current water quality criteria derived to protect fish may be inadequate  to
protect wildlife who consume contaminated fish.  Wildlife are especially  at
risk from chemicals which  biomagnify  because they are frequently exposed  to
very high levels of,these  contaminants because many species feed primarily
from aquatic food webs.  For this reason, emphasis was placed on selecting
piscivorous wildlife species  (i.e., those which eat fish) for the derivation
of wildlife  criteria as to represent  species likely to experience significant
contamination  through an aquatic  food web.  Wildlife species may also be
uniquely  susceptible to some chemicals, as  compared to aquatic species.

      EPA acknowledges that  other factors,  such as habitat destruction  and
introduction of  exotic species are  impacting species viability in the Great
Lakes System.  EPA has programs  in place that are attempting to  address many
of these  problems; however,  this  does not mean that toxic  chemical discharges
into the  Great Lakes basin are not  also adversely affecting species  viability.
As discussed in  section I, and in U.S. EPA  (1995b), effects of toxic chemicals
on wildlife  species  continue to  be  a  problem.

-------
                                Section VI: Wildlife                             183
      CommentB:  Many commenters expressed general concerns over the
scientific soundness of the proposed methodology.  In particular, some were
critical of the use of the noncancer human health paradigm as a model for
ecological risk assessment (which might include an assessment of all
ecological stressors on all components of the ecosystem).  Those commenters
were also concerned-that the effect component of the noncancer human health
paradigm methodology is based on individual-level, rather than population-
level measurement endpoints.   Some commenters recommended either delaying
implementation of the methodology, or making the methodology just guidance.

      Responses:  EPA agrees that the use of the noncancer human health
paradigm does not consider all potential stressors on an ecosystem.  However,
the final wildlife methodology is not designed to be a comprehensive model for
assessing all ecological risk to the entire Great Lakes ecosystem.  Instead,
the intent of the methodology is to initially focus attention on those avian
and mammalian species in the System which are likely to experience significant
exposure to contaminants through aquatic food chains.  The methodology
specifically excludes reptiles and amphibians because there is currently
insufficient toxicity data for these species and insufficient understanding of
exposure routes to estimate risks from toxic contaminants to include these
groups in the methodology.  While it would be better to provide a
comprehensive ecological risk assessment approach for chemical and non-
chemical stressors to the System, it is currently not possible given the many
data gaps and enormous resources required to develop such an approach.  As
explained below, EPA finds the noncancer human health paradigm to be
appropriate for the avian and mammalian species on which the method focuses.

      EPA considered not including the methodology in the final Guidance until
a more comprehensive multi-stressor risk assessment approach could be
designed.  However, this option was rejected because such a program will take
many years to develop and a sound wildlife methodology was available at this
time to provide protection to those species at greatest risk from persistent,
bioaccumulative pollutants.  EPA selected a reasonable approach to address the
adverse ecological effects from toxic contaminants in the Great Lakes System.
In addition, based on the results from the two National meetings and the April
1994 EPA Science Advisory Board  (SAB) (U.S. EPA, 1994a) commentary (discussed
below),  EPA concludes that the paradigm is a scientifically reasonable
approach to address'impacts from bioaccumulative compounds on avian and
mammalian species in the Great Lakes at this time.

      During the development of the wildlife methodology, EPA hosted two
public meetings held in December 1989 and April 1992,  (U.S. EPA, 1989, 1994b)
as part of a national effort to develop methodologies to protect wildlife
criteria.  During both of these meetings, there was general consensus that the
proposed Guidance methodology was fundamentally sound from bioaccumulative
contaminants.  In addition, other concepts developed for the national program
were extensively used in the development of the wildlife portion of the final
Guidance.

      Finally, EPA discussed the use of the proposed paradigm for developing
water quality criteria to protect wildlife with EPA's SAB in February 1992 and
April 1994  (U.S. EPA, 1992, 1994a).  The report from the February 1992 SAB
(U.S. EPA, 1992) meeting indicated concern with the wildlife criteria concepts
being formulated around the perceived requirements of the noncancer human
health paradigm, which might be inadequate for wildlife.  In response to the
SAB's commentary, EPA made several changes that were discussed in the proposed
Guidance  (see 58 FR 20882) .  The April 1994 SAB commentary (U.S. EPA, 1994a)
stated that, while the use of the noncancer human health paradigm for the
development of wildlife criteria is in the early stages of development, it
promises to be an innovative and valuable new method for understanding the
fate and effects of contaminants in the environment.  Based on the changes
made to the methodology in response to the February 1992 SAB commentary (U.S.
EPA, 1992) and the support for the use of the methodology expressed by the
report from the April 1994 SAB commentary  (U.S. EPA, 1994a),  the paradigm

-------
184    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

being pursued is appropriate for the  species and stressors EPA is currently
addressing.

      EPA also does not agree that the wildlife methodology focuses too
extensively on the protection of individuals.  The methodology focuses  on
population-level impacts by restricting the toxicological measurement
endpoints on which .a criterion is based to those that are likely to adversely
affect populations.  If these toxic responses were observed in wildlife
populations in the Great Lakes System, the continuation of breeding
populations of the wildlife species could be jeopardized.  The SAB  (U.S. EPA,
1994a) endorsed the basis of the wildlife criteria on the protection of
wildlife populations from the direct  effects of chemical stressors.

      There are two distinctions between the noncancer human health paradigm
and the wildlife approach used in the proposal.  Because the wildlife approach
is designed to protect populations and not individuals, the wildlife paradigm
does not include an intraspecies UF (although exceptions can be made in cases
where toxicological or exposure data  suggest that species listed pursuant to
section 4 of the Endangered Species Act will not be protected by system-wide
criteria) to ensure.better protection of toxicologically sensitive members of
a given population.  Further, the selection of toxicological endpoints  in the
wildlife methodology is restricted to gross endpoints likely to adversely
affect population dynamics  (i.e., reproductive or developmental effects).
This approach is consistent with the  recommendation from the SAB  (U.S.  EPA,
1994a), but is different from the human health methodology which focuses on
the effects on individuals.  This is  illustrated by comparing the Human Health
criteria documents with the DDT section in the final Wildlife Criteria
Documents.  Both documents reference  the same study, but the same dose  level
considered a No Observed Adverse Effect Level  (NOAEL) for wildlife criteria
derivation, based on reproductive endpoints, is cited as a Lowest Observed
Adverse Effect Level  (LOAEL) in the noncancer human health criteria document,
based on liver lesions as an endpoint.  Examples of acceptable endpoints are
made available in the final Wildlife  Criteria Documents, as well as in  the
proposal (58 FR 20882).

      Comments:  The use of Tier II wildlife values was generally criticized
by many commenters' who believe there  were technical weaknesses with the
methodology and because of insufficient data being used to derive these
values.  Some commenters were concerned with the resource requirements  to
derive wildlife criteria or values for the entire universe of pollutants.
Some commenters supported the use of  Tier II values for wildlife.

      Responses:  The methodology for deriving Tier II values is sound.  EPA
agrees, however, with commenters that the data for deriving wildlife values is
currently limited.  In addition, EPA  agrees with those commenters who
cautioned against advancing too rapidly with a new methodology before
additional field validation can be made.  Therefore, during review of comments
received on the proposal, EPA reconsidered the scope of application of  the
wildlife methodology.  EPA decided to limit the methodology to bioaccumulative
chemicals for which the determining route of exposure is through the diet.
EPA still agrees, however, that the methodology can be modified to derive
reasonable wildlife-values where other exposures become significant.  For non-
bioaccumulative chemicals, it may be  more appropriate to select different
representative species which are better examples of wildlife species with the
greater exposure for a given chemical.

      In addition, EPA decided to limit the methodology to require developing
only Tier I criteria for several reasons, including concerns that a Tier II
value, which is based on toxicity data from only one taxonomic class, could
not be protective of wildlife species in other taxonomic classes when there is
evidence of wide differences in sensitivities across classes.  Further, a Tier
II value is based on an interclass UF that  is not used in deriving a Tier I
criterion, making the uncertainty inherent  in the value potentially
unreasonably greater.   Finally, the proposed wildlife methodology was a new

-------
                                Section VI: Wildlife                             185
approach to ecological risk assessment for wildlife.  EPA believes, however,
that the taxonoznic class-specific wildlife values provide adequate protection
for species of that-class, even where toxicity data from other classes are
missing.

      3.     Final Guidance:  Based on the above discussion, the use of the
proposed methodology is appropriate and the methodology does not emphasize on
individual effects.

      However, EPA has decided to modify the scope of the final wildlife
methodology.  The final Guidance requires the States or Tribes to use the
methodology to derive Tier I numeric wildlife criteria for only the
bioaccumulative chemicals of concern  (BCCs)  listed in Table 6-A of part 132.
EPA considered making the methodology optional for all chemicals, as advocated
by some commenters,  but decided that the methodology was advanced enough to
use for those chemicals of greatest concern (BCCs) to the higher trophic level
wildlife species feeding from the aquatic food webs in the Great Lakes basin.

      For the development of Tier II wildlife values for all pollutants, use
of the proposed Tier II methodology is encouraged, but not required.  To
derive Tier I numeric wildlife criteria for chemicals not listed in Table 6-A
of part 132, the methodology contained in the final Guidance is also
encouraged, but not required.  While States or Tribes may develop and
implement additional Tier I criteria or Tier II values as deemed necessary,
any derived Tier I criterion remains subject to EPA review and approval at 40
CFR part 131.  In the event that the methodology is used to derive Tier I
criteria for pollutants not listed in Table 6-A of part 132 or to derive Tier
II values, States and Tribes are also encouraged to use the methodology for
deriving bioaccumulation factors (BAFs), described in appendix B to part 132.

C.    Effect Component

      As with the noncancer human health methodology, the wildlife methodology
consists of both an effect and an exposure component.  The effect component  is
determined by the toxicity data and the UFs used to account for uncertainties
in predicting an appropriate test dose (TD)  for wildlife species.

1.    Minimum Data Recruirements

      a.     Proposal:  The effect component of the proposed methodology was
defined by the KOAEL, which is the maximum concentration of the toxicant in
the food of the test species which did not cause adverse effects to those test
organisms.  A NOAEL is derived from published studies from which dose-response
curves can be developed.  EPA proposed that the NOAEL selected must be based
on studies of adequate length (i.e., a minimum of 90 days for mammals, and 28
days for birds) so that chronic effects may be reasonably expected to be
expressed.  The use of a LOAEL,  adjusted by an UF, could be used to derive
criteria where an acceptable NOAEL was not available.  A LOAEL is defined as
the lowest concentration of the toxicant in the diet of the test species which
produces an adverse effect on the test organisms.

      The acceptable observed endpoints in these toxicity studies were those
that are directly or indirectly related to maintaining viable wildlife
populations.  Examples of endpoints which would reasonably be expected to be
related to the reproductive or developmental success of the species included
the number of viable young per female, or hatching or whelping success.

      The proposal also established an order of preference for selecting the
appropriate NOAEL or LOAEL to be used to calculate individual wildlife values
for each taxon.  Field study data were preferred over laboratory data, but the
latter could be used in place of field study data if best professional
judgement deemed it to be of better quality. • Examples of circumstances where
laboratory data may be more appropriate than field study data include cases
where:  dose-response curves or cause-effect relationships cannot be

-------
186    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

established for the field study because of the design of the field  study;
because the effects of other stressors cannot be identified or controlled;
where laboratory data are determined  to be more consistent with  other
published data than are field data; or, where there are quality  control
concerns.

      b.    Comments:  Many commenters argued that the minimum exposure
durations required for toxicity studies were too short to adequately evaluate
the potential reproductive effects of chemicals.  Other commenters  either
supported or challenged the use of field studies in the development of
wildlife criteria.  Commenters in support argued that field studies provide a
reasonable estimate of the impact of  chemicals in a natural setting,
incorporating actual exposures and metabolic impacts of the chemicals.
Commenters opposed to the use of field studies believed that current studies
were not constructed to provide quality data or adequate controls to satisfy
the intent of the final Guidance.  In addition, EPA received comments
recommending that existing field studies be used to validate criteria as they
are derived.

      Responses:  The endpoints of concern, as defined in the proposed
methodology, were those expected to impact adversely the reproductive or
developmental success of a species.  The intent of establishing  the minimum
study duration was to limit the use of short-term acute toxicity information
because it may not fully reflect potential impacts on the endpoints of  concern
(i.e., reproduction and development), and, therefore, could result  in an
under-protective criterion.  EPA agrees that longer-term studies, including
multi-generational studies, are desirable and should be used where  available.
It is also important not to make the  study duration requirements so long that
most available data, which could be used to derive wildlife criteria or
values, was eliminated.

      The study duration of 90 days for mammals is consistent with  the minimum
requirements established in the 1980  Human Health National Guidelines  (45 PR
79347) and in the final Guidance for  developing noncancer human  health
criteria  (Appendix C.II.l to part 132).  In that guidance, "subchronic"
toxicity tests are defined as continuous or repeated exposures for  a period of
90 days, or approximately 10 percent  of a rat's lifespan.  EPA acknowledges
that the test species used for development of wildlife criteria  could have
significantly different life-spans than a rat, and therefore permit some
flexibility in the selection of an appropriate study.  However,  rat data are
allowed to be used in deriving wildlife criteria with the proper use of UFs;
therefore, it is reasonable to use a  minimum 90-day study duration  for
wildlife.

      The minimum study duration for  bird taxa was changed from  28  days to 70
days.  The 70-day period was selected to conform with established EPA test
protocols for reproductive effects on avian species, described in U.S.  EPA
(1986).  If a study evaluates impacts on growth or mortality of  chicks, post-
hatching, 28 days may be an adequate  exposure duration.

      EPA continues to support the use of field studies in the development of
wildlife criteria because such studies can be used to predict the impacts of
chemicals in the environment, by integrating food web, dietary preferences,
and metabolic considerations.  EPA cautions that both field study data  and
laboratory study data must be carefully reviewed and evaluated for  their
usefulness and adequacy in deriving wildlife criteria.  Where appropriate, and
based on best professional judgement, EPA supports the use of laboratory data
over field study data when it has been determined that the laboratory data are
of better quality and are more likely to predict impacts on wildlife species.
EPA also supports the use of any other appropriate data to validate not only
the derived criterion, but any and all UFs and exposure parameters  used in the
derivation of that criterion.

-------
                                Section VI: Wildlife                             187
      The final Wildlife TSD provides further guidance on the appropriate
selection of toxicity data, including proper evaluation of studies.

      c.    Final Guidance:  A Tier I criterion is based on toxicological data
of sufficient duration  (generally 90 days or more for mammals and 70 days or
more for birds) from which a dose-response curve may be developed.  Assuming a
dose-response or a cause-effect relationship can be established, field studies
remain the preferred type of data for deriving a TD, although laboratory data
continues to be allowed if it is determined to be of better quality and more
appropriate, or where field data are lacking.

2.    UFs

      a.    Proposal:  EPA proposed to allow the use of UFs to address
uncertainty in any extrapolation of toxicity data to an appropriate endpoint.
For chemicals lacking an acceptable NOAEL, the LOAEL could be substituted,
with the application of an UF (ranging from 1 through 10) to extrapolate to an
estimated NOAEL.  The minimum test durations for the use of LOAELs were the
same as for NOAELs  (28 days for birds, and 90 days for mammals); these differ
from the noncancer Human Health methodology for Tier I criteria, which
requires a LOAEL to be based on a one year or longer rodent study.  The one-
year requirement provides a level of  conservatism which is not needed for the
protection of wildlife populations, where species typically have shorter life-
spans and reproductive cycles (see 58 FR 20879).

      In cases where a chronic endpoint was not observed, but subchronic
effects were expressed which could reasonably be expected to lead to chronic
effects in a longer study, these data could be used with the application of a
subchronic-to-chronic UF.  The range proposed for this factor was 1 through 10
(see 58 FR 20879).

      Because the species tested may not necessarily reflect the sensitivities
of a representative wildlife species in the Great Lakes System, an additional
UF, the species sensitivity factor, could be applied to extrapolate to protect
species of greater sensitivity than the test species.  In the proposal, this
factor ranged from 0.01 through 1 and was applied as a multiplier, rather than
a divisor, as were the other UFs described above (see 58 FR 20880) .

      Finally, in the proposal (58 FR 20881) EPA requested comment on an
alternate formula.  This formula was functionally the same as the proposed
formula, except that each of the UFS was explicitly included in it.

      b.    Comments:  Several commenters considered the use of UFs
inappropriate because it could result in overly conservative criteria because
the factors are multiplicative in nature, thereby resulting in large
uncertainties being'applied.  Commenters also suggested that there is an
insufficient data set available to evaluate the appropriateness and range of
these factors.  Other commenters complained that there was insufficient
guidance to choose appropriate UFs in the ranges provided,  A number of
commenters expressed a preference for using the alternate equation described
in the proposal.  They believed this equation more clearly described the
application of the UFs in the methodology.

      Responses:  EPA believes that the use of UFs is appropriate.  Wildlife
UFs are needed for several reasons, including the limited amount of available
toxicological data,, test duration,  observed endpoints, and differences in the
sensitivities among wildlife taxa.   All of these factors introduce uncertainty
into the criterion derivation process, making the precise determinations of
the effect of ambient concentrations of specific chemicals on wildlife
populations difficult.

      EPA agrees that any UF must be applied >with careful consideration of the
magnitude of the factor; however, the ranges for the UFs proposed for the
wildlife methodology are reasonable.  In the proposal, the data to support the

-------
188    Water Quality Guidance for the Great Lakes System — Supplementary Information Document


appropriateness and range of specific UFs was, in most cases, from work that
had been completed for human health.

      Subsequent to the proposal, EPA performed work that confirms the  ranges
of the UFs used for wildlife.  The analysis to support the range of
interspecies UFs of 1 to 100 involved examinations of acute and chronic
toxicity data..  The analysis compared median lethal doses for numerous  species
for a variety of chemicals.  The results indicate that approximately 90
percent of the median lethal doses among the species for the same chemical
tested were within a factor of 20.  The chronic toxicity analysis involved a
review of data for four chemicals  (including DDT and mercury).  It was
determined that 90 percent of the chronic endpoints for each species were
within two orders of magnitude of the corresponding NOAELs for the other
species tested.

      Regarding the basis of the LOAEL-to-NOAEL UF, results indicate that 95
percent of the ratios of LOAELs to NOAELs for birds and mammals are less than
10.  Finally, the. recommended range of the subchronic-to-chronic UF is
supported by previous reviews on the toxicity of chemicals to laboratory
mammals, and a new analysis of toxicity to birds.  In two separate reviews
more than 95 percent of the ratios of the NOAELs for subchronic exposures
(approximately three months) to NOAELs for chronic exposures (approximately
two years) were less than 10 (Weil and McCollister, 1963; McNamara, 1976).
More detailed information is provided in the final Wildlife TSD and in  U.S.
EPA  (1995b).

      Although a cumulative UF of 10,000 is possible, the range of the
combined UFs for the four chemicals listed in Table 6-A of part 132 range from
six to 10.  In comparison, the Human Health methodology, section V.C.4.b.ii,
estimates a likely maximum composite UF of 3,000.  Therefore, while it  is
possible to use a large cumulative UF for a chemical, in practice the
magnitude of the UFs in the final criteria are likely to be small.  EPA
considers it unlikely that a State or Tribe will derive cumulative UFs  that
are significantly higher.  To ensure that UFs remain reasonable, EPA
recommends that States or Tribes consider a cap on the maximum composite UF of
1,000 because composite UFs greater than 1,000 may indicate a level of
uncertainty that is unacceptable.

      EPA believes it has provided sufficient guidance on the selection of
UFs.  Even so, it is important to note that the selection of UFs will,  in many
situations, be based on best professional judgement.  EPA anticipates that the
Clearinghouse described in section II of this document will provide a forum
for assisting in selecting appropriate UFs for chemicals.

      EPA agrees that the alternate equation is preferred because  it
eliminates the potential confusion of having to multiply the TD by the  species
sensitivity factor, while at the same time dividing the subchronic-to-chronic
and LOAEL-to-NOAEL factors into the TD.  Instead, all UFs are divided directly
into the TD.  In addition, the term "species sensitivity factor" has been
changed to  "interspecies UF" to reduce potential confusion regarding its use.

      Several changes were made to the alternate equation as proposed.  The
intraspecies UF was removed, which was to be applied when additional
protection of the individual or more sensitive members of a species was deemed
appropriate.  Guidance for modifying the criteria to provide for this added
protection is provided in procedure 1 of appendix F to part 132, and in the
final Wildlife TSD.  In addition, the subscripts for the three remaining UFs
were changed to be consistent with U.S. EPA  (1991) .  Finally, several of the
representative species feed at two or more trophic levels, which was not
readily apparent in the proposed equation.  To more clearly characterize
uptake through food, the food ingestion rate for each trophic component has
been separately calculated, and these are presented in Table 1  (discussed
below).  Contaminant uptake through the food is calculated by summing all the

-------
                                Section VI: Wildlife                              189
products of the trophic level -specific food  ingestion  rates and  the
appropriate RAF.

      During its review of the UFs, EPA considered  including  in  the  equation
 (described in the final Guidance section, below) an allometric scaling factor
to derive doses which would be more toxicologically equivalent when
extrapolating from test animals to the representative  species, based on
differences between the species' body weights and metabolic rates.   This  is
because the current state of the science indicates  that  some  of  the
variability in sensitivity across species can be related to general
physiological and anatomical differences observed across organisms within the
same taxonomic class  (e.g., mammals) .  The rates of such processes,  such  as
basal metabolic rates, cardiac output, renal clearance,  oxygen consumption,
food consumption, and water consumption, tend to vary  across  species according
to allometric scaling factors that can be expressed as a non- linear  function
of body weight.  The relationship of these and other physiological processes
to toxicokinetics has led to the explicit use of allometric scaling  for
estimating more toxicologically equivalent doses in EPA's human  health cancer
methodology and its implicit use in the human health noncancer methodology.
EPA recommends that in the determination of  an interspecies UP,  States or
Tribes apply the equation below to assess the allometric scaling factor for
each representative species and to consider  that assessment as one component
in the determination of an appropriate interspecies UF.   This equation was
endorsed by EPA in 1992 (57 FR 24152) .  In the derivation of  wildlife
criteria, allometric scaling is useful in adjusting for  some  of  the
toxicokinetic differences across species.  However, it may not accurately
reflect the toxicokinetics of all chemicals nor encompass all the
toxicodynamic differences among species.  Therefore, in  determining  an
interspecies UF, allometrically- derived TDs  should  be  considered in
conjunction with chemical class -specific information on  sensitivity,
toxicokinetics, and toxicodynamics across species.  This is consistent with
the guidance provided in the SAB commentary  (U.S. EPA, 1994a) which  stated
that allometric relationships should not be  the sole basis for selecting  an
interspecies UF.


                                           fft  1/4
where :  TDR =  Test Dose scaled for the given representative species in
question  (mg/kg-d) . '
               TD for the test species (mg/kg-d) .

        WtT =  Body Weight of the test species (kg) .

        WtR =  Body Weight of the given representative species  (kg; presented
in Table 1) .

      c.    Final Guidance:  The use of the LQAEL-to-NQAEL, subchronic-to-
chronic, and interspecies UFs described in the proposal  (58 FR  20879)  remain
in the final Guidance, although they are applied as divisors in the equation,
consistent with the preferred alternate equation,  as well as for clarity as to
their application.  Their ranges remain as in the  proposal, 1 through  10,
except for the interspecies UF which ranges from 1 through 100  (the
mathematical inverse of its counterpart, the species sensitivity factor,
contained in the proposal) .

      In addition, the alternate equation, as modified in accordance with the
above discussion, has 'been selected for the final  Guidance:

-------
190    Water Quality Guidance for the Great Lakes System — Supplementary Information Document
                                UFA x UFS x UF,
                                  A     s     L
                                ** + £
Where:

      WV = Wildlife' Value in milligrams of substance per liter  (mg/L) .

      TO = TD in milligrams of substance per kilograms per day  (mg/kg-d)  for
the test species.  This shall be either a NOAEL or a LOAEL.

      UFA = UF for extrapolating toxicity data across species (unitless) .  A
species -specific DF shall be selected and applied to each representative
species, consistent with the equation.

      UFS = UF for extrapolating from subchronic to chronic exposures
(unitless) .

      UFL = UF for LQAEL to NOAEL extrapolations (unitless) .

      Wt = Average weight in kilograms  (kg) for the representative  species.

      W = Average daily volume of water consumed in liters per  day  (L/d)  by
the representative species .

      FTU = Average daily amount of food consumed from trophic level i in
kilograms per day  (kg/d) by the representative species .
            = RAF for wildlife food in trophic level i in liters per kilogram
 (L/kg) , developed using guidelines for wildlife presented in appendix  B  to
part 132, Methodology for Development of Bioaccumulation Factors.   For
consumption of piscivorous birds by other birds  (e.g., herring  gulls by
eagles) , the BAF is derived by multiplying the trophic level 3  BAF  for fish by
a biomagnification factor  (BMP) for biomagnification from fish  to birds.

D.    Exposure Component

      The proposed exposure component of the wildlife methodology consisted of
three general areas :  selection of the representative species ;  exposure
parameters for those representative species; and use of BAFs specific  for
wildlife diet.  The first two areas will be discussed below.  For a discussion
on the BAFs refer to section IV of this document.

1.    Representative Species

      a.    Proposal:  During the development of the proposed methodology, EPA
considered using a hypothetical model wildlife species on which to  base  the
derivation of wildlife criteria or values.  EPA decided, however, to use
actual Great Lakes System species representing various foraging behaviors at
upper trophic levels and, therefore, greater exposure than  other wildlife
species inhabiting the Great Lakes System.  In addition, the use of
representative species allowed a basis for deriving an appropriate
interspecies UF in cases where it is known that  there are more  sensitive
species than the species from which the NOAEL was  derived.  This would have
been difficult to determine if a hypothetical model wildlife  species was used.

      Five species were proposed as the representative species: two mammals,
the mink  (Mustela vison) and river otter  (Lutra  canadensis) ;  and three birds,
the belted kingfisher  (Cervle alcvon) , osprey  (Pandion haliatus) ,  and  bald
eagle  (Haliaeetus leucocephalus) .  These species were proposed based on

-------
                                Section VI: Wildlife                             191
geographic distribution, dietary habits, and the trophic level of their prey.
EPA did not consider routes of exposure other than through the aquatic food
web in selecting representative species because of the focus on
bioaccumulative pollutants.

      b.    Comments:  While some commenters expressed support for the
selection of the five representative species, other commenters argued that
these species are not ecologically representative of the Great Lakes wildlife
species because they may not be widespread throughout all parts of the Great
Lakes System, nor are they resident in all ecological habitats.  Other
commenters recommended replacing or adding species, such as gulls, cormorants,
or the raccoon.

      Responses;  EPA agrees that the proposed species selected may not be
ecologically representative of all the possible wildlife species in the Great
Lakes System, but they were not selected for this reason.  Rather, EPA's
intent was to select disparate species most likely to be exposed to
environmental contaminants from aquatic ecosystems to serve as surrogates of
wildlife species that are highly exposed to toxicants from the aquatic food
chain.  The representative species are not necessarily the most
toxicologically sensitive species, nor do they represent species most likely
to be exposed to bioaccumulative environmental contaminants from terrestrial
ecosystems, or through other routes of exposure.  It was not EPA's intent to
select species to account for every available niche.

      After reviewing comments which suggested the use of different
representative species, EPA decided to replace the osprey with the herring
gull  (Larus arcrentatus) based on a re-evaluation of the exposure parameters of
Great Lakes wildlife species, including species identified by commenters for
inclusion on the list  (e.g., raccoons, herring gulls, common terns, and
double-crested cormorants).  EPA first re-evaluated all five of the proposed
species to determine if they are truly representative of the species that are
the most likely to be exposed to environmental contamination through the
aquatic ecosystem.  EPA showed that, although the osprey is one of the more
highly exposed piscivorous birds, it is not one of the most exposed and that
the gull is potentially more exposed.  In addition, the osprey's foraging
behavior is similar to the bald eagle's and a bird with foraging behaviors
different from the kingfisher and eagle would be preferred.  The herring gull
was selected because its body weight is roughly in between the eagle and the
kingfisher, its food ingestion rate is greater than the osprey, and the
trophic levels at which it feeds are higher  (72 percent at trophic level 3, 18
percent at trophic level 4 and 10 percent terrestrial prey, compared to 100
percent at trophic level 3 for the osprey).

      EPA also considered replacing the kingfisher with the common tern;
however, the estimated exposure parameters for the two species were
essentially the same,  and either species could be used.  Distribution
attributes of the common tern that may make it appear a more representative of
the Great Lakes ecosystem do not affect the derivation of the a wildlife
criterion by the methodology in the proposed or final Guidance.  EPA,
therefore, retained the kingfisher as a representative of a small piscivorous
bird feeding entirely on trophic level 3 fish.

      No changes were made to the mammalian representative species because all
other appropriate species in the Great Lakes System have much lower rates of
consumption of aquatic organisms  (e.g., raccoon) than either the mink or
otter.  (See U.S. EPA, 1995a.)

      c.    Final Guidance:  Elements of the exposure component remain largely
the same as the proposal:  the use of five representative species, as
described in the proposal, with the replacement of the osprey by the herring
gull.  The focus on the food and water uptake routes of exposure remains as in
the proposal.

-------
192    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

2.    Exposure Parameters

      a.    Proposal:  The routes of  exposure to wildlife  considered in the
proposal were uptake through drinking the ambient water and uptake  through
dietary consumption of prey in the aquatic food web. EPA presumed that other
routes of exposure  (e.g., inhalation  or trans-dermal uptake)  were
insignificant for these two taxonomic classes where bioaccumulative chemicals
are concerned 'as in the case for the  final Guidance.   Exposure parameters were
derived for each representative species, consisting of body weight,  and food
and water ingestion rates.  In each case, the "average" individual  was
assumed.  Where the ingestion rates could not be determined from published
literature because of the lack of available data, use  of allometric equations
from Nagy (1987) were proposed.  Additional information on the food and water
ingestion rates selected may be found at 58 PR 21005.

      The degree of accumulation of a contaminant at different steps in the
aquatic food chain was determined through the application  of  BAF calculated
pursuant to the proposed BAF methodology  (58  PR 21022) .

      b.    Comments:  Several commenters criticized the use  of  the allometric
equations to determine food and water ingestion rates  because rates change
throughout the life of the organism;  others criticized the values of the rates
presented in the proposal.  Some commenters stated that the bald eagle diet
consists in part of herring gulls, and that the methodology should  be modified
to reflect that route of exposure, taking into account the proper trophic
level of the gull.

      Responses:  Where available, EPA continues to support the  use of
empirically-derived ingestion rates.  EPA conducted a  review  of  the available
literature to re-evaluate ingestion rates used for the representative species.
In this re-evaluation, EPA used empirical data  (where  available)  or allometric
equations to determine basal metabolic rates  for free-living  animals.  Food
ingestion rates were determined from  the basal metabolic rates so different
caloric contents of wildlife food could be considered. The revised exposure
parameters are presented in Table 1,  below.

-------
 g
•H
JJ
 U
 0)
 a
•0
 0)
•H
 4J
 C
 m
 d)
-H
 u
 01

 &
-H
4J
td
4J
 C
 0)
CO
0)
^
 ft
 d)

 u
 O
14-1

 m
 n
 0)
4J
 0)
 n)
 (X

 a)
 m
 O
1
rH >,
0) 0)

d) PI
i4
U O
•H
.c
g

fr1




0) £ iH
4J U 0)
rt ea >
Oi U 0)
C jjj
0 O U
-H (M -H
• I .-I
•*-' f*H
DQ ^^ QJ

0) O 4J
4J-rH «J
* 4J oi
S CO
Oi
Oi
c
H
3*4
*O CQ "rl
•a; CD
m
0)
-H
o
0)
a
CO





w
0)
-H
•o

o
JJ
c
0)
u
0)




1
Oi









a
•0
l_3




Ol

m
-•H
g







**

O O



m M
H) 0)
4J
O




a\ t*
in i-i
H H
h^ 0)
^J
O


rH
00
O
0




O
00 .
O
C
-H
s
^






0*0 <*>

O O
00 (N


ro ^j*
HH






P- 'J'
(Tt CN
O O


ro *»•
H EH




o
o
vo
0




"!
p-
Otter







*•

o
o
rH

PO
hi






OJ
p-
VO
o
o


m
H




P-
o
0




in
H
0
n
5
10
-H
IM
Oi
C
-H
^


*»*•* *

O O O O
CM 00 M H


m ^<
HJt3
•• EH H
£ M
01 d)
-H £
fc 4J



O P-
(N 00 VO
H O 0
0 O O


m if to

4J
o


m
vo
o
o




H

rH
rH
01
C
•rl
^
^4
0)
X
*>*•*> **.*.

MOO 00 O O
erv oo IN P- ro


n TC CO U
^ J O.-H
£ m id
m -a 3
-H W O1
fc -H BJ
ffl •
0
C
(T\ CO rH
rH CN 00 CN
P- O> tN H
ro o o o
O O O O


r*i 4< CQ M
r3 rl a (I)
H H .C

O


o
vo
H
O




VO

0)
rH
S1
0)

•o
rH
Iti
PQ


 OQ  CO     OQ
-rt-H    'O
IM IH     ^
       tn-H
m ^ TJ J3
       ^
rH rH -H  U
 0)  (IJ ja-H
 >  >     W
 v  a) m  nt
                                                                                                    U U  H  10
                                                                                                    •H -H  O   '
                                                                                                    J   -  -   c
                                                                                                    H  ^4  CO
                                                                                                    4J 4-> -H  H
                                                                                                           Oi
                                                                                                     II  II     to
                                                                                                           II  0)
                                                                                                    ro ^*    jr«

                                                                                                    H H So
                                                                                                    0)
                                                                                                    JJ

-------
194    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      Because ingestion rate data, however, are very limited, EPA considers
the use of the allometric equations presented in the methodology to be
adequate when more specific measured values or more appropriate  allometric
equations are not available.  In addition, EPA also supports the use of the
estimation methods described in U.S. EPA  (1993a) or the information contained
in U.S. EPA  (1988) .  The States or Tribes can at their discretion use any of
the estimation methpds or the allometric equations to derive ingestion rate
data.

      In the sensitivity analysis sections for each pollutant in the proposed
Wildlife Criteria Documents, EPA considered the ingestion of mammals, non-fish
eating birds, and fish-eating birds by the eagle.  To define better the
magnitude of the ingestion of these prey, EPA conducted an analysis described
in the final Wildlife TSD and in U.S. EPA  (1995a), which characterized the
diet of all five representative species, including the eagle and other birds
in the Great Lakes basin.  Based on this analysis, the composition of the
eagle's diet for the Great Lakes basin was modified.  The method also
incorporates the use of chemical-specific biomagnification factors to account
for the accumulation of contaminants in piscivorous birds which  serve as prey
for the eagle.  The final Wildlife TSD and the final Wildlife Criteria
Documents further discuss this method for incorporating the biomagnification
factor into the criteria.

      c.    Final Guidance:  For use where ingestion data are lacking, the
final methodology retains the use of the allometric equations from Nagy  (1987)
or the estimation methods contained in U.S. EPA  (1993a).

      EPA also made.modifications to the body weight, and food and water
ingestion rate exposure parameters for each of the five representative
species.  These modifications were based on the continued review of the
technical literature, and as requested by commenters  (see Table VT-1) .

      EPA also reviewed the dietary composition of each of the five
representative species.  EPA found that aquatic organisms comprise virtually
all of the prey for two of the species:  the river otter and kingfisher;
however, the mink, herring gull, and eagle obtain a significant  portion of
their diets from mammals and birds.  The contribution of these aquatic
organisms to the uptake of the contaminant from the aquatic system by these
three species is reflected through an appropriate weighting of the energy
intake through the aquatic and terrestrial components of the diet, and by
assuming zero uptake of the contaminant through the terrestrial  component of
the food chain.

      To properly account for the exposure through the food chain, the portion
of fish-eating prey in the diet must be quantified, and an appropriate
adjustment factor  (a BMP) for the fish-eating birds  (prey) must  be determined.
The derivation and application of the BAF and BMP, and examples  of how these
calculations can be made are described in the final Wildlife TSD and  in  the
final Wildlife Criteria Documents.

E.    Protection of Individual Members of a Population

      1.    Proposal:  The proposed methodology also made allowance for  the
protection of individuals in cases where  decreasing population size or density
threatened the continuing existence of the  species in the Great  Lakes System.
An intraspecies UF, ranging from  1 through  10 could be applied in the effect
component of the methodology, in  a manner  similar to the LOAEL-to-NOAEL,
interspecies, or  subchronic-to-chronic UFs.

      2.    Comments:  Several commenters were  concerned that the wildlife
criteria did not allow enough flexibility in making site-specific
modifications to existing criteria.  Some  commenters argued that the  criteria
should be waterbody-specific, and other  commenters believed that site-specific
modifications less restrictive than the  System-wide criteria  should be

-------
                                Section VI: Wildlife                             195
allowed.  EPA's response to this issue is addressed in section VIII.A. of this
document.

F.    Wildlife Criteria

      1.    Proposal:  The proposed Guidance presented Tier I wildlife
criteria for four pollutants:  mercury (including methylmercury); PCBs;
2,3,7,8-tetrachlorodibenzo-p-dioxin  (TCDD); and DDT and its metabolites.  A
discussion of the TD, UFs, and other assumptions were fully described in the
Wildlife Criteria Documents.  No Tier II values were derived in  the proposal.

      2.    Discussion:  The four Tier I criteria listed in the  proposal have
been modified based on reconsideration of the applicable UFs, on revised
exposure parameters for the final five representative species, and
recalculation of the BAFs pursuant appendix B to part 132 of the final
Guidance.  Additional changes from the proposal, including changes to the UFs
and the BAF used to account for uptake from piscivorous birds fco the eagle are
discussed below..

      In general, values of one, three, or 10 were used as the UFs, to reflect
relative degrees of conservatism.  Theoretically, the UFA could range to 100,
but that did not occur in the derivation of these four criteria.  A value of
one for any of these UFs indicates that there is reasonable certainty that the
TD portrays the expected toxicological condition (i.e., the result is a true
chronic condition with an adequate duration, the test species is reasonably
expected to be among the most sensitive,  and the endpoint reflects a
reasonable NQAEL); a value of 10 indicates that there is significant
uncertainty that the TD portrays the expected toxicological condition; and a
value of three indicates an intermediate level of certainty.

      Additional justification for the toxicological and exposure data used  in
the derivation of the four wildlife criteria may be found in the final
Wildlife Criteria Documents.

      a.    Mercury.

      Comments:  Several commenters stated that the proposed mercury criterion
of 180 pg/L is significantly below natural background, and therefore, is
erroneous.  These commenters cite natural background levels averaging 1 ng/L,
with upper bounds of near 7 ng/L.  Other commenters stated that  the proposed
mercury criterion is too stringent and that it is orders of magnitude below
the level of detection.  These commenters believed compliance with the
criterion would be impossible.  Some commenters believed that the avian LOAEL
for mercury should be based on the ingestion rates of the test animals, not
the ingestion rates of the controls.

      Responses:  EPA revised the numeric criterion for mercury  to reflect
adjustments made to the exposure parameters for the five final representative
species and modifications to the BAFs  (particularly the trophic  level 3 BAF).
A re-evaluation of the interspecies UF for the avian class was also made.  The
resultant value is 1300 pg/L, approximately the same as the natural background
concentrations cited by commenters.

      EPA conducted a study to determine the natural background  levels of
mercury (U.S. EPA,  1993b).  In general, there is an unfortunate  lack of
reliable data because the global transport of mercury from anthropogenic
sources has created concern over whether pristine areas receive  significant
loads of mercury from atmospheric deposition into freshwater ecosystems.

      Data contained in Noreheim and Forslie (1978), Wren (1983), and Vermeer
et al.  (1973) indicate a BMP range of three to 12 between fish and piscivorous
birds (i.e.,  bird prey species for the eagle).   The value 10 was selected as
being a reasonable BMP.

-------
196    Water Quality Gujdance for the Great Lakes System - Supplementary Information Document

      The dose-response curve from the studies from which the avian TD  was
based (Heinz, 1974, 1975, 1976a, 1976b, and 1979) indicate that  the TD  level
was close to the NOAEL.  In addition, the UFL selected in the proposal was
two.  For these reasons the UFL remains at two.   The UFA  was  changed for all
three representative species from 10 to three because reconsideration of the
ranges of sensitivities of various avian species to wildlife indicates  that
the mallard is relatively sensitive, but not the most sensitive  species. -The
avian LOAEL was modified from 64 /zg/kg-day to 78 /zg/kg-day to accurately
reflect the ingestipn rate of the test animals, rather than the  controls as
had been done in the proposal.

      The mammalian DFs were not changed from the proposal.  EPA notes  that
the UFS  was set  to 10  because of the test length;  in the  study selected
(Wobeser et al.f 1976) there was histopathological effects observed at  the  low
doses and the authors indicated that if the exposure period was  longer  death
may have resulted.

      EPA considered the concern on the level of detection in ambient systems,
and the impact on compliance monitoring by discharging facilities.  The final
Guidance sets forth-an appropriate mechanism to describe reasonable compliance
goals in cases where the criterion is below the level of detection.  Section
303(c)(2)(A) of the Clean Water Act states that water quality standards are to
be adopted which are protective of the uses designated for the waters
affected.  In compliance with this provision, aquatic life criteria for metals
have been promulgated in recent years by several states  in the Great Lakes
System that are more restrictive than the level of detection.  Implementation
of the water quality standards, however, does take into  account  the ability to
detect the pollutant in the waste stream.  Procedure 8 of appendix F to part
132 provides that the water quality-based effluent limit must be derived from
the water quality criterion; compliance with that limit, however, may be based
(at the State's discretion) on the level of quantification, defined in
procedure 8 of part 132.

      EPA agrees that the ingestion rates of the test animals, not the
controls should be used in deriving a LOAEL for the avian value.  The LOAEL
was revised to account for the different ingestion rate.

      b.    DDT.

      Comments:  Several commenters argued against the use of the Anderson  et
al.  (1975)  study from which the avian wildlife value for DDT was derived.
They argued that a dose-response curve could not be generated from the  data
presented in that study.  Another commenter argued that  the different
toxicities of the various DDT congeners should be accounted for  in the  Tier I
criterion.

      Responses:  The Anderson et al.  (1975) study used  to derive a NOAEL
represents an adequate basis from which to derive an effect level for use in
deriving a wildlife water quality criterion for DDT and  its metabolites.  The
study analyzed DDT concentrations in the food items of West Coast pelicans
over several years and  correlated those concentrations to reproductive
effects, including hatching success, egg shell thinning, and fledgling
success.  From these data a dose-response curve can be developed, with  the
appropriate application of a UFL to account for uncertainties in extrapolating
to a NOAEL and for uncertainties in time lags for eliminating the toxicant
from the parents.

      The Anderson et al.  (1975) study did not differentiate the effects  of
DDT and those of its metabolites, and therefore, the avian value is based on
total DDT  (DDT, DDE, and DDD) at a near  steady-state condition  in an  aquatic
system.  It is important to account for  the  differences  in bioaccumulation  of
each metabolite through the food web  and establish appropriate  BAFs  for the
DDT mixture in the Great Lakes.  Based on data from Oliver and  Niimi  (1988),
EPA derived composite DDT and metabolite BAF values for  aquatic trophic levels

-------
                                Section VI: Wildlife                             197
3 and 4, based on weighting the BAF for DDT, DDE, and ODD  (derived for the
dissolved fraction), in accordance with the fraction of each compound in the
Great Lakes fish species  (see section IV).  The values for the two BAFs for
trophic levels 3 and 4, respectively, are 1,687,000 and 9,357,000 L/kg.

      Data contained in Braune and Norstrom (1989) indicate a value of 85 for
a BMP for DDE for trophic level 3 fish to herring gulls is a reasonable, "yet
conservative value.  The BMP used in the calculation for DDT and metabolites
was 63  (see appendix K of the GLWQI TSD for the Procedure.to Determine
Bioaccumulation Factors for a discussion on how the BMP of 63 was derived).
This value was obtained from weighting the proportions of DDT, DDE, and ODD in
the tissues of the organisms, and was used in assessing that portion of the
eagle's diet that is comprised of piscivorous birds.  Additional information
may be found in the final Wildlife Criteria Documents.

      The avian UFL was changed from 10 to three because of the relatively low
level of response noted in the study as the LOAEL  (Anderson et al., 1975).  A
value of one is inappropriate because there is a clear indication of an
adverse impact at the TD.  The UFA was also changed from 10 to one for each of
the representative species because published data indicate that the test
species used, the pelican, is among the most sensitive birds in terms of
available LOAELs, particularly when compared to the three avian representative
species.

      The mammalian test dose value was changed from 0.5 mg/kg/day to 0.8
mg/kg/day, based on a recalculation of the food ingestion rate of the test
animals.  Further the contaminant used in the mammalian study was DDT, not a
mixture of DDT and metabolites.  Therefore, the BAFs used to derive the
mammalian wildlife value were based on DDT only, not the mixture weighted BAF
used for birds.  The mammalian UFs were not changed from the proposal.

      The criterion'is reported for DDT and its metabolites because the
criterion is based on the avian wildlife value, which used a test dose that
included DDT, DDD, and DDE.

      c.    PCBs.  Data contained in Braune and Norstrom  (1989)  indicate a
value of 90 for a BMP between fish and piscivorous birds  (i.e., bird prey
species for the eagle)  is reasonable, yet conservative.

      The UFL for birds was changed from 10 to three because a re-evaluation
of the dose-response relationship suggested that a factor of 10 was overly
conservative.

      The UFS for mammals  was adjusted from 10 to one because the duration for
the study selected (Aulerich and Ringer, 1977) was 300 days which was
considered of sufficient length for manifestation of chronic effects.

      d.    2.3.7.8 TCDD.  Data contained in Braune and Norstrom (1989)
indicate a value of 30 for a BMP between fish and piscivorous birds (i.e.,
bird prey species for the eagle) is reasonable, yet conservative.

      The avian UFS was adjusted from one to 10 because the test  duration of
10 weeks (Nosek et al., 1992) was only a small fraction of the reported half-
life for TCDD elimination in non-egg laying adult pheasants  (U.S. EPA, 1993c).
The UFA was  adjusted from 10 to one for each representative species because
other species showed only slightly greater sensitivities than the test
organism,  particularly when compared to the three avian representative
species.

      No changes were made to the values selected for the UFs for mammals.
EPA notes that the UFA  value for each representative species was  selected
based on a comparison' of single-dose lethality data for the rat and mink.

-------
198    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      3.     Final Guidance:  The final values for each of the four Tier I
criteria, modified as discussed above,  'are presented in Table 2.  States and
Tribes must adopt criteria for these four pollutants that are no less
stringent than EPA;s final criteria.  Additional information may be found in
the final Wildlife Criteria Documents.

G.    Comparison of Wildlife Criteria and Methods to National Program and to
      Great Lakes Water Quality Agreement

      The proposal at 58 FR 20884 contained a discussion of the relationship
of this methodology to both the National Program and to the Great Lakes Water
Quality Agreement (GLWQA).  The final Guidance largely supports that
discussion, including the statement that the four wildlife criteria continue
to be more restrictive than existing aquatic life values.

      The proposal described that the wildlife criterion for DDT and
metabolites at 0.87 pg/L was more stringent than the corresponding GLWQA Annex
1 value of 3.0 pg/L.  The final criterion is now 11 pg/L, which is slightly
less stringent than'the Annex 1 value.   For the reasons described in the
aquatic life section of this document,  EPA continues to believe that the DDT
criterion adequately conforms to the GLWQA.

H.    References

      Anderson, D.W., J.R. Jehl, R.W. Risebrough, L.A. Woods, L.R. Deweese,
and W.G.  Edgecombe.  1975.  Brown pelicans:  improved reproduction off the
southern California coast.  Science 190:  806-808.

      Aulerich, R.J. and R.K. Ringer.  1977.  Current status of PCB toxicity
to mink,  and effect on the reproduction.  Arch. Environ. Contam. Toxicol.  6:
279-292.

      Braune, B.M. and R.J. Norstrom.  1989.  Dynamics of organochlorine
compounds in herring gulls:  III.  Tissue distribution and bioaccumulation in
Lake Ontario gulls.  Environ. Toxicol.  Chem.  8:  957-968.

      Heinz, G.H. ' 1974.  Effects of low dietary levels of methyl mercury on
mallard reproduction.  Bull. Environ. Contam. Toxicol.  11:  386-392.3

      Heinz, G.H.  1975.  Effects of methylmercury on approach and avoidance
behavior of mallard ducklings.  Bull. Environ. Contam. Toxicol.  13:  554-564.

      Heinz, G.H.  1976a.  Methylmercury:   second-year feeding effects on
mallard reproduction and duckling behavior.  J. Wildl. Manage.  40(1):  82-90.

      Heinz, G.H.  1976b.  Methylmercury:   second-year reproductive and
behavioral effects on mallard ducks.  J. Wildl. Manage.  40(4):  710-715.

      Heinz, G.H.  1979.  Methylmercury:  reproductive and behavioral effects
on three generations of mallard ducks.  J.  Wildl. Manage.  43:  394-401.

      McNamara, B.P.  1976.  Concepts in Health Evaluation of Commercial and
Industrial Chemicals.  In;  Mehlman, M.A. et al.,  (eds).  Advances in Modern
Toxicology, Vol.1, Part  1.  New Concepts in Safety Evaluation.  John Wiley and
Sons, New York.

      Nagy, K. A.  1987.   Field Metabolic Rate and Food Requirement Scaling in
Mammals and Birds.  Ecological Monographs.   57(2):   111-128.

      Noreheim, G. and A.  Forslie.   1978.   The degree of methylation and
organic distribution in  some birds  of prey.  Acad. Pharmacol. Toxicol.  43:
196-204.

-------
                                Section VI: Wildlife
                       199
      Nosek, J.A., S.R. Craven, J.R. Sullivan, J.R. Olson, and R.E. Peterson.
1992.  Metabolism and disposition of 2,3,7,8-tetrachlorodibenzo-p-dioxin in
ring-necked pheasant hens, chicks, and eggs.  J. Toxicol. Environ. Health.
35:  153-164.

      Oliver, E.G. and A.J. Niimi.  1988.  Trophodynamic analysis of
polychlorinated biphenyl congeners and other chlorinated hydrocarbons in the
Lake Ontario ecosystem:  Environ. Sci. Technol.  22:  388-397.

      Stephan C.E., D.I. Mount, D.J. Hansen, J.H. Gentile, G.A. Chapman, and
W.A. Brungs.  1985.  Guidelines for deriving numerical National water quality
criteria for the protection of aquatic organisms and their uses.  Office of
Research and Development.

      U.S. Environmental Protection Agency.  1986.  Hazard Evaluation
Division:  Standard Evaluation Procedure -- Avian Reproduction Test.  EPA
540/9-86-139.

      U.S. Environmental Protection Agency.  1988.  Recommendations for, and
documentation of biological values for use in risk assessment.  NTIS-PB88-
179874.
      U.S. Environmental Protection Agency.  1989.
Protect Wildlife Resources.  EPA/600/3-89/067.
Water Quality Criteria to
      U.S. Environmental Protection Agency.  1991.  General Quantitative Risk
Assessment Guidelines for Noncancer Health Effects.  EPA Document No. ECAO-
CIN-538.

      U.S. Environmental Protection Agency.  1992.  An SAB Report:  Evaluation
of the Guidance for the Great Lakes Water Quality Initiative.  EPA-SAB-
EPEC/DWC-93-005.

      U.S. Environmental Protection Agency.  1993a.  Wildlife Exposure Factors
Handbook, Volumes I^and II.  EPA/600/R-93/187a and b.

      U.S. Environmental Protection Agency.  1993b.  Assessment of Mercury
Occurrence in Pristine Freshwater Ecosystems.  Draft.

      U.S. Environmental Protection Agency.  1993c.  Interim Report on Data
and Methods for Assessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin Risks to
Aquatic Life and Associated Wildlife.  Office of Research and Development.
EPA/600/R-93/055.

      U.S. Environmental Protection Agency.  1994a.  Advisory on the
Development of a National Wildlife Criteria Program.  EPA-SAB-EPEC-ADV-94-001.

      U.S. Environmental Protection Agency.  1994b.  Draft Proceedings of the
National Wildlife Criteria Methodologies Meeting.

      U.S. Environmental Protection Agency.  1995a.  Trophic Level and
Exposure Analyses for Selected Piscivorous Birds and Mammals.  Volumes 1 and
3.  Draft.

      U.S. Environmental Protection Agency.  1995b.  Technical Basis for
Recommended Ranges of Uncertainty Factors used in Deriving Widlife Criteria
for the Great Lakes"Water Quality Initiative.  Draft.

      U.S. General Accounting Office.  1987.  Wildlife Management:  National
Refuge Contamination is Difficult to Confirm and Clean Up.  Gaithersburg, MD.
GAO/RCED-87-128.

      Vermeer, K.,  F.A.J. Armstrong, and D.R.M. Hatch.  1973.  Mercury in
aquatic birds at Clay Lake, Western Ontario.  J. Wildl. Manage.  37:  58-61.

-------
200    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      Weil, C.S. and D.D.  McCollister.   1963.  Relationship Between Short- and
Long-term Feeding  Studies  in Designing'an Effective Toxicity Test.
Agricultural and Food Chemistry.   11(6):  486-491.

      Wobeser, G.f N.D.  Nielsen,  and B. Schiefer.  1976.  Mercury and mink:
II.  Experimental  methyl mercury intoxication.  Can. J. Comp. Med.   40:   34-
45.

      Wren, C.D.,  H.R. MacCrimmon,  and B.R. Loescher.  1983.  Examination of
bioaccumulation and  biomagnification of metals in a precambrian  shield lake.
Water, Air, and Soil Pollut.   19:  277-291.  1983.

-------
                                 Section VI: Wildlife
201
Table  2.   Final Tier  I  Wildlife Criteria.
Pollutant
mercury (including methylmercury)
PCBs
2,3,7,8-TCDD
DDT and metabolites
Criteria (/ig/L)
1.3 x Id'3
7.4 x 1CT5
3.1 x 1CT9
1.1 x ICr5

-------
202     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

-------
                             Section VII: Antidegradation                         203
                  VH.  ANTIDEGRADATION


A.    General Discussion/Background

      This document explains the intent of the final Guidance on
antidegradation and.EPA's expectations of the Tribes and States in the
development and implementation of antidegradation policies and procedures.
This document also includes guidance to Tribes and States in the
implementation of the agreement entitled "A Bi-National Program to Restore  and
Protect the Lake Superior Basin." (September, 1991)   Tribes and States may
extend additional protection to the waters of the Lake Superior Basin through
the use of the special designations contained in the final Guidance if they so
choose.

1.    History of the Great Lakes Antidegradation Guidance

      The history of the Federal policy on antidegradation is discussed in
detail in the preamble of the proposed Guidance; readers are referred there
for a complete discussion of the origins and evolution of the Federal
antidegradation policy.  EPA's Water Quality Standards Handbook (September,
1993) also contains useful background information.

      Current national policy is found in the Federal regulations  at 40 CFR
131.6 and 40 CFR 131.12.   Federal regulations at 40  CFR 131.6 specify that
Tribal or State water quality standards must include an antidegradation
policy.  Federal Regulations at 40 CFR 131.12 identify the elements of an
acceptable antidegradation policy.  The Federal antidegradation policy is
composed of three levels of protection commonly referred to as tiers.  The
first element identified at 40 CFR 131.12(a)(1)  protects the minimum level  of
water quality necessary to support existing uses and applies to all waters.
This element establishes the ultimate limit on the extent to which water
quality can be lowered in a water body.  Lowering of water quality to the
point where existing uses are impaired is prohibited.  The second  element is
found at 40 CFR 131.12(a)(2), and protects water quality where water quality
is better than that needed to support fish and aquatic life and recreation  in
and on the water.  Where these conditions exist, the water body is considered
high quality and water quality must be maintained and protected unless
lowering water quality is necessary to support important social and economic
development.  The third element at 40 CFR 131.12(a)(3)  involves the protection
of water quality in water bodies that are of exceptional ecological,  aesthetic
or recreational significance.  Water quality in such water bodies,  identified
as Outstanding National Resource Waters (ONRW),  must be maintained and
protected.

      The protection of high quality waters under antidegradation  causes
considerable confusion and controversy.  It is often interpreted incorrectly
as an absolute prohibition on lowering of water quality in high quality
waters.  Such a prohibition would amount to a "no growth" policy which is not
consistent with EPA"1 s position as expressed in the regulations. Neither the
existing Federal regulations nor the final Guidance  prohibit activities that
would lower water quality in high quality waters.  The antidegradation
provisions contained in the final Guidance provide a structure for the
systematic evaluation of activities that are expected to lower water quality.
Implementation of antidegradation allows Tribes and  States to arrive at a
decision that considers all the available information regarding the social,
economic and environmental impacts of lowering water quality.   Review of such
activities under a Tribe's or State's antidegradation policy is intended to
ensure that any lowering of water quality is necessary,  that the lowering of
water quality is minimized and that desirable economic and social  benefits

-------
204    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

accrue to the area affected by the lowered water quality as a result of the
lowering of water quality.

      The applicability of antidegradation to nonpoint sources is also a
source of confusion.  EPA policy is that water quality standards, including
antidegradation, are applicable to any activity that might affect water
quality (see "Interpretation of Federal Antidegradation Regulatory
Requirement," memorandum from Tudor Davies, Director, Office of Science and
Technology, to Water Management Division Directors, dated February 22, 1994) .
However, despite the broad applicability of water quality standards,
mechanisms to implement water quality standards may not exist in all
circumstances.  Neither the antidegradation provisions contained in the final
Guidance,  nor existing regulations, confer any additional authority upon
States, Tribes or EPA to regulate nonpoint sources of pollution.  However,
where independent regulatory authority over nonpoint sources exists requiring
compliance with water quality standards, compliance with the antidegradation
provisions of the final Guidance is required.

      From the inception of the joint State-EPA effort to develop consistent
water quality standards for the Great Lakes, the participants recognized
antidegradation guidance was an essential element of such an effort.  The
Great Lakes Critical Programs Act of 1990 also recognized this by requiring
that the final Guidance published by EPA include antidegradation provisions at
least as stringent as existing Federal requirements.

2.    Summary of the Proposed Guidance

      The proposed Guidance consisted of four components each of which were to
be adopted by Tribes and States:

      a.    An Antidegradation Standard;
      b.    Antidegradation Implementation Procedures;
      c.    An Antidegradation Demonstration; and
      d.    An Antidegradation Decision.

Each of the components is discussed in detail below.

a.    The Antidegradation Standard

      As proposed, the Great Lakes antidegradation standard was derived from
the existing Federal antidegradation policy at 40 CFR 131.12.  The proposed
Guidance provided additional detail on antidegradation to assist Tribes and
States in implementing the standard and to  encourage consistency across the
basin.

      The protection of existing uses contained in the proposed Guidance
refers to the definition found at 40 CFR 131.3.  The intent of this reference
was to make clear that water quality necessary to support existing uses,  as
well as designated uses, is protected under antidegradation.  The proposed
Guidance prohibited lowering water quality  if, as a result, either existing or
designated uses wouj.d be impaired.

      The proposed Guidance and existing regulations at 40 CFR 131.12 provide
similar protection for high quality waters.  The proposed Guidance expanded
upon Federal regulations, however, in that  it specified that high quality
waters should be identified on a parameter-by-parameter basis, whereas
existing Federal regulations do not specify a particular method  for
identifying high quality waters.  The parameter-by-parameter approach was
selected for the proposed Guidance because  it was  considered the most
reasonable and workable way of identifying  high quality waters and because it
ensured that water quality improvements would be protected.  Also,  a
parameter-by-parameter  approach allows a water body to be considered high
quality for one pollutant even if  the criteria for another is exceeded,  as is
the case in many water  bodies within the Great Lakes basin.  Finally, this

-------
                             Section VII: Antidegradation                          205
approach is consistent with how waters are assessed and regulated in other
aspects of States' and Tribes' water quality programs.

      The final major element of the standard is the protection of Outstanding
National Resource Waters  (ONRWs).  The standard contained in the proposed
Guidance is identical to existing Federal regulations.

b.    Antidegradation -Implementation Procedures

      The implementation procedures contained in the proposed Guidance were
intended to provide direction to States and Tribes on how the antidegradation
standard should be applied.  The implementation procedures included
definitions of key terms and descriptions of how the Tribes and States were to
carry out the requirements of the Great Lakes antidegradation standard.  The
proposed Guidance stated explicitly that designated uses must reflect existing
uses and that where designated uses are impaired, water quality may not be
lowered at all with respect to the pollutant or pollutants causing the
impairment.

      For high quality waters, the proposed Guidance defined the concepts of
de minimis lowering of water quality and significant lowering of water
quality.  The implementation procedures recognized that considerable variation
was possible in the effects of different activities on water quality.  The
proposed Guidance included a mechanism for distinguishing between activities
based on the extent to which water quality was anticipated to be lowered.
Small reductions in water quality were identified as de minimis and not
subject to antidegradation review, whereas larger reductions were identified
as significant and subject to antidegradation review.  Both concepts were
intended as mechanisms to allow Tribes and States to differentiate between
activities that are likely to have an inconsequential effect on water quality
and those that are likely to have significant effects and to focus their
efforts on those that are of the most consequence to water quality.  The
definition of de minimis in the proposed Guidance provided criteria to be used
by Tribes and States to identify when a lowering of water quality could be
considered de minimis.  The proposed Guidance also provided criteria for
identifying when an activity would result in a significant lowering of water
quality subject to' antidegradation review.

      The proposed Guidance recognized that the significance of a potential
lowering of water quality depended not just on the magnitude of the lowering
of water quality, but also on the types of pollutants involved.  In keeping
with the emphasis of the Guidance on BCCs, the proposed Guidance established a
more restrictive threshold of significance for BCCs than for non-BCCs.  The
concept of existing effluent quality (EEQ) was specified in the proposed
Guidance as the means to implement antidegradation for BCCs.  EEQ served as
the basis for either permit limits or notification requirements which, if
exceeded, triggered antidegradation review.  Each of these concepts is
discussed in greater detail below.

      It should also be noted that for high quality waters and ONRWs, the
implementation procedures in the proposed Guidance allowed short-term and
temporary lowering of water quality without review under the antidegradation
provisions of the Guidance.  This is consistent with existing national policy
as described in USEPA's Water Quality Standards Handbook (August, 1994) and
allows Tribes and States greater flexibility in how they use the
antidegradation procedures as regulatory tools.

c.    Antidegradation Demonstration

      When an action is considered that could lower water quality in a high
quality water,  Federal regulations require that a determination be made that
the action is necessary to support important Social and economic development
{40 CFR 131.12(a) (2)}.  The proposed Guidance explained the process for
evaluating such activities and the information a discharger considering such

-------
206    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

an action must provide to the regulatory agency making the determination.  The
antidegradation demonstration process consisted of two sub-demonstrations:

1.    a demonstration that the significant lowering of water quality is
      necessary and,

2.    a demonstration that the significant lowering of water quality will
      support -important social and economic development.

Under the proposal, the two demonstrations were intended to be performed
sequentially, with performance of the social and economic development
demonstration contingent on a successful demonstration that the significant
lowering of water quality was necessary.  To assess the need for a significant
lowering of water quality, a person proposing an action that would lower water
quality would first determine whether or not existing treatment, pollution
prevention, additional treatment or some combination within a defined cost
range could avoid the need to lower water quality.  If this was the case, the
significant lowering of water quality would not be permitted.  If not, the
discharger would be required to demonstrate that important social benefits,
economic benefits or both would accrue to the community as a result of the
activity responsible for the significant lowering of water quality.

      The proposal also included special requirements applicable to RCRA and
Superfund actions.  In addition, requirements for implementing optional
special designations for Lake Superior were included.

d.    Antidecrradation Decision

      The proposed Guidance described how Tribes and States should evaluate
the data provided by a discharger in making a decision concerning whether or
not to allow a significant lowering of water quality.  The specifics of this
section are discussed in greater detail below.

B.    Overview of the Final Guidance

      The final Guidance includes Great Lakes-specific antidegradation
requirements for BCCs only.  A lack of specific requirements in the final
Guidance for non-BCCs should not be construed as relieving States and Tribes
of their responsibilities under the CWA and Federal regulations to adopt water
quality standards consistent with the CWA and Federal regulations.      As
stated at 40 CFR 131.6, antidegradation provisions consistent with 40 CFR
131.12 are compulsory elements of any State's or Tribe's water quality
standards; States and Tribes must adopt antidegradation policies and
implementation procedures just as they must adopt designated uses and
criteria.  For non-BCCs, the antidegradation policy and implementation
procedures developed by States and Tribes must define key terms,, specify what
types of information will be required to demonstrate that a significant
lowering of water quality is both necessary and will support important social
and economic development, how that information will be used to arrive at a
final decision regarding a request to lower water quality, and how public
participation will be factored into the final decision consistent with the
antidegradation standard and 40 CFR 25.  This SID includes EPA's
recommendations as to how antidegradation should be implemented for non-BCCs.

      For BCCs, States and Tribes must adopt specific provisions consistent
with the antidegradation standard and implementation elements contained  in
appendix E.  Special requirements for BCCs are necessary because of the
demonstrated sensitivity of the Great Lakes  System to such pollutants.
Consequently, imposing these requirements on Great Lakes States and Tribes is
warranted to protect the shared resources of the Great Lakes System.

      A second major change to the final Guidance is that the final Guidance
does not require the use of existing effluent quality  (EEQ)-based limits to
implement antidegradation for BCCs.  EEQ-based limits are not included in  the

-------
                             Section VII: Antidegradation                          207
final Guidance because EPA believes that it is possible that the use of EEQ-
based limits to implement antidegradation could have unintended effects
contrary to the goals of the CWA and the intent of antidegradation.

      The final Guidance retains the overall structure of the proposed
Guidance, that is, it contains an antidegradation standard, implementation
procedures, demonstration requirements, and a final decision.  Very few
changes were made tb the antidegradation standard between the proposal and the
final Guidance.  An implementation framework, consisting of implementation
procedures, demonstration requirements and the requirement for a final
decision is provided in the final Guidance.  However, many of the details
contained in the proposed Guidance, such as specific benchmarks for assessing
affordability of alternative treatment technologies, are included elsewhere in
this document as guidelines only.

C.    Detailed Discussion of the Final Regulation

1.    The Antidegradation Standard

      The antidegradation standard contained in the final Guidance differs
little from either the proposed Guidance or existing regulations on
antidegradation at 40 CFR 131.12.  The section pertaining to the protection of
high quality waters was reworded to cross-reference the definition of high
quality waters at 40 CFR 131.3 (See discussion below).  States and Tribes are
required to adopt an antidegradation standard consistent with the Standard
contained in appendix E applicable to increased loadings of BCCs to the Great
Lakes System.  States and Tribes may adopt the antidegradation standard as
applicable to all waters and pollutants.  As required by 40 CFR 131.6, water
quality standards adopted by a State or Tribe must include an antidegradation
policy consistent with 40 CFR 131.12.

2.    Antidsgradation Implementation Procedures

      The antidegradation implementation procedures were the second major
component of the proposed Guidance.  The implementation procedures consisted
of definitions of key terms and the processes to be used by Tribes and States
in applying antidegradation.  The proposed Guidance provided procedures for
protecting existing uses, high quality waters and ONRWs.  In addition, the
implementation procedures included methods for applying optional special
protection designations to waters in the Lake Superior Basin.

a.    De Minimis Lowering of Water Quality

i.    Background

      The proposed Guidance made provision for identifying certain small
increases in loading as de minimis and not subject to the requirements for
antidegradation review.  The "de minimis test" identified three criteria to be
used in classifying an increased loading as de minimis.  These were:

1.    only non-BCCs will be released as a result of the proposed activity
      responsible for the anticipated lowering of water quality;
2.    the proposed lowering of water quality uses less than 10 percent of the
      available assimilative capacity; and
3.    for pollutants included in 40 CFR 132.2, Table 5, at least ten percent
      of the total assimilative capacity remains unused following the lowering
      of water quality.

The proposal also required that any decision to allow a de minimis lowering of
water quality comply with the requirements in the TMDL section of the
implementation procedures to maintain a margin of safety (MOS).  Finally, the
proposal specified how and when the total and unused assimilative capacity of
a water body should be calculated.

-------
208    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      EPA's intent in including the de minimis test in the proposed Guidance
was to recognize that certain activities, although they may result in some
lowering of water quality, will not lower water quality to such an extent as
to result in a significant lowering of water quality.  EPA's goal in allowing
States and Tribes to identify certain increases as de minimis was to provide a
means of reducing the administrative burden on all parties associated with
activities of little or no consequence to the environment.  De minimis
determinations were made at the discretion of the State or Tribal Director;
the Director could choose to treat as significant any lowering of water
quality that would otherwise be de minimis where the circumstances of a
particular discharge warrant more thorough review.

ii.   Discussion of Significant Comments

      EPA received a number of comments on the proposed de minimis provisions.
Significant comments are discussed below.  For responses to specific comments,
see the comment response document in the docket of this rulemaking.

      Comment:  Some commenters stated that the de minimis provision is too
broad and should be"narrowed to include fewer activities.

      Response:  De minimis provisions are not part of existing regulatory
requirements, nor are they authorized for BCCs under the final Guidance.  For
non-BCCs, States and Tribes may include de minimis provisions in their
antidegradation policy at their discretion.  Although de minimis provisions do
involve non-conservative assumptions, the de minimis provisions included in
the proposed Guidance are not likely to seriously undermine the protection
afforded a high quality water body through antidegradation.De minimis
provisions provide a means for States and Tribes to differentiate between
actions that will result in an increased loading of a pollutant to a receiving
water that is likely to have a significant impact on water quality and those
that are unlikely to do so and focus review efforts on actions that will
degrade water quality.  It is reasonable to assume that loading increases of
non-BCCs that will use less than  ten percent of the remaining assimilative
capacity in a water body will have a negligible effect on ambient water
quality.

      Comment:  Some commenters felt that the de minimis provisions are too
narrow and should be broadened to reduce the number of activities subject to
antidegradation review.

      Response;  Given that existing Federal regulations are silent on the
concept of de minimis and that allowing any de minimis provision involves non-
conservative assumptions about the effects of increased loadings on water
quality, an expanded de minimis provision would be incompatible with the
antidegradation standard that States and Tribes are required to adopt under 40
CFR 131.6.  De minimis provisions must be narrow in order for antidegradation
to function as it was intended.   Some activities or actions have a negligible
impact on water quality and may be exempted appropriately from antidegradation
review through a de minimis provision; however, allowing loading increases
that use more than ten percent of the remaining assimilative capacity of a
water body to be considered de minimis strains the credibility of a State's or
Tribe's antidegradation implementation procedures and increases the chance
that a significant lowering of water quality could occur without
antidegradation review.

      Comment:  The de minimis provisions should be extended to include BCCs.

      Response;  EPA does not agree that even small increases in the loadings
of BCCs to the Great Lakes Basin  can be considered de minimis.  Low levels  of
BCCs in the Great Lakes have adverse impacts on the organisms that inhabit
them.  Further, because BCCs are  both resistant to degradation  and
hydrophobic, they tend  to accumulate in sediments and biota, amplifying  their

-------
                             Section VH: Antidegradation                          209
effects.  For these reasons, even small increases in loadings of this type of
pollutant must be considered significant.

      Comment:  There is no need to reserve a portion of the unused
assimilative capacity as a MOS for either the pollutants included in the final
Guidance or those listed on Table 5 and excluded from the Guidance.

      Response:  EPA does not agree with this comment.  A MOS is essential to
preserve high quality waters in their high quality state.  It is inconsistent
with the intent of antidegradation to allow such waters to be degraded without
scrutiny to the point where criteria are only just achieved.  Consequently, a
de minimis provision must include a MOS to ensure that such a situation does
not occur.  When a State or Tribe chooses to define de minimis as a percentage
of the remaining assimilative capacity, there will always be an opportunity
for a de minimis increase, albeit a smaller and smaller one.  Absent a MOS
requirement, this could lead to a situation where the water quality in a water
body is reduced to the point where the water body is effectively no longer
high quality, without antidegradation review.  To prevent this, a MOS similar
to the requirements contained in the final Guidance should be a part of a
State's or Tribe's antidegradation implementation procedures if the procedures
include a de minimis provision.

      Comment:  Many commenters felt that EPA should either do more to account
for the cumulative effects of multiple de minimis lowerings of water quality
or, conversely, that special protection against such cumulative effects is not
necessary.

      Response:  EPA agrees with those commenters concerned over the
cumulative impacts of a State's or Tribe's de minimis provisions.  MOS
requirements ensure that in no case will the water quality of a high quality
water body be degraded to the point where the water body is no longer high
quality without review and should be a part of a State's or Tribe's
antidegradation implementation procedures.

      MOS requirements, however, address only the more extreme cumulative
effects of de minimis increases.  Several commenters expressed concern that a
single discharger could request numerous de minimis increases resulting in a
significant impact on water quality that would not be subjected to
antidegradation review.  Similarly, several dischargers to the same water body
could receive permission for de minimis increases leading to a significant
lowering of water quality.  Such concerns are inherent in any decision to
allow for de minimis provisions.  However, the benefits of the de minimis
provisions in allowing States and Tribes to reduce administrative burdens and
prioritize their efforts outweigh the possibility for abuse, especially given
the MOS requirements and Tribal, State and EPA oversight of water quality
program implementation.  Abuse of the de minimis provisions can also be
prevented if States and Tribes adopt a definition of significant lowering of
water quality that allows the Director to identify any lowering of water
quality as significant on a case-by-case basis.

iii.  The Final Guidance

      Because the final Guidance does not address non-BCCs, it also does not
make provision for de minimis lowering of water quality.  Where a State or
Tribe wishes to include a de minimis provision in the State's or Tribe's
antidegradation policy and procedures for non-BCCs, EPA recommends adopting an
approach based on the proposed Guidance.  Any de minimis provision should be
based on a percentage of the unused assimilative capacity to protect against
over-allocation of the water body. In addition, the de minimis provision
should include a MOS based upon a percentage of the total assimilative
capacity such that once more than a certain percentage of the total
assimilative capacity is used,  any further lowering of water quality is
subject to antidegradation review.  If a TMDL is in place, total assimilative
capacity and loading capacity are functionally equivalent; it is not necessary

-------
210    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

to recalculate the total assimilative capacity of a water body where loading
capacity has already been calculated.  'Unused assimilative capacity must
account for background loading and should be determined based on permitted
rather than actual discharge levels at point and nonpoint sources.  As written
in the proposal, unused assimilative capacity could have been overestimated if
background loadings were not considered or if dischargers to the water body
were discharging below permitted levels.   Finally,the unused assimilative
capacity must be recalculated prior to each request for a lowering of water
quality to ensure that it reflects current conditions in the water body.

b.     High Quality Waters

i.     Background

      As explained above, existing Federal regulations at 40 CFR 131.12
establish a framework for antidegradation consisting of three levels of
protection.  The second level creates the category known as high quality
waters and defines high quality waters as those water bodies with water
quality better than minimum levels necessary to support fish and other aquatic
life and recreational uses.  Federal regulations provide that where water
quality is better than that needed to support propagation of fish, shellfish,
and wildlife and recreation in and on the waters, that water quality shall be
protected unless lowering of water quality is necessary to support important
social and economic development.

      The proposed Guidance followed closely the provisions contained in the
Federal regulations.  In addition, the proposed Guidance specified that the
level of protection afforded a water body under antidegradation be determined
on a parameter-by- parameter basis, considering each individual pollutant
separately from the others present in a water body.  Under the proposed
Guidance, a discharger contemplating an action that would result in an
increased loading wpuld identify the constituents of its effluent that would
increase as a result of the action.  Then, the ambient level of the pollutants
of interest would be determined and compared to the applicable criteria.
Where ambient concentrations of the pollutants in question are less than
criteria concentrations, the water body would be considered high quality for
those pollutants and increases in those pollutants would be subject to the
requirements applicable to high quality waters.

      EPA selected this method of identifying high quality waters for the
proposed Guidance for several reasons.  First, it establishes a clear, easily
evaluated criterion for determining whether or not a water body is high
quality.  The proposed method is also consistent with existing EPA guidance on
the subject of identifying high quality waters for purposes of antidegradation
(see "Application of Antidegradation Policy to the Niagara River," memorandum
from Martha G. Prothro, Director, Office of Water Regulations and Standards,
to Richard L. Caspe, Director, Water Management Division, Region II, dated
August 4, 1989, available in the docket for this rulemaking) and the current
approach taken by many States in the Great Lakes basin to implementing
antidegradation.  Also, the parameter-by parameter approach ensures internal
consistency between the mechanisms used in the final Guidance to identify
impaired waters and those used to identify high quality waters.  Finally,
identifying high quality waters on a parameter-by-parameter basis protects
high quality water wherever  it exists in the basin, even when individual
criteria may not be met and ensures that the water quality of the Great Lakes
System will not degrade below that necessary to support fish and other aquatic
life and recreation in and on the waters.

ii.   Discussion of Significant Comments

      Comment;  A number of  commenters stated that factors other  than just the
ambient pollutant concentration should be considered in determining whether  or
not a water body is high quality.

-------
                             Section VII: Antidegradation                          211
      Response:  EPA does not agree with this comment as a general matter.
The language of the,final Guidance conforms with existing Federal regulations
that refer to ambient water quality as the measure of whether or not a water
body is high quality.  In addition, ambient water quality provides a
quantitative measure for identifying high quality waters with regard to a
specific pollutant that is compatible with implementation through a permit to
discharge or other control document.  Finally, in general, the water bodies
found in the Great Lakes basin are, or should be, able to support fish and
aquatic life and recreational uses.  Therefore, where water quality is better
than necessary to support fish and aquatic life and recreation, that water
quality should be protected.

      Comment:  Some commenters stated that the parameter-by-parameter
approach to identifying high quality waters should be required for BCCs only.

      Response:  The commenter does not provide any compelling reasons for
employing a different approach to identify high quality waters for non-BCCs
than is used for BCCs.  On the contrary, the rationale for using a parameter-
by-parameter approach is equally valid for both BCCs and non-BCCs.  While the
final Guidance requiring the pollutant-by-pollutant approach pertains only to
BCCs, EPA believes that consistent with current EPA policy, States and Tribes
should implement antidegradation on a parameter-by-parameter basis for non-
BCCs, and identify high quality waters on this basis.

      Comment:  The identification of high quality waters for purposes of
antidegradation should consider the attainability of fishable/swimmable uses
as well as ambient water quality.

      Response;  Given that final Guidance applies only to increased
discharges of BCCs, and given the properties of BCCs, it is not reasonable to
assume that there exists anywhere in the Great Lakes System a water body where
an increased loading of BCCs does not have the potential to ultimately affect
water quality in the a water body where fish and aquatic life uses and
recreation in and on the water occur.  Consequently, there is no basis for
distinguishing between different water bodies within the Great Lakes System
for purposes of antidegradation.  In order the protect the Great Lakes System
as a whole, any water body where water quality is better than the minimum
necessary to support fish and aquatic life and recreation in and on the water
is considered high quality and subject to full protection under
antidegradation.

      Attainability of a use should not be confused with impairment of a use.
In some instances, a designated use may be impaired due to pollution resulting
from historic activities, such as fish consumption advisories due to PCB
contamination,  and yet the use may be considered attainable with control of
the contamination.  Identifying high quality waters on a parameter-by-
parameter basis provides Tribes and States the flexibility to both encourage
restoration by prohibiting increased loadings of the pollutant causing the
impairment, and at the same time provide antidegradation protection for
otherwise high quality waters.  As a result, water quality problems will be
kept from worsening through the protection of existing and designated uses
while new water quality problems will be prevented through the protection of
high quality waters.

      Comment:  The. definition of high quality waters should recognize that
some waters that meet the requirements on a parameter-by-parameter basis may
not be high quality resources.  The final Guidance should allow for less
rigorous review for activities affecting such waters.

      Response:  As stated above, any release of BCCs to the Great Lakes
System has the potential to affect water quality throughout the system,
because of their persistence and tendency to'bioaccumulate.  In order to
maintain the Great Lakes System as a whole in a high quality condition, all
water bodies within the system must be treated as high quality.  However, EPA

-------
212    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

agrees with the concerns raised by this comment with respect to non-BCCs.  EPA
may approve implementation procedures for non-BCCs which include an exemption
from the requirements for identifying high quality waters on a parameter-by-
parameter basis for non-BCCs for water bodies that are of limited ecological,
recreational and aesthetic significance.

iii.  The Final Guidance

      Much of what was included in the proposed Guidance is incorporated in
the final Guidance.  For BCCs, the final Guidance requires States and Tribes
to identify high quality waters on a parameter-by-parameter basis.  For
pollutants other than BCCs, EPA recommends that, in general, high quality
waters be identified on a parameter-by-parameter basis in the same manner as
for BCCs.

      The final Guidance includes one additional change to improve its
consistency and clarity.  A detailed definition of high quality water is
included in the definition section.  In the proposed Guidance, the definition
was found only in the antidegradation standard.

c.    Lake Superior Basin - Outstanding International Resource Waters

      The final Guidance includes a definition of the term "Lake Superior
Basin - Outstanding International Resource Waters."  This term, which was not
defined in the proposed Guidance, is defined in the document, "A Bi-National
Program to Restore and Protect the Lake Superior Basin,"  (September, 1991)
available in the docket to the rule.  A definition is included in the final
Guidance in response to comments requesting the inclusion of the definition.

d.    Outstanding National Resource Water

i.    Background

      The proposed Guidance included a definition of Outstanding National
Resource Water (ONRW).  The definition was derived from existing Federal
regulations at 40 CFR 131.12(a)(3).  The definition specified that designation
of ONRWs is at the discretion of States and Tribes.  The definition also
included examples of the types of waters that might be afforded ONRW status by
States and Tribes.

ii.   Discussion of Significant Comments

      Comment:  Some commenters stated that the proposed Guidance broadened
the definition of ONRW beyond existing Federal regulations.  The commenters
stated that the examples of types of waters that could be considered for ONRW
designation could encompass any water body within the Great Lakes basin and
that this could lead to delays in the permitting process as disagreements
arose over whether or not a water body was or should be an ONRW.

      Response:  EPA does not agree with this comment.  The final Guidance
does not contemplate any additional requirements beyond existing Federal
regulations for ONRWs.  The types of waters subject to listing were included
in the proposal as examples of what States and Tribes should consider  in
designating ONRWs.  Identification of an ONRW is inherently site- and  case-
specific and depends not only on  the water body but also on its  importance as
a resource to the State or Tribe.  Any cold water stream may be  an ONRW where
such resources are rare; elsewhere, where they may be more common, the
presence of a cold water community may not  in itself be sufficient to  warrant
designation as an ONRW.  Designation  of ONRWs occurs through the normal water
quality standards review process  and  should not delay permit issuance.

-------
                             Section VII: Antidegradation                          213
iii.  The Final Guidance

      The content of the final Guidance was not changed from the proposal.
The definition included in the final Guidance represents accurately what  is
intended by an ONRW designation.

e.    Significant Lowering of Water Quality

i.    Background

      The proposed Guidance identified criteria for determining when a
lowering of water quality is significant.  For all regulated discharges of
BCCs, any increase in loading to a high quality water, measured as a change  in
EEQ, was considered significant and subject to antidegradation review.  For
point source discharges of non-BCCs to high quality waters, a significant
lowering of water quality was defined generally as an increase in applicable
permit limits greater than a de minimis increase.  For nonpoint sources of
non-BCCs, a significant lowering of water quality was defined generally as an
increase in the rate of mass loading authorized by the governing nonpoint
source program.  Increased mass loadings of non-BCCs that would not change the
concentration of the pollutant outside of a designated mixing zone were not
required to undergo antidegradation review.  Finally, there was a "safety
provision" that allowed the Director to consider any action as significant on
a case-by-case basis.

ii.   Discussion of Significant Comments

      Comment:  Several comments stated that there is a need for a succinct
definition of an action in the definition of a significant lowering of water
quality.

      Response:  EPA takes the term "action" or "activity" to have a broad
meaning with respect to determining what may or may not be subject to
antidegradation.  Although the final Guidance includes examples of actions and
activities, it is not possible to anticipate each activity or action that
might result in an increased loading of pollutants or otherwise lead to a
deterioration of water quality.  Anything more than a broad definition could
constrain States and Tribes from requiring antidegradation reviews where
appropriate.  Therefore, EPA does not agree that including a definition of the
terms "action" or "activity" would be appropriate; States and Tribes need to
retain a measure of flexibility to adapt their antidegradation policies and
implementation procedures to the circumstances encountered in the day-to-day
operation of a water quality management program.  However, based on comments
received, EPA has determined that some commenters believe certain activities
to be potentially considered an "action" which EPA believes do not constitute
an action, accordingly, EPA has listed these activities in the final rule, and
noted that they are not subject to antidegradation review.

      Comment:  A number of commenters stated that it is inaccurate to assume
that an increase in.the loading of a pollutant to a water body will result in
a lowering of water quality in that water body.

      Response;  An antidegradation review is required whenever a
"significant" lowering of water quality is considered.  For BCCs, because of
their persistence and ability to exert profound impacts through accumulation
in the food chain, the final Guidance defines any increased loading of such
pollutants as a significant lowering of water quality.

      Comment;  A number of commenters stated that the "catch-all" provision
allowing any action to be considered significant is too broad.

      Response;  As stated above in the discussion of the term "action,"
States and Tribes need a certain amount of flexibility to be able to respond
to all the various types of situations that may arise as they attempt to

-------
214    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

implement this Guidance.  The "catch-all" provision gives them the ability to
consider whether individual actions that might lower water quality and are not
covered explicitly by the final Guidance are of concern, and to require an.
antidegradation review, as appropriate.  This flexibility is important and is
therefore retained in the final Guidance.  However, EPA has revised this
provision to clarify that it applies to deliberate activities.

      Comment r  A number of commenters stated that the term "significarit" is
ambiguous and should be deleted.

      Response:   The term "significant" does not stand alone, but is rather a
part of the term "significant lowering of water quality."  EPA has provided,
in the final Guidance, additional language to the definition of the term so
that the perceived ambiguity is minimized.

iii.  The Final Guidance

      The final Guidance differs from the proposed Guidance in several
important ways.   First, the EEQ provisions contained in the proposed guidance
for identifying significant lowering of water quality for BCCs were dropped in
favor of requirements that facilities provide notification and an
antidegradation demonstration in support of actions that are expected to
result in any increase in loading of BCCs prior to taking the action.
Although the underlying concept that any increase in loadings of BCCs to the
Great Lakes System warrants review under antidegradation remains intact, the
mechanism for triggering reviews was modified.  This change addresses comments
received and is discussed in greater detail in the section of this document
addressing antidegradation implementation.

      With respect to non-BCCs, the most significant change is that States and
Tribes are no longer required to adopt a definition of significant lowering of
water quality that is consistent with the definition contained in the final
Guidance  States and Tribes are only required to adopt definitions as needed
to support their implementation procedures, consistent with their
antidegradation standard and implementation procedures.  At a minimum, EPA
recommends that any activity that is expected to increase loadings of
pollutants such that relaxed permit limits are required, or which will result
in a new discharge of pollutants, should receive an antidegradation review.

      If a State or Tribe chooses to link antidegradation review to changes in
permit limits for pollutants other than BCCs, the State or Tribe must also
ensure that when an action at a point source increases the loading of a
pollutant that was not subject to a previous permit limit, that  increase is
subject to antidegradation review.  Examples of activities that  could receive
an antidegradation review include installation of a new source, a change in
process at an existing source that results in the addition of new pollutants
into the waste stream or the tie-in of a new industrial user to a municipal
treatment plant.  Similarly, for loadings of non-BCCs from nonpoint sources,
antidegradation review could be required when a new project seeks
authorization from the State or Tribe.  The revisions discussed above make the
implementation of antidegradation more equitable by subjecting both new and
existing dischargers to the same requirements.

      Antidegradation review is not required when a new limit is imposed as a
result of new monitoring data, improved monitoring techniques or a change in a
wasteload allocation or TMDL that results in no increase in loading or a de
minimis increase in loading.  Under the final Guidance, antidegradation review
for new limits is only necessary when there  is or will be a significant
lowering of water quality.

f.    Deleted Definitions

      Two definitions were included in the antidegradation section of the
proposed Guidance that are not included in the final Guidance.   These were  the

-------
                             Section VH: Antidegradation                          215
definitions of  "bioaccumulative chemical of concern" and "pollutant."
Definitions for these terms are included in the general definition section of
the final Guidance at section II B, making it unnecessary to include them in
appendix B.

g.    Implementation Procedures

i.    Background

      The proposed Guidance required that, for all waters, the level of water
quality necessary to protect existing uses be maintained.  The proposed
Guidance also specified that the uses designated by Tribes and States must
include existing uses.  Where water quality criteria for a particular
parameter are not attained, the proposed Guidance prohibited further
degradation of water quality for the parameters in question.

      For high quality waters, the proposed Guidance restricted when
significant lowering of water quality could occur to those instances where the
discharger demonstrates that the significant lowering of water quality is both
necessary and will support important social and economic development.
Different measures of significance were proposed for BCCs and non-BCCs.

      For BCCs, the proposed Guidance defined any increase in loading as
significant.  EEQ was selected to serve as the baseline against which
increased loadings would be measured.  EEQ was structured so as to track
effluent quality and could be reduced in subsequent permits to reflect load
reductions or increased to reflect loading increases approved under
antidegradation.  Tribes and States were given the option of either imposing
EEQ-based effluent limits in discharge permits or including provisions in
discharge permits requiring an antidegradation review if an EEQ value were
exceeded.  Tribes and States were also required to prohibit actions by
dischargers that would increase loadings of BCCs without prior approval of the
Director pursuant to section IV of the proposed Guidance.

      Where the proposed increase involved non-BCCs, a significant lowering of
water quality was defined as an increase in permit limits greater than de
minimis.  If a discharger was able to operate below permit limits such that an
increased loading from the discharger would not exceed existing permit limits,
no antidegradation review would be required.  Similarly, if the proposed
increase in permit limits was less than a de minimis amount, no
antidegradation review would be required.

      For ONRWs, the proposed Guidance stated that water quality must be
maintained and protected.  The proposed Guidance did not permit permanent
degradation of water quality in ONRWs.  Short-term and temporary degradation,
replacement of a failing septic system for example, could be permitted at the
discretion of the Director.

      The proposed Guidance also included provisions for Lake Superior if
Tribes and States wished to adopt special protection designations for the
lake.  Two special protection designations were included in the proposed
Guidance, Lake Superior Basin - Outstanding National Resource Waters (LS-ONRW)
and Lake Superior Basin - Outstanding International Resource Waters  (LS-OIRW).
The LS-ONRW designation could be used by Tribes and States to prohibit new or
increased discharges of certain listed pollutants, known as Lake Superior
bioaccumulative substances of immediate concern (BSICs), to portions of the
Lake Superior basin so designated.  The LS-OIRW designation could be used by
Tribes and States to require more stringent treatment for new or expanding
sources of BSICs.

      In addition, the implementation section included a description of the
circumstances under which the proposed implementation procedures would not
apply.  These included short-term and temporary lowering of water quality,

-------
216    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

bypasses not prohibited at 40 CFR 122.41(m) and certain actions under the
Comprehensive Environmental Response, Compensation and Liability Act.

ii.   Discussion of Significant Comments

      The majority of the comments received on the antidegradation
implementation, procedures addressed EEQ.  The comments received are summarized
below.

      Comment:  Use of EEQ is a disincentive to optimal waste water treatment
beyond the minimum necessary to achieve permit limits.  Dischargers that do so
will be "penalized" by receiving more stringent permit limits, whereas
dischargers that do the bare minimum receive less stringent effluent limits.

      Response;  EPA modified the final Guidance to address this concern.
Effluent limits are no longer required as the mechanism for implementing
antidegradation.  Instead, the final Guidance builds off of existing reporting
requirements and -requires an antidegradation review prior to commencement of
any activity that has the potential to result in an increased loading of BCCs
to waters within the Great Lakes System.  There are a number of advantages to
the mechanism included in the final Guidance for identifying when a
significant lowering of water quality will occur over what was in the
proposal.  First, the mechanism included in the final Guidance requires
consideration of antidegradation at the appropriate point in the process,
before the lowering-of water quality occurs and while the project is still in
the planning stages.  EEQ, on the other hand, serves a mechanism for
identifying failure to consider antidegradation and thus has a. punitive rather
than preventative function.  Second, unlike EEQ, the mechanism contained in
the final Guidance does not expose dischargers to enforcement actions solely
as a result of unusual effluent variability.  Also, by linking antidegradation
requirements to actions taken by a discharger, there is no danger of a
discharger being forced to undergo spurious antidegradation reviews to justify
apparent increases in loadings.  Finally, since antidegradation is independent
of effluent limits, there is no disincentive to efforts by dischargers to
optimize waste water treatment.

      Comment:  A significant number of commenters stated that use of EEQ
deprives dischargers of compliance benefits of improved waste water treatment.
Dischargers frequently achieve better effluent than is required by their
permit limits in order to ensure compliance with effluent limits and account
for process variability.  The EEQ provisions in the proposed Guidance
eliminated this option and exposed dischargers to greater enforcement
liability.

      Responses  Revisions to the implementation procedures in the final
Guidance address this concern.  Under the final Guidance, antidegradation
considerations for BCCs are uncoupled from  effluent limits.  As a result, if
dischargers wish to reduce loadings to improve compliance, they may do so
without a corresponding change in applicable effluent limits.  Consequently,
dischargers will be able to build as large  a margin between the limits
contained in their control documents and their actual effluent quality as they
choose.

      Comment:  EEQ is burdensome to dischargers and regulators.  The EEQ
requirements in the proposed Guidance will  necessitate  increased monitoring by
dischargers.  In addition, the need to analyze data and recalculate EEQ  at
each permit reissuahce will be difficult for regulators and slow the permit
issuance process.

      Response;  The final Guidance does not require calculation of permit
limits to implement antidegradation for BCCs.  Thus, no additional monitoring
or reporting is  likely to be necessitated  by the final  Guidance.  Although  the
final Guidance does specify monitoring requirements for discharges that  are
known or suspected  containing  BCCs,  this requirement should not be onerous  as

-------
                             Section VII: Antidegradation                          217
in most case, regardless of the antidegradation provisions, monitoring would
be necessary to assess compliance and provide data for States' and Tribes'
reasonable potential procedures.

      Comment;  Many commenters felt that the EEQ provisions create
uncertainty, making it difficult for dischargers to anticipate regulatory
requirements.

      Response:  The changes in the antidegradation implementation procedures
incorporated in the final Guidance should reduce the uncertainty of
dischargers regarding future discharge limits.  Dischargers will receive
limits for BCCs based on the reasonable potential and limit calculation
procedures adopted by States and Tribes consistent with the final Guidance.
For BCCs, antidegradation will be administered separately from numeric limits
on BCCs.  Further, the procedures contained in the final Guidance remove the
uncertainty associated with the proposed approach that resulted from limits
that became tighter over time as a result of antidegradation.  As a result,
dischargers will have greater certainty about what future control document
requirements will be.  For a discussion of the anticipated costs of
implementing the final Guidance, see the discussion of Costs and Benefits,
section IX of this document.

      Comment:  Commenters believed that monitoring associated with an EEQ-
based approach will be excessive.

      Response:  While it could be argued that the proposed approach would
have led to increased monitoring, under the final Guidance monitoring
requirements for dischargers should not be increased significantly as a result
of the antidegradation provisions.  In most cases, the requirement to monitor
for BCCs that is included in the final Guidance will be met through normal
monitoring to measure compliance with control document requirements and
support control document reissuance.  No unusual or more sensitive monitoring
is envisioned as a result of antidegradation.

      Comment:  A number of commenters stated that the EEQ provisions are
inconsistent with the TMDL process and the watershed approach.  Commenters
stated that the antidegradation procedures undermined the TMDL process and
EPA1s watershed protection goals by allowing limits more stringent than those
derived through the TMDL process, making the TMDLs irrelevant.

      Response:  EPA does not agree that antidegradation is incompatible with
TMDLs.  First, TMDLs and antidegradation address different and complementary
components of- the water quality program and both are necessary and required
for the protection of water quality in the Great Lakes System.  TMDLs, in the
simplest sense, establish the maximum amount of a pollutant that may be
discharged into a water body without exceeding water quality criteria
necessary to protect designated uses.  Antidegradation, on the other hand,
protects existing water quality where water quality is better than the
criteria.  Thus, conceptually, protection of high quality waters under
antidegradation is concerned with water quality that is better than the
minimum required by a TMDL.  Second, TMDLs and antidegradation are likely to
address different subsets of water bodies.  At present, TMDLs are only
required to be calculated for threatened or impaired water bodies and result
in imposition of water quality-based effluent limits sufficient to ensure
compliance with water quality standards.  In contrast, the effects of the
final antidegradation Guidance will be felt most strongly by dischargers to
waters that attain water quality standards and whose effluent limits are
technology-based.  Finally, unlike the proposed Guidance, in the final
Guidance, antidegradation is not implemented through effluent limitations for
BCCs.  In practice, antidegradation and TMDLs each reinforce the other, and
either one or the other of the two processes will be the important controlling
mechanism in a given water body for a given pollutant.

-------
218    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

iii.  The Final Guidance

      The final Guidance incorporates a number of changes in response to
comments received during the public comment period on the proposed Guidance.
These include revisions to the treatment of BCCs in high quality waters.

      EPA received a significant number of comments on the use of EEQ in the
proposed Guidance as a means of controlling loadings of BCCs.  Numerous
comments opposed the use of EEQ in the proposal on the grounds that it would
be a disincentive to optimal waste water treatment and would deprive
dischargers of any "cushion" they create between their effluent  limits and
actual effluent quality to ensure compliance.  However, other commenters
supported the reasoning in the proposed Guidance that led to the proposed use
of EEQ, namely that the environmental risk from BCCs is so great that any
increased loading should be considered significant.

      To reconcile these two positions, the final Guidance employs an approach
that, while not linked to specific, numeric limits, requires an
antidegradation review any time a discharger undertakes an activity that could
result in an increased loading of BCCs.  The new approach will be implemented
through notification requirements in a discharger's control document.
Antidegradation review will occur when an action is proposed that is likely to
result in an increased loading of BCCs.  No regulated facility would be
permitted to take an action that would result in an increased loading of BCCs
without completing an antidegradation demonstration and receiving approval
from the State or Tribe with jurisdiction over the affected waters.  Taking an
action without receiving approval would be a violation of the applicable
control document and would be subject to enforcement.   EPA selected this
approach because it eliminates perceived problems with EEQ and yet is
consistent with defining any increase in loadings of BCCs as a significant
lowering of water quality.  Also, it places greater emphasis on  the prevention
and minimization of any significant lowering of water quality and less on use
of antidegradation as punitive measure against dischargers.  As  envisioned by
the final Guidance, antidegradation is a tool to be used by States and Tribes
to make rational, well-supported decisions regarding activities  that affect
water quality.  For dischargers, antidegradation should be viewed as a reality
check on a proposed project, that all possible opportunities to minimize
impacts on water quality have been considered and implemented, as appropriate.
States and Tribes may employ EEQ as a means of implementing antidegradation if
they so choose.

      A further benefit of the approach taken in the final Guidance is that it
does not require quantifiable effluent concentrations for implementation.
Where the available data are insufficient to calculate a loading level but
BCCs are known to be discharged, as in the case where discharge  concentrations
are below levels of quantification, increased loadings of BCCs are controlled
by a requirement in the control document prohibiting a discharger from taking
any deliberate action that would increase loadings of BCCs without first
providing an antidegradation demonstration and receiving approval for the
increase from the Director.  Thus, where loadings of BCCs are occurring, but
are not quantifiable, increases are still subject to antidegradation.

      It should be noted that requiring a discharger to receive  approval prior
to taking an action that will result  in a lowering of water quality is not a
new requirement.  The intent of the antidegradation policy contained in the
Federal regulations at 40 CFR 131.12  is that antidegradation review of  an
activity that will lower water quality occur before the activity takes place,
when the antidegradation review will be most effective in identifying ways
that degradation of water quality may be minimized.

      For non-BCCs, EPA recommends that, at a minimum antidegradation review
be required any time a permit limit is made less stringent.  States and Tribes
must also ensure that the antidegradation policy and procedures  they adopt
address new sources and increasing sources where no permit limit currently

-------
                             Section VII: Antidegradation                          219
exists for a pollutant.  States and Tribes may also include a provision in
their procedures that allows them to review any action that is not
specifically addressed to provide the flexibility to respond to unforeseen
circumstances.

3.    Antideqradation Demonstration

a.    Background

      Existing Federal regulations at 40 CFR 131.12 require that water quality
in a high quality water be maintained and protected unless a lowering of water
quality is necessary to support important social and economic development.
The phrase "antidegradation demonstration" is used as shorthand for the
information gathered for review by the permitting agency to show whether or
not an action proposed by a discharger that will result in a significant
lowering of water quality both is necessary and will support important social
and economic development.

      The proposed Guidance required an antidegradation demonstration be
performed prior to any significant lowering of water quality.  The proposed
Guidance interpreted the Federal regulations at 40 CFR 131.12 as requiring two
distinct demonstrations; first, that the significant lowering of water quality
was "necessary," meaning the proposed activity could not occur without a
concomitant significant lowering of water quality, and second, that the
proposed activity would result in social and economic development.  Consistent
with Federal regulations, the proposed Guidance organized the antidegradation
demonstration into a hierarchy of three separate demonstrations to be made by
the discharger.  The discharger was required to demonstrate that the
significant lowering of water quality could not be reduced or prevented
through the application of prudent and feasible pollution prevention
alternatives, that the significant lowering of water quality could not be
prevented through the application of alternative or enhanced treatment within
a defined cost range and finally, if, as a result of the preceding
demonstrations, a significant lowering of water quality was shown to be
necessary, that the proposed activity would result in social and economic
development.  Each of the demonstrations included certain required elements to
be considered by dischargers conducting an antidegradation demonstration.

b.    Discussion of Significant Comments

      Comment:  Tribes and States lack the technical expertise required to
review antidegradation demonstrations.

      Response:  The final Guidance identifies the broad areas that should be
a part of any antidegradation demonstration.  This document identifies key
areas that should be considered in evaluating an antidegradation
demonstration, which will assist Tribes and States in their reviews.  Using
the final Guidance, Tribal and State water quality personnel should be capable
of determining whether or not an antidegradation demonstration prepared by a
discharger considers all of the essential elements.

      It should also be noted that preparing a demonstration may in and of
itself be useful to the proponent of an activity that would significantly
lower water quality.  In order to request a significant lowering of water
quality to accommodate a particular activity, a discharger must investigate
alternatives to lowering water quality.  This investigation may lead the
discharger to cost-effective alternatives to lowering water quality.

      Comment:  More detailed guidance is needed on all aspects of the
demonstration.

      Response;  In reviewing the comments received, EPA is convinced that it
is not possible to write guidance that would cover adequately every possible
situation.  In addition, including great detail would hinder efforts by Tribes

-------
220    Water Quality Guidance for the Great Lakes System - Supplementary Information Document

and States to adapt the final Guidance to existing regulatory structures and
thereby slow its ability to respond to requests to lower water quality.
Consequently, the final Guidance provides a framework within which Tribal and
State programs must, operate with respect to BCCs.  This general  approach will
also prove more versatile as new circumstances arise that were not foreseen
when the Guidance was developed.  Further, general guidance allows Tribes and
States to tailor antidegradation reviews, including level of detail and
documentation,- to the specific circumstances encountered.

      Comment:  A number of commenters noted that the costs of pollution
prevention need to be considered.

      Response;  The final Guidance has been revised to require  that a
discharger identify "cost-effective" pollution prevention alternatives that
will eliminate or greatly reduce the extent of the significant lowering of
water quality.  Alternatives that would cause the proponent of the activity to
incur excessive costs would not be considered cost-effective; however,
pollution prevention can be the most cost-effective and environmentally benign
approach to protecting the environment.

      EPA's intent in having pollution prevention figure so prominently in the
antidegradation demonstration is to focus attention on alternatives that will
not lead to the release of pollutants to the environment rather  than on those
that depend upon treating the pollution after it is generated.   Such an
emphasis is consistent with the hierarchy outlined in EPA's National Pollution
Prevention Policy (see discussion in section I of this document).

      Comment;  Several commenters stated that the ten percent mandatory
expenditure requirement for enhanced and/or alternate treatment  is arbitrary.

      Response:  EPA agrees that the ten percent mandatory expenditure
requirement may limit the ability of States and Tribes to consider individual
circumstances in reviewing antidegradation demonstrations.  As a result, the
ten percent additional cost benchmark is included in this document as guidance
only.  EPA realizes that the determination of what represents affordable
treatment options is specific to the case in question.  Therefore, a strict
cut-off at ten percent additional costs is not realistic.  Greater costs may
be affordable in some cases; in others, ten percent may be too expensive.  In
the final Guidance, the determination of what treatment alternatives are
practicable is left to Tribes and States.  EPA is developing additional
National guidance on a variety of issues related to economic considerations in
water quality standards that will provide direction to Tribes and States as
they implement their antidegradation policies.  It is important  to note that
the affordability measures discussed above are separate and distinct from any
determination of penalties or ability to pay within the context  of an
enforcement action.

      Comment;_  Some commenters suggested that the cost-effectiveness
alternative described in the preamble to the proposal is preferable to the
approach taken in the proposed Guidance.

      Response:  Although EPA considers the ten percent additional cost
benchmark to be the easiest measure of affordability to implement, the final
Guidance does not mandate any one approach to identifying available treatment
options.  Tribes and States are free  to use any approach they choose that is
appropriate to the specific situation under consideration.  The  increased
flexibility of the antidegradation requirements of the final Guidance will
make it more sensitive to the circumstances of each individual  situation.  Any
benchmark chosen by a State or Tribe  to determine whether or not additional or
alternative treatment is affordable is relevant only within the  context of an
antidegradation demonstration for purposes of evaluating information provided
in support of a request to lower water quality in a high quality water.

-------
                             Section VH: Antidegradation                          221
      Comment:  Some commenters believed that expenditures should be required
even if improvements are only incremental.

      Response:  The final Guidance does not preclude Tribes or States from
granting only partial approval of requests to lower water quality.  Where a
Tribe or State believes that a project may occur with a less extensive
lowering of water quality than was requested, control documents may be issued
with limits that reflect the lowering of water quality necessary to
accommodate the activity.

      Comment:  Several commenters stated that the antidegradation
demonstration should establish a direct link between the degradation of water
quality and the social and economic development.

      Response:  EPA agrees with this comment in principle.  In the final
Guidance, a demonstration of the important social and economic development
that will result fr6m the proposed significant lowering of water quality is a
precondition to approving a significant lowering of water quality.  The other
components of the antidegradation demonstration, the cost-effective pollution
prevention analysis and the enhanced and alternate treatment analysis are
intended to ensure that social and economic development takes place with a
minimum of environmental impact.  However, it is not always possible to
quantify exactly how much degradation is associated with a certain activity
that will produce social and economic development.

c.    The Final Guidance

      The final Guidance retains the framework contained in the proposed
Guidance, but with less detail.  The demonstration is divided into pollution
prevention, alternative and enhanced treatment and social and economic
development components that are performed in series.  States and Tribes are
required to adopt the demonstration elements contained in the final Guidance
for purposes of regulating new and increasing discharges of BCCs.  For non-
BCCs, States and Tribes may either adopt the procedures contained in the final
Guidance, or develop their own antidegradation demonstration requirements
provided they are consistent with the requirements of 40 CPR 131.12.

      The first two- components of the demonstration, the pollution prevention
alternatives and the enhanced and alternative treatment demonstrations,
address the question of whether or not water quality must be lowered to
accommodate the proposed activity (i.e., is the significant lowering of water
quality necessary).  EPA's intent in mandating these tests is to ensure that,
when they are available, feasible alternatives that would allow an activity to
occur with little or no degradation of water quality are identified and
implemented and the lowering of water quality is minimized.  In keeping with
EPA's preference for source reduction over waste treatment, dischargers are
first directed to identify any and all cost-effective pollution prevention
alternatives that might eliminate or reduce the proposed significant lowering
of water quality,  if the significant lowering of water quality cannot be
eliminated entirely through pollution prevention, then alternative or enhanced
waste water treatment should be considered.  Only after pollution prevention
and alternative and enhanced treatment are examined and no feasible
alternative to the significant lowering of water quality is found can it be
considered necessary.  A more detailed discussion of the components follows.

i.    Identification of Cost-Effective Pollution Prevention Alternatives to
Prevent or Reduce the Significant Lowering of Water Quality

      This is the starting point for the antidegradation demonstration.  The
proponent of the activity should consider five broad categories of pollution
prevention activities in determining whether or not alternatives exist that
would reduce or eliminate the anticipated significant lowering of water
quality.  These include:

-------
222    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      -- Substitution of non-bioaccumulative or non-toxic chemicals for BCCs.
The primary objective of this evaluation is to determine if the source of a
BCC, which would otherwise cause or contribute to a significant lowering of
water quality, can be eliminated in favor of a less environmentally
problematic substance, especially one that is not a BCC.

      -- Application of water conservation methods.  The objective of this
portion of the pollution prevention analysis is to determine whether or not
reductions in the overall volume of waste water are possible and would reduce
pollutant loadings so that the proposed activity could occur without a
significant lowering of water quality.

      -- Waste source reduction within process streams.  The objective is to
evaluate all waste streams involved in the process affected by the proposed
activity.  Opportunities to control more carefully the use of raw materials
and reduce waste should be identified and implemented where feasible.

      -- Recycle or reuse of waste byproducts, either liquid, solid or gas.
The objective is to identify ways in which recycling and reuse of internal
waste streams can be employed to reduce the loadings of pollutants to the
environment.  This is a common practice in industry and can reduce energy, raw
material and waste disposal costs.  The proponent of the proposed activity
should investigate the process involved to determine whether or not
opportunities exist to implement such changes where they would alleviate the
need for a significant lowering of water quality.

      -- Manufacturing Process Operational Changes.  The focus of this part of
the investigation should be to identify different means of achieving the
desired end that will produce either smaller quantities of toxic waste
products or waste products that are less toxic.  All of the processes that
will contribute to the significant lowering of water quality should be
examined and alternatives that would reduce or eliminate the need to lower
water quality significantly should be identified.

      The pollution prevention possibilities discussed above are not intended
to be all inclusive.  Dischargers seeking approval for an action that will
result in a significant lowering of water quality should evaluate all aspects
of the proposed activity for opportunities to reduce pollutant loadings.  The
categories of pollution prevention discussed above should be viewed as
guidance to those performing an antidegradation demonstration and those
evaluating it.

      Pollution prevention is applicable to municipal as well as industrial
dischargers.  Where a significant lowering of water quality will result from
increased industrial use of a municipal treatment plant, the municipality
should ensure that the loadings from the industrial users are minimized
through pollution prevention to the extent possible.  Tools available to
municipalities to encourage pollution prevention include local limits and fees
levied on industrial users.  Municipalities should also consider public
education and bans on certain substances to reduce loadings from nonindustrial
sources.

      The determination that an alternative or combination of alternatives is
cost-effective is the decision of the Director, and not the entity that is
seeking to significantly lower water quality.  States and Tribes are
encouraged to develop their own guidelines  to follow when evaluating pollution
prevention alternatives identified by the entity to select those that are
cost-effective.

ii.   Alternative or Enhanced Treatment to  Eliminate the Significant Lowering
      of Water Quality

      The second part of determining whether  or not a significant lowering of
water quality is necessary is the alternative or enhanced treatment analysis.

-------
                             Section VII: Antidegradation                          223
This analysis should be undertaken after the pollution prevention analysis is
completed and should focus on removing the remaining incremental increase in
pollutant loadings after cost-effective pollution prevention measures are
taken.  If application of pollution prevention techniques alone are sufficient
to eliminate the significant lowering of water quality, the alternative or
enhanced treatment analysis need not be performed.

      The objective-of the alternative or enhanced treatment analysis is to
ensure that the actual degradation of the high quality water is reduced to the
greatest extent practicable.  The analysis proceeds by identifying the least
costly options for additional treatment under which the proposed activity
could occur without resulting in a significant lowering of water quality.  The
costs of the different treatment options are determined and compared to the
costs of the treatment needed to achieve all applicable standards, including
Federal effluent guidelines, water quality-based effluent limits and all other
applicable Federal and State or Tribal requirements.  Where treatment options
are identified that are comparable in cost to baseline treatment costs and
allow the proposed activity to occur without leading to a significant lowering
of water quality, those treatment options should be implemented in lieu of
lowering water quality.

      In the proposed Guidance, EPA considered using a ten percent increase in
treatment costs as a benchmark for determining whether or not alternative or
enhanced treatment options identified through this analysis were affordable.
Where treatment options with costs up to ten percent more than the cost of
meeting all applicable discharge requirements were identified, the discharger
would have been required to implement the treatment option identified and the
request to lower water quality would be denied.  In the final Guidance, the
ten percent cost benchmark is provided as guidance only.  Tribes and States
may use the ten percent value to assist them in evaluating antidegradation
demonstrations.  Tribes and States may also use other appropriate means of
identifying practicable treatment options such as the cost-effectiveness
approach described in the preamble to the proposed Guidance (see 58 FR 20910).
The final Guidance does not provide specific direction to Tribes and States on
how to determine the affordability of a treatment option under consideration
because to do so would limit their ability to respond to the unique
circumstances of each antidegradation demonstration.  Tribes and States retain
the ability to require treatment beyond the benchmarks discussed in the
preamble to the proposal, this document (SID), proposed or final Guidance, or
to require treatment that will only reduce the magnitude of the significant
lowering of water quality.  Tribes and States should also see EPA's draft,
Economic Guidance for Water Quality Standards Workbook. (November 1993) for
additional guidance on how to evaluate the affordability of treatment options.

iii.  Important Social and Economic Development

      In the final part of the antidegradation demonstration,  the proponent of
the activity that will result in a significant lowering of water quality must
show that the significant lowering of water quality proposed will support
social and economic, development.  This part of the demonstration should occur
only if no pollution prevention or alternative treatment options are
identified that will eliminate the need for a significant lowering of water
quality.  In determining whether or not a proposed activity will support
important social and economic development, Tribes and States should consider
the geographic area in which the significant lowering of water quality will
occur, the current or baseline economic condition of that area, the net
positive impacts that will result from the proposed activity and the
possibility of other development occurring in the area that will result in
similar economic and social benefits but will not cause a significant lowering
of water quality.

-------
224    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      -- Identification of the Affected Area

      The area in which the economic benefits occur should correspond with the
area in which water quality is lowered.  Determining the area is a case-by-
case decision, made taking into account the pollutants involved as well as the
location of the discharge.

      - - Baseline Economic Condition

      Once the affected area is identified, the baseline economic condition of
the area should be assessed.  Factors that may be useful include unemployment
rates, percentage of the population living below poverty levels, percentage of
the population that*are elderly and average household income relative to State
and National averages.

      -- Net Positive Impact

      In determining net positive impact the Tribe or State should attempt to
assess the benefits of the proposed activity corrected for any negative
economic impacts of the activity.  The types of benefits from the activity to
be considered include an increase in the number of jobs, an increase in
personal income and/or wages, reduction in unemployment rates or social
service expenses, increased tax revenues and provision of necessary social
services.  Other measures may be relevant on a case-by-case basis.

      Adverse economic impacts may also result from an activity that supports
social and economic development.  For example, a new industrial facility may
provide additional jobs in a community; however it may also make the receiving
water less attractive for recreation and cause a loss in tourism dollars.
Such impacts should be considered in determining whether or not a project or
activity that will result in a significant lowering of water quality will also
support important social and economic development.

      - - Other Developments

      Tribes and States should also consider whether a proposed activity will
preclude another activity that may not affect water quality yet yield
comparable social and economic benefits.  In the example above, the siting of
an industrial plant may preclude water front development or building of a
marina that would provide comparable social and economic development at less
cost to the environment.

4.    Antidegradation Decision

a.    Background

      The final section of the proposed Guidance established the process to be
used by the Director for arriving at a final decision regarding whether or not
to allow a significant lowering of water quality.  The proposed Guidance
specified how the information obtained through each of the antidegradation
demonstration components was to be evaluated and presented two alternatives
for factoring public comment into the final decision.  Under the proposed
Guidance, public comment could be sought following a tentative analysis of the
available data, or the data analysis could be deferred and public comment
sought on a tentative decision to deny the request to lower water quality.

b.    Discussion of Significant Comments

      Comment:  Many commenters stated that Tribes and States should be
allowed greater flexibility in how data are used and public comment is sought
in arriving at a final decision regarding an antidegradation demonstration.

      Response:  EPA agrees with  the views expressed by the commenters.  The
final Guidance no longer specifies how States and Tribes  should consider the

-------
                             Section VII: Antidegradation                          225
data and public input they obtain in making a final decision.  Also, States
and Tribes are required to adopt a procedure consistent with the final
Guidance for BCCs only; States and Tribes may develop their own decision-
making procedures for non-BCCs consistent with the requirements 40 CFR 131.12.

c.    The Final Guidance

      The final Guidance is less prescriptive than the proposed Guidance in
that the process for arriving at a decision is left to individual Tribes and
States.   States and Tribes are only required to adopt procedures consistent
with the Guidance for BCCs.  However, the essential requirements remain the
same and the options presented in the proposed Guidance may be used by Tribes
and States.  The Director must examine each of the demonstrations submitted by
the proponent of the activity that will lower water quality.  Also, an
opportunity for public comment must be provided and public comment must be
factored into the final decision, consistent with the antidegradation
standard.  EPA expects that the process of making a decision will parallel the
antidegradation demonstration process.  The Director should first evaluate the
analysis of cost-effective pollution prevention alternatives and determine the
extent to which pollution prevention can reduce or eliminate the significant
lowering of water quality.  The Director should then review the information
generated through the alternative and enhanced treatment analysis.  If the
analysis identifies affordable treatment options that, combined with the
pollution prevention alternatives, will eliminate the need to lower water
quality significantly, the Director should deny the request to lower water
quality.  If, however, the pollution prevention and alternative treatment
analyses are unable to eliminate the need to lower water quality, the Director
should weigh the social and economic development that will result from the
significant lowering of water quality in the affected area.  If the proposed
activity is found tp support important social and economic development, the
Director may decide to grant all or part of the requested significant lowering
of water quality, provided water quality sufficient to protect existing and
designated uses is maintained and provided the decision is subject to public
comment.

      Opportunity for public comment is an essential element of the
antidegradation decision making process and is required under Federal
regulations at 40 CFR 131.12.  If the tentative decision relates to an
activity subject to a NPDES permit,  the public participation requirements may
be fulfilled by the public notice of the draft permit and fact sheet.  In any
event,  the public notice of the tentative decision must either set forth the
extent to which water quality will be significantly lowered and the basis for
the tentative decision to allow the lowering,  or, if analysis of the
demonstration has been deferred, a tentative decision to deny the request to
lower water quality pending public comment and analysis of the information
obtained through the antidegradation demonstration.

-------
226     Water Quality Guidance for the Great Lakes System — Supplementary Information Document

-------
                        Section VELA: Site-Specific Modifications                     227
             .  IMPLEMENTATION PROCEDURES


A.    Site-Specific Modifications to Criteria and Values

1.    General

      As discussed in the proposed Guidance,  there  currently exists National
guidance to modify aquatic life 'criteria on a site-specific basis, but there
is no such guidance for modifying human health or wildlife criteria on a site-
specific basis (see 58 PR 20918).  The proposed Guidance continued to allow
more stringent site-specific modifications  to the wildlife and human health
criteria or values and to bioaccumulation factors  (BAFs).  Sections 1 through
4 below discuss the changes to the site-specific modifications for aquatic
life, wildlife, BAFs, and human health.

2.    Aquatic Life

      a.    Proposal:  The proposed Guidance,  consistent with the current
national Guidance, provided that  Great Lakes States and Tribes may adopt site-
specific modifications allowing more stringent or less stringent aquatic life
criteria or values when local water quality characteristics such as pH,
hardness, temperature, color, etc.,  alter the biological availability or
toxicity of a pollutant; or criteria when the sensitivity of the local aquatic
organisms (i.e.,  those that would live in the water absent human-induced
pollution) differs significantly  from the species actually tested in
developing the criteria.  The proposal suggested use of the existing national
guidance provided in the U.S. EPA Water Quality Standards Handbook (1983)
(1983 Handbook) when modifying the criteria.   (Since the proposal, the 1983
Handbook has been reformatted and republished as the U.S. EPA Water Quality
Standards Handbook, Second Edition - Revised,  1994)  (Revised Handbook) The
Revised Handbook is available in  the docket for this rulemaking.

      The proposed Guidance went  beyond existing national guidance by also
allowing the Great Lakes States and Tribes  to develop site-specific
modifications to chronic aquatic  life criteria for  the Great Lakes System to
reflect local physical and hydrologic conditions. Such conditions might
include natural features of a water body, such as lack of proper substrate,
cover, flow, depth, .pools, and riffles, unrelated to ambient water quality.
Specifically, such conditions would include any local physical or hydrological
condition which precluded aquatic life from remaining at the site for 96 hours
or more.  The proposal stated that sites where conditions precluded all but a
few forms of aquatic life from living may be protected by less stringent
chronic criteria.  EPA expected that this provision would typically be used
for waters where a full aquatic life use is unattainable.  The proposal did
not include such a provision for  acute aquatic life criteria.

      b.    Comments:  Many commenters supported the proposal for allowing
more or less stringent site-specific modifications  to aquatic life
criteria/values when local water  quality parameters alter the biological
availability 'or toxicity of a pollutant; or criteria when the sensitivity of
the local aquatic organisms differs significantly from the species actually
tested in developing the criteria.  A few commenters requested that EPA not
allow any less stringent site-specific modifications to aquatic life criteria.
These commenters were concerned that less stringent criteria in localized
areas in the Great Lakes System would cause declines in downstream water
quality.

      Several commenters suggested that EPA also allow less stringent site-
specific modifications for acute  aquatic life criteria to reflect local
physical and hydrologic conditions.   They reasoned  that when physical or

-------
228    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

hydrological conditions preclude aquatic life from remaining at a site  for a
period of time in which acute effects may occur, less stringent site-specific
aquatic life criteria should be allowed.  Further, commenters pointed out that
EPA offered no support for why this exception was allowed for chronic
criteria, but not acute criteria.

      EPA agrees with the commenters suggesting that there may be instances
where aquatic life might not inhabit a site due to the site's physical  or
hydrological characteristics (e.g., natural features of a water body, such as
lack of proper substrate, cover, flow, depth, pools, and riffles, unrelated to
ambient water quality).  To make such modifications, EPA will expect the State
or Tribe to demonstrate that aquatic organisms do not inhabit a site or do not
spend sufficient time at the site to experience acute effects.  The final
Guidance provides that the acute criteria may be modified to be less stringent
to reflect local physical and hydrological conditions.  Such modification may
be made to the criterion using the recalculation procedure provided in  Chapter
3 of the Revised Handbook.

      Like the National program, State and Tribal programs must ensure  that
less stringent site-specific modifications do not impair the water quality of
downstream waters. ,In addition, allowance for less-stringent site-specific
criteria must still'ensure protection of species which do "occur at the site"
(as defined in appendix F.A.l.a.ii of this rulemaking); in other words, the
absence of acute and chronic criteria is unacceptable.

      c.    Final Guidance:  The final Guidance maintains the requirements in
the proposal with the following changes:  States or Tribes are allowed  to
develop site-specific modifications to both acute and chronic criteria  to
reflect local physical and hydrological conditions.  This is a change from the
proposal which only allowed modifications for physical or hydrological
conditions to chronic criteria.  The final Guidance still allows modifications
to both acute and chronic criteria/values to reflect water chemistry and to
criteria to reflect species sensitivity differences.  The final Guidance
provides the State or Tribe with some options for achieving the requirement to
protect endangered or threatened species, such as:  (i) if the Species  Mean
Acute Value (SMAV) for a listed species, or for a surrogate of a listed
species, is lower than the calculated Final Acute Value  (FAV), such lower SMAV
is used instead of the calculated FAV in developing site-specific modified
criteria; or  (ii) the site-specific criteria may be calculated using the
recalculation procedure described in Chapter 3 of the Revised Handbook.  In
option  (ii) , acceptable toxicity data  (as defined in appendix A to part 132
for the aquatic life Tier I criteria methodology) for the listed species or a
surrogate for the listed species must be included in the data set in which the
criterion is recalculated.

      In defining a site the State or Tribe should consider the known range of
the threatened or endangered species.  Either option will ensure protection
for species listed under the Federal Endangered Species Act when adequate data
are available.  If there is a critical food source which is also an aquatic
animal, option  (ii) might provide a greater level of protection for that
listed species if that food organism were more sensitive than the listed
species itself.  By including data for the organism which is a critical food
for the listed species, option  (ii) will ensure protection of that food source
from water quality effects.  Option  (i) only considers the  sensitivity  of the
listed species itself.  EPA wishes to clarify that the recalculation procedure
contains a provision that the recalculated FAV, Criterion Maximum
Concentration  (CMC), and/or Criterion Continuous Concentration  (CCC) should be
lowered if necessary to ensure that the criterion is not likely  to jeopardize
the continued existence of any endangered or threatened species  listed  under
section 4 of the Endangered Species Act or result in  the destruction or
adverse modification of such species' critical habitat.

      For metals criteria, a State or Tribe may wish  to use total recoverable
metals criteria rather  than dissolved criteria to provide a greater  level of

-------
                        Section Vm.A: Site-Specific Modifications                     229

protection for such species.  EPA recommends use of total recoverable metals
criteria when it is known that an endangered or threatened species is
chronically sensitive to metals and lives in or on sediments and whose diet is
substantially comprised of benthic aquatic organisms.

      If the SMAV for a listed species, or for a surrogate of a listed
species, is not lower than the calculated FAV or if the recalculated criteria
(with data for the listed species included) is less stringent than the
established Tier I criteria or Tier II value, then the existing Tier I
criteria or Tier II value provides adequate protection for the listed species.

3.    Wildlife

      Many comments were received on the proposed use of site-specific
modifications to wildlife criteria.  The significant comments, including EPA
responses and changes to the proposal, are set out below.  Responses to all
comments are contained in the Response to Comments Document, which is part of
the docket for this action.

a.    site-specific Modifications

      i.    Proposal:  The proposed procedure 1 allowed wildlife criteria and
values to be modified on a site-specific basis to provide an additional level
of protection.  This additional protection could be provided through the use
of an intraspecies uncertainty factor  (UF), which was applied in addition to
the other UFs in the wildlife methodology (58 FR 20880).  The proposal did not
allow for less-stringent modification to wildlife criteria or values.

      ii.   Comments:  Many commenters agreed with the need for site-specific
modifications for wildlife, citing such factors as waterbody-specific
characteristics driving the bioavailability of the chemical in question, the
geographical representation of the representative species, and the dietary
habits of the representative species.  Many of the commenters criticized the
proposal for allowing only those modifications which would be more stringent
than the Great Lakes System-wide criteria, stating that it is not technically
defensible because certain taxa may be missing from specific locations, the
assumptions of the environmental fate of the chemical of concern may not be
valid for all sites in the system, and wildlife mobility may affect exposure
assumptions.

      EPA recognizes that there may exist situations in which a site-specific
modification, resulting in a less stringent criterion or value, may be
appropriate; therefore, the final Guidance allows for the development of site-
specific modifications to wildlife criteria that may be more or less stringent
than the Great Lakes System-wide criteria.  The following discussion provides
additional guidance for the development and application of site-specific
modifications that are more or less stringent than system-wide wildlife water
quality criteria.

      (A).  Less Stringent Site-specific Modifications

      Modifications to wildlife criteria that result in less stringent site-
specific criteria may be allowed when a site-specific BAF is derived, which is
lower than the System-wide BAF derived under appendix B to part 132.  At the
same time, it is important that a site-specific relaxation of system-wide
criteria not produce off-site impairments of designated uses due to wildlife
and prey mobility,  pollutant transport, and interbreeding of populations of
varying sensitivities.  To safeguard against that, the final Guidance requires
that a showing be made before approval of any site-specific modification is
granted that:  any increased uptake of the toxicant by prey species utilizing
the site will not cause adverse effects in wildlife populations in the Great
Lakes System; wildlife populations utilizing.the site, or boundary or
interconnecting waters will continue to be fully protected; and, the mobility

-------
230    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

of all prey organisms and wildlife populations have been adequately described
in justifying the site.

      EPA considered allowing less stringent modifications to the wildlife
criteria based on site-specific distributions of wildlife species, and their
associated dietary habits.  EPA believes that in most cases a less stringent
site-specific modification based on these factors will be difficult because
the representative species listed in the final Guidance were not selected to
be the sole species targeted for protection; rather, they were selected to
exemplify the highly exposed wildlife species resident in the Great Lakes
basin.  Hence, even if a representative species is not found at a specific
geographic site, or has a diet different than that listed in appendix D to
part 132.  other highly exposed wildlife species would still need the
protection granted by the criteria derived under appendix D to part 132, or
those criteria listed in Table 4 to part 132.  In addition, because in many
cases it is difficult to determine the relative sensitivities of the species
within a specific area, the use of the highly exposed species to represent all
species ensures that the majority of species will be protected if they are
more sensitive.  Although, EPA believes it would be difficult to demonstrate
that there are no species within an area that are not as exposed or not as
sensitive as the representative species, EPA believes it should not preclude
calculating wildlife criteria using a different set of species than those used
in deriving the four wildlife criteria.

      An important component of a site-specific modification is the definition
of the site to which the modification is applicable.  A site may range from
being a portion of a watershed to the entire part of the Great Lakes System
under the jurisdiction of the State or Tribe proposing the modification.

      (B).  More Stringent Site-specific Modifications

      States retain the authority through section 510 of the Clean Water Act
(CWA) to develop site-specific modifications that are more stringent than
system-wide criteria.  Use of the wildlife methodology described in appendix D
to part 132 is recommended, with the appropriate changes to the toxicological
or exposure parameters that are contained in that methodology.  The allometric
equations provided in appendix D to part 132, section III.E may be used if
feeding or drinking rates for the species are not available.  Because the
site-specific modifications subject to this subsection will be more stringent
than the corresponding system-wide value or criterion, the concerns behind the
demonstration described above for less stringent modifications are not
applicable.

      iii.  Final Guidance:  A Tier I wildlife criterion may be modified to be
either more or less stringent than the corresponding system-wide criterion or
value.  Where a site-specific modification  is proposed that is less stringent
than the system-wide criterion, a demonstration is needed that the site-
specific modification will not only protect on-site wildlife species but also
protect off-site wildlife species and their prey.  The demonstration must
consider the mobility of both the wildlife  species and the wildlife prey which
utilize the site.

      b.    Protection of Endangered or Threatened Species

      i.    Proposal:  The proposed Guidance did not contain any specific
provisions regarding protection of endangered or threatened wildlife species.

      ii.   Comments:  Some commenters stated that protection of the
individual, rather than the population, is  warranted in cases of endangered or
threatened species, but were concerned that the proposed Guidance was too
broad in allowing the application of  site-specific modifications for
protecting individuals beyond those listed  under the ESA.

-------
                        Section VELA: Site-Specific Modifications                     231

      Procedure 1.A of the final Guidance provides that States or Tribes must
develop site-specific modifications to protect endangered or threatened
species listed under section 4 of the ESA; it also describes a recommended
methodology to accomplish such a modification  (procedure 1.A.2.C).  The final
Guidance recommends the use of the methodology contained in appendix D to part
132, modified as described below, to provide a greater degree of protection
for endangered or threatened species.  Where available, toxicological or
exposure data for the species in question shall be used.  Where ingestion
rates are not available, the allometric equations contained in appendix D to
part 132 may be used.  Also, EPA strongly encourages the use of the
methodology for determining BAFs (Appendix B to part 132) to derive
appropriate BAFs for the site-specific modification.

      The equation from appendix D to part 132 is modified by the inclusion of
an intraspecies UF to provide additional protection for the benefit of
individual members of populations.  The recommended range of this UF is 1 to
10 and additional discussion is provided in the final Great Lakes Technical
Support Document for Wildlife Criteria.  In addition, the selection of
endpoints for protection of listed species may be broader than those described
in appendix D to part 132; however, the selection of an endpoint must be
defended by establishing its relevancy to the viability of an individual or
its ability to reproduce.

      In developing'the site-specific modification, it is important to
delineate the geographic area to which the modification shall apply.  In
designating the appropriate site, it is important to consider fully not only
the current and potential ranges of the wildlife species in question, but also
the range and mobility of the prey organisms.  Therefore, a site could, by
necessity, be defined as the entire Great Lakes System within the jurisdiction
of the State or Tribe.

      iii.  Final Guidance;  Because the equation contained in appendix D to
part 132 has been revised in accordance with comments received (see section VI
of this document), the following equation is recommended in place of the
equation presented in the proposal.  In addition, where available and
appropriate, toxicological, epidemiological, or exposure information for the
species in question is recommended to be used in the methodology.


                                       10
                             UFA x UFL x UFsx  UF,
Where the terms are defined in section II, appendix D to part 132, and also
include:

      UF, = Intraspecies Uncertainty  Factor for  extrapolating  to the  most
sensitive individuals within a population (unitless).

      The lowest of the wildlife values calculated from the above methodology
contained in appendix F.I.2 to part 132, and the two class-specific wildlife
values derived for £he system-wide wildlife water quality criterion should be
selected as the site-specific modification.

4.     Bioaccumulation Factors

      a.     Proposal:  The proposed Guidance allowed only more  stringent site-
specific modification for BAFs pursuant to authority reserved to the States
and Tribes under CWA section 510.  BAFs could be modified on a  site-specific
basis where reliable data showed that local bioaccumulation was greater than
the system-wide value.

-------
232    Water Qualify Guidance for the Great Lakes System — Supplementary Information Document

      b.    Comments:  Many commenters stated that it was not scientifically
justifiable to allow only more stringent criteria  (a higher BAF) and that EPA
should also allow less stringent criteria  (a lower BAF).

      Many commenters stated that less stringent modifications to the BAFs
should be allowed based on both chemical- and site-specific characteristics,
such as particulate organic matter and dissolved organic matter which may
modify the bioavailability of an organic chemical at a specific location; and
the partitioning of the chemical between the water column and sediment since
this may be a significant factor affecting the bioavailability of the
contaminant.

      Several commenters stated that they wanted the flexibility to adjust the
percent lipid value.to reflect that of fish consumed from a particular
location.  Other commenters were concerned that the fish species used in
deriving the BAF may not be present at a particular site, which may lead to a
more stringent value than needed at that site.

      Many commenters stated that the current BAF methodology precludes using
site-specific information.  The commenters wanted EPA to instead allow for
calculation of a BAF using field data from the site of discharge or other
modifications to reflect the characteristics of the site.  In particular,
commenters suggested modifications to the food-chain model based on chemical-,
site- and species-specific data.   Some commenters wanted EPA to develop both
the methodology and'the data for deriving site-specific BAF values.

      EPA agrees that both more and less stringent modifications of the BAFs
should be allowed on a site-specific basis if there is scientific
justification.

      EPA agrees with commenters that modifications to the BAFs should include
consideration of the bioavailability of the chemical. EPA does not agree that
the partitioning of the chemical between the water column and sediment, and
other fate processes affect the BAF.  However, modifications based on
chemical-specific characteristics are not specific to a particular site, and
therefore not appropriate for procedure 1 of appendix F to part 132.
Chemical-specific considerations are addressed in the derivation of the
system-wide BAF for the chemical.  Because the final criteria are based on the
total concentration of the chemical in the water,  the final BAF)  used in the
calculation of the criteria may be modified for site-specific particulate and
dissolved organic carbon concentrations.

      EPA also agrees that site-specific modifications should be allowed if it
can be demonstrated that the percent lipid of aquatic organisms is different
than the percent lipid values used in the derivation of BAFs.  The percent
lipid of 1.82 for trophic level three and 3.10 for trophic level four in
edible tissue for use in determining human health BAFs and the percent lipid
of 6.46 for trophic level three and 10.31 for trophic level four in whole fish
for use in determining wildlife BAFs for an organic chemical are protective of
most sites in the Great Lakes.  In cases where it  can be documented that the
percent lipid for the fish species consumed at a site differs from these
values, modifications can be made to the BAF.

      In the final Guidance, derivation of Food Chain Multipliers  (FCMs) based
upon the model of Gobas  (1993) is used because the input parameters are easily
defined and measured, the calculated BAFs are in better agreement with
measured BAFs for chemicals with very high K^s than  the  Thomann model  (1989)
used in the proposal, and the model uses equilibrium partitioning theory to
predict chemical residues in benthic organisms.  The model of Gobas  (1993) can
also be adjusted for site-specific considerations.  The FCMs in Table 1 of
appendix B to part 132 were calculated using Great Lakes-specific data.  If  it
can be demonstrated that the values for input parameters used by EPA are not
appropriate for a given site, use of other values  is permitted.  In
particular. States and Tribes can use site-specific data for the food web and

-------
                        Section VELA: Site-Specific Modifications                     233


accompanying lipid content of the aquatic species, and concentrations of the
dissolved organic carbon and particulate organic carbon.  If one input
parameter is modified, site-specific values must be used for all input
parameters.  Selective modification of the FCM is not allowed because it will
not accurately represent the characteristics of a specific site.

      For the final Guidance, a distinct single BAF for each trophic level of
concern for derivation of human health criteria and a separate distinct single
BAP for each trophic level of concern for derivation of wildlife criteria were
calculated using the concentrations of POC and DOC in Lake Superior as
reasonable worst-case conditions.  As noted in the above responses and in the
methodology set forth in procedure 1 of appendix F to part 132, and
subsequently in appendix B to part 132,  EPA is allowing for modifications to
the BAF based on site-specific characteristics.

      c.    Final Guidance:  EPA is allowing site-specific modifications to
the BAF based on the procedure set forth in procedure 1 of appendix F to part
132.

5.    Human Health

      a.    Proposal:  The proposed Guidance stated that there is no specific
EPA guidance regarding site-specific modifications to human health water
quality criteria.  (The 1983 Handbook does not offer any guidance in the area
of human health site-specific criteria development.)  Consistent with the
Steering Committee proposal, the proposed Guidance restricted site-specific
modification to human health criteria/values to only those which would
increase the level of protection for humans. The proposed Guidance also stated
that human health criteria or values shall be modified on a site-specific
basis to provide additional protection appropriate for highly exposed
subpopulations.

        EPA invited comments on whether the proposed approach for humans and
wildlife was reasonable or whether less stringent site-specific modifications
could be allowed under certain circumstances.

      b.    Comments;  EPA received numerous comments that States should be
allowed to make a demonstration for less stringent human health criteria based
on site-specific  conditions, if scientifically supported.  The majority of
commenters suggested that two exposure components, fish consumption and fish
lipid content should be evaluated on a site-specific basis.  These same
commenters believed that there could be parts of the Great Lakes System
(especially tributaries) where one could demonstrate that the fish consumption
rate and/or fish lipid concentration are significantly different (presumably
lower) from those in the proposed Guidance.   Other commenters argued that the
body weight of sensitive subpopulations (women of childbearing age and
children)  should be accounted for in developing human health criteria on a
site-specific basis.   Commenters also believed that site-specific criteria
should be generated only if the resulting criterion were protective of
potential uses at the site and all downstream uses.  Commenters stated that
EPA did not provide scientific justification for allowing only more stringent
site-specific human health criteria.

      A few commenters were opposed to the idea of less stringent site-
specific human health criteria.  Commenters stated that less stringent
criteria should not be allowed for the open waters of the Great Lakes, since
this would result in inconsistency throughout the basin, but was in favor of
allowing less stringent modifications for tributaries (waters other than the
open waters of the Great lakes).  Other commenters believed that EPA should
not allow site-specific modifications for fish consumption of subsistence and
recreational angler;  but should select a fish consumption rate protective of
sport anglers and other highly exposed subgroups.  Still other commenters
pointed to the impractical nature of conducting meaningful site-specific
studies because the Great Lakes Basin is large with free movements of

-------
234    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

fisheries and distribution of pollutant loadings, believing the practice would
only result in States imposing different health advisories and ultimately non-
attainment of water uses.

      EPA agrees with many of the comments, and has changed the final Guidance
to allow for demonstrations for less stringent site-specific human health
criteria as related,to fish consumption rate and BAFs.  Other exposure
parameters such as drinking water consumption rate, body weight and incidental
ingestion rate can fluctuate on a population basis, but are not considered
parameters likely to change on a site-specific basis.  For example, the mean
body weight of a State may be greater or less than the 70 kilograms assumed in
the criteria methodology,  but making a determination that a specific site has
a higher or lower mean body weight may be implausible and impractical.  To
make a showing for a less stringent criterion, a State or Tribe would have to
demonstrate that there was a site-related population of people weighing more
than 70 kilograms.  In addition, EPA has provided guidance on choosing
protective body weights for those parts of the population considered more
vulnerable to the toxicological effects of environmental contaminants.  With
regard to toxicological assessments, EPA does not believe there are likely
conditions under which a site-specific toxicological assessment can be made.
For example, to make a site-specific toxicological assessment, it would have
to be shown that a particular human population at a specific site was more or
less sensitive to an environmental contaminant due to genetic predisposition
or site-related conditions which mitigate or enhance the toxicological effects
of a specific chemical contaminant.   Again, EPA believes making such a
showing is highly unlikely and probably implausible.

      i.    Fish Consumption Rate:  EPA believes, on the other hand, that it
is conceivable that there might be isolated tributaries and populations of
people who do not consume as much fish as the rate presented in the proposal.
Such a demonstration would likely be difficult to make, due to the transience
of people in the Great Lake area, but if a State or Tribe can demonstrate
based on data that a group of people who are the exclusive users of a
waterbody has a significantly lower fish consumption rate than the rest of the
Great Lakes population, they may apply that lower rate in developing their
human health criteria for that waterbody.  The States and Tribes must ensure
that fish migration from the waterbody in question will not lead to increased
exposure to other human populations.  The State or Tribe must also demonstrate
that the specified waterbody is not associated with a known or anticipated
group of individuals who may consume more fish, such as a sport or subsistence
angler population.  To ascertain such information, the State or Tribe must
conduct a site-specific fish consumption survey.  When determining a site-
specific fish consumption rate, a site-specific fish lipid percentage must
also be determined.

      In response to commenters who stated  EPA should not allow for site-
specific criteria adjustments due to the difficulty in conducting meaningful
site-specific surveys,  EPA does not believe it is fair to disallow such a
demonstration if a State or locality believes there is a basis for such a
showing.  While it is true that a well conducted fish consumption survey is
costly and sometimes open to various interpretations, the data from such a
survey may be more useful in determining subpopulation consumption trends than
a default value or a generalized  assumption for an entire population of
people.

      ii.   Percent Fish Lipid:   EPA has concluded that the percentage  of fish
lipid can be adjusted on a site-specific basis.  This can be  done  in
conjunction with a fish consumption survey  or localized monitoring data on
fish species of the area.  From a site-specific  fish  consumption  survey, the
predominant fish  species consumed,  the percentage  of  each  species  consumed,
and the lipid concentration  of  those species can be  determined.   If the
recalculated weighted mean percent  lipid is significantly  different than  that
in the final Guidance,  it may be  used  in calculating site-specific human
health criteria.

-------
                        Section Vffl.A: Site-Specific Modifications                     235

      iii.  BAF:  EPA has concluded that the BAF can be adjusted on a site-
specific basis.   (See section A.4 of this document on BAFs to determine how a
site-specific BAF can be derived.)

      In response to the commenters who stated site-specific criteria should
be limited to tributaries only, EPA concludes that a showing can be made for
any part of the Great Lakes System as long as it is scientifically justified.
Site-specific conditions must be the basis for the criteria adjustment '
regardless of geographical location in the Great Lakes System.

      With regard to the Guidance which states that human health criteria or
values shall be modified on a site-specific basis to provide additional
protection appropriate for highly exposed subpopulations, EPA believes the
best way to determine whether a highly exposed subpopulation exists is through
the use of fish consumption surveys targeted toward waterbodies and sites
which are known (qualitatively, if not quantitatively) for high levels of
sport and subsistence angler usage.  Once a site/waterbody is established as a
high-use site, a fish consumption survey should be conducted to determine
average and high level (90th or 95th percentile consumption rates) consumption
rates for the sport and subsistence angler subpopulations.  Once it is
established that a highly exposed subpopulation does exist at a distinct site,
the criteria or values must be modified using the fish consumption rate, the
percent lipid, and the resulting BAFs associated with the higher consumption
rates attributed to the subpopulation in question.   For information on how to
conduct a fish consumption survey and how to analyze the results of such a
survey, refer to the following EPA document: Consumption Surveys for Fish and
Shellfish.  A Review and Analysis of Survey Methods.  Feb. 1992.  EPA 822/R-
92-001.

      An alternative to conducting site-specific fish consumption surveys is
to use default assumptions for fish consumption by recreational and
subsistence anglers.  Several commenters recommended default assumptions for
recreational and subsistence fish consumption.  For recreational fish
consumption, the commenters recommended using a range of 45-150 grams/day in
developing water quality criteria protective of sport anglers.  For
subsistence fishing, the commenters recommended using a range of 90-165
grams/day in developing water quality criteria protective of subsistence
anglers.  EPA believes the use of default assumptions are acceptable.  The
actual default values used, however, will depend on the circumstances in that
State or Tribal area.  For more details on the commenters1 recommendations
refer to the Environmental Defense Fund/ Penobscot Indian Nation/NAACP Legal
Department document entitled: The Protection of Sport and Subsistence Fishing
Populations in the United States: Recommendations to the Administrator, EPA,
For Implementation of the President's Executive Order on Environmental Justice
and the Subsistence Consumption of Fish and Wildlife, June 1994 which is
available in the docket for this rulemaking.

      c.    Final Guidance;  The final Guidance allows human health criteria
or values to be modified on a site-specific basis based on differences in fish
consumption, or BAF where the resulting criteria may be less stringent than
the final criteria listed in Tables 3, for the reasons stated above.

-------
                              Section Vffl.B: Variances                           237
B.    Variances from Water Quality Standards for Point Sources

      The final Great Lakes Water Quality Guidance allows Great Lakes States
and Tribes to include water quality standards variance provisions  in their
water quality standards, and grant variances based on those provisions,  as
long as they are consistent with procedure 2 of appendix F.  These water
quality standards variance procedures provide a mechanism for States and
Tribes to maintain the basic standards  (uses and criteria) as goals and  assure
compliance with sections 301(b)(1)(C) and 402(a)(1) of the CWA that require
NPDES permits meet applicable water quality standards, while granting
temporary relief to point source dischargers under appropriate circumstances.
This Guidance does not require the States or Tribes to include a variance
provision as part of their standards program.

      The intent of the variance provision is to: provide a mechanism by which
permits can be written to meet a modified standard where compliance with the
underlying water quality standard is demonstrated to be infeasible; encourage
maintenance of original standards as goals rather than removing uses that may
be ultimately attainable; identify conditions under which variances may  be
granted; identify the requirements for variance applications; and  ensure the
highest level of water quality achievable while the variance is in effect.

      The final variance procedures included in procedure 2 of appendix  F
differ little from the proposal and provide for consistent application of
water quality standards variances for Great Lakes States and Tribes.
Variances can be requested for any of the same reasons which justify removing
designated uses as described at 40 CFR 131.10(g).

      In the final Guidance, States and Tribes retain the discretion to  define
what specific information they will require in a permittee's variance
demonstration and application.  States and Tribes also have the discretion to
define the decision criteria to use when approving or disapproving a variance,
as long as they are at least as stringent as the requirements in procedure 2
of appendix F and subject to EPA review and approval.

      A State or Tribe choosing to adopt variance procedures as part of  its
part 132 submission*will provide information on the requirements for the
variance demonstration and application as well as the evaluation criteria that
the State or Tribe will use to approve or disapprove specific variances.  A
variance procedure should assure that: the public has sufficient information
to comment on the appropriateness of a State's or Tribe's WQS variance
process; EPA has sufficient details to determine if the State or Tribe
procedures comply with the CWA and are approvable; and both EPA and the  public
have adequate information on which to judge State or Tribal compliance with
its own procedures when making individual variance decisions.  The final
guidance does not require States or Tribes to grant variances in any specific
circumstance.

      In addition to reviewing individual variances granted by a State or
Tribe,  EPA will review the State or Tribal variance procedure itself as  part
of its review of the State or Tribe's GLI submission.

1.    Appljcabi1itv

      a.    Proposal:  The proposed Guidance limited the availability of a
variance to the permittee requesting the variance and only for the
pollutant(s)  specified.  The water quality standards for the affected water
body would not otherwise be changed by a variance.  The proposal also did not
allow variances for new or recommencing dischargers as those terms are defined
at 40 CFR 122.2.

      In procedure 2.C of the proposal, EPA also did not allow a variance to
be granted if standards would be attained by implementing effluent limitations
required under sections 301(b)  and 306 of the CWA and by the permittee

-------
238    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

implementing cost-effective and reasonable best management practices.  (In the
final Guidance, this requirement has been moved to the Applicability section
for clarity.)

      b.    Discussion of Significant Comments

      Comment:  Several commenters suggested that EPA also allow basin-wide or
waterbody variances often citing the need for relief from "ubiquitous"
pollutants.  Other commenters objected to allowing waterbody variances.

      Response:  EPA considered these comments and has decided not to
incorporate waterbody variances into the final guidance for the following
reasons:  (i) EPA has not developed methodologies for applying waterbody
variances and has not allowed this practice nationally,  (ii) EPA is concerned
that waterbody variances could provide inappropriate relief to nonpoint
sources and cause ari additional regulatory burden to point source dischargers,
and  (iii) EPA believes that the appropriate long term solution to the problem
of "ubiquitous" pollutants is the total maximum daily load  (TMDL) process
which can account for all sources of pollutant loading both point and
nonpoint, and that the appropriate short term relief consists of discharger
specific variances and determinations made pursuant to the Reasonable
Potential portion of the final guidance.  See sections VIII.C. and VIII.E. of
this document for a discussion of TMDLs, Reasonable Potential and "ubiquitous"
pollutants.  In order to provide the most efficient short term relief to
"ubiquitous" pollutants, EPA encourages States to consider multiple discharger
specific variance requests on a watershed basis where appropriate.    Multiple
discharger variances are similar to, but not the same as a waterbody variance.
Where all point source discharges on a waterbody request a variance, both
approaches will have the same effect on the point sources.  The difference is
that a waterbody variance applies to all nonpoint sources of the pollutant as
well, whereas a multiple discharger variance does not.

      Comment:  Some commenters stated that variances should be available to
dischargers who applied for NPDES permits on or before the effective date of
the rule.

      Response;  EPA agrees.  The new definition of "new discharger" at §132.2
developed for purposes of compliance schedules and mixing zones will
accomplish this.

      Comment:  Several commenters objected to excluding new or recommencing
dischargers for consideration for WQS variances, citing  the presence of
"ubiquitous" pollutants in the Great Lakes  System.  Other commenters agreed
with the proposal to exclude new and recommencing discharges from variances.

      Response:  As was stated in the preamble to the proposed Guidance, and
reiterated above, WQS variances are used to allow existing dischargers to
comply with a  modified standard when compliance with the underlying  standard
is infeasible.  Variances are not intended  to allow water quality that is
already below  standards to be further degraded, which would be the case if new
or recommencing dischargers  add increased loads of a pollutant to a  waterbody.
This concept is reinforced in the proposed  and final Guidance at procedure
2.F.I which requires that the NPDES permit  limitation be no less stringent
than that achieved under the previous permit.  Since new or recommencing
dischargers have no previous load, they could not add  any load under the
variance procedure's terms.  If new, recommencing or existing dischargers
desires  relief from  requirements to control "ubiquitous" pollutants and will
not add  increased loads of the pollutant to the waterbody,  the appropriate
mechanisms  are found in the  TMDL  procedure  3 and the Reasonable  Potential
procedure  5  in appendix F.

      Comment:   Some commenters requested that EPA more clearly define
"recommencing  discharger," and  specify  "...that facilities  that  cease

-------
                              Section Vm.B:  Variances                           239
discharging temporarily, but retain their NPDES permits, are not recommencing
dischargers."

      Response:  EPA disagrees that further definition of "recommencing
discharger" is needed.  40 CFR 122.2 defines the term as "...a source that
recommences discharge after terminating operations."  EPA interprets
terminating operations to mean eligible for NPDES permit revocation.  Hence,
sources that retain'their NPDES permits would not be considered "recommencing
dischargers."

      Comment;  Several commenters suggested that EPA should eliminate the
requirement of Best Management Practices as a condition for obtaining a
variance.  Other commenters stated that the BMP requirement should be
clarified, or that BMPs should be limited to those that may be implemented by
a particular discharger on a reasonable and cost-effective basis.

      Response:  EPA disagrees that the BMP requirement should be eliminated.
EPA agrees, however, that the BMPs that must be implemented before a variance
may be granted should be limited to those that may be implemented by a
particular discharger.  WQS variances are not intended to allow water quality
that is already below standards to be further degraded.  In addition, as
stated in procedure 2.F.I, the purpose of variances are to improve water
quality as much as possible by requiring effluent limitations that represent
the level of water quality achievable by the permittee.  If the permittee can
implement cost effective and reasonable BMPs for nonpoint sources, over which
it has control, that will attain water quality standards, the permittee should
implement those BMPs rather than requesting a variance for its point source
discharge.  If implementing such BMPs will improve water quality but not meet
the standards, implementation by the permittee will result in a reduced
variance request and an overall improvement in water quality.

      c.    Final Guidance:  For the reasons stated above, EPA has retained
paragraph A, the Applicability section of procedure F.2, in the final variance
procedures except:  (i) the term "new discharger" has been changed to "new
Great Lakes discharger" and defined at §132.2, (ii) a new paragraph 2.A.2 has
been added to assure compliance with section 7 of the Endangered Species Act,
and (iii) the final sentence from procedure 2.C of the proposal has been moved
and re-numbered as procedure 2.A.3.

2.    Maximum Timeframe for Variances

      a.    Proposal:  Procedure 2.B of the proposal, specified a variance
term not to exceed three years.  EPA proposed the three-year term to reinforce
the triennial review required of all water quality standards in accordance
with section 303(c) of the Clean Water Act.

      b.    Discussion of Significant Comments

      Comment:  This section received a large number of comments.  Some
commenters favored retaining the three-year requirement but the majority
favored extending the life of a variance to five years or the term of the
NPDES permit for which it is granted.  Several commenters stated that a three
year variance term would present a substantial administrative burden on both
the discharger and the permitting authority,  especially for situations where
a variance renewal is appropriate.

      Response:  EPA's 1979 Guidance (see the preamble to the proposed
Guidance (58 FR 20921) for discussion of EPA's 1979 Guidance) clearly
indicated that variances are granted for a specific time period and must be
reviewed every three years, but does not suggest that variances must expire at
the same three-year interval.

      c.    Final Guidance:  Based on the comments and the rationale above,
EPA has extended the allowable term for a water quality standards variance to

-------
240    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

five years or the term of the NPDES permit, whichever is less.  This reduces
the administrative burden of a facility in preparing a variance request since
the facility can prepare the request at the same time it prepares the NPDES
permit application.  To ensure that variances are indeed reviewed trienriially
with the rest of the State or Tribal standards, and modified as appropriate,
EPA has added that requirement to procedure 2.B.  In addition, a new
requirement has been added at procedure 2.F.4 to provide for reopening and
modifying the NPDES'permit as appropriate based on changes made during the
water quality standards review, e.g. if the variance is terminated or
modified.

3.     Conditions to Grant a Variance

      a.    Proposal:  Variances under the proposed Great Lakes Guidance were
applicable if any of five specified types of waterbody conditions exist and/or
the affected community would encounter substantial and widespread economic and
social impacts as a result of the point source having to install controls,
beyond technology-based requirements, necessary to meet the WQS.

      The permittee was required to make two other demonstrations.  The first
demonstration was that the requested variance is consistent with State or
Tribal antidegradation procedures.  In the second demonstration, the applicant
was required to characterize the extent of any increased risk to human health
and the environment associated with granting the variance compared to the
original water quality standards, and the State or Tribe was required to find
that any such increased risk is consistent with the protection of the public
health, safety and welfare before granting the variance.  Because variances
are from water quality standards that meet the goals and requirements of the
Clean Water Act, this language was intended to ensure that the general
requirement of section 303(c)(2)(A) of the CWA be met even though specific
protective criteria may be temporarily exceeded.

      Consistent with other approaches for regulatory relief, the permittee
was responsible for providing sufficient relevant information, pursuant to
State or Tribal requirements, to make a variance demonstration for the
pollutant(s) in question.  Failure of the permittee to make an adequate
demonstration or to provide sufficient information was sufficient for a State
or tribal denial of the variance.

      b.    Discussion of Significant Comments

      Comment:  Some commenters suggested adding additional reasons for
granting a variance to those proposed at procedure 2.C including technological
infeasibility and background pollutants.

      Response;  Based on current EPA guidance  (discussed in the preamble to
the proposed rule at 58 FR 20921}, the justifications for a WQS variance are
the six reasons for removing a designated use found at 40 CFR 131.10(g).
Because the final Guidance is to be no less  restrictive than national guidance
 (CWA §118(c) (2) (A))> EPA has decided that it would be inappropriate to provide
additional justifications for variances at  this time.  However, it is EPA's
intent to provide the Great Lakes States  and Tribes reasonable discretion in
defining, characterizing and developing decision criteria for the conditions
for granting a variance, subject  to EPA review and approval.  It is EPA's
position that by maintaining this discretion, States and Tribes will be able
to address specific concerns in  the State/Tribal procedures developed pursuant
to procedure 2.  Flexibility for  addressing background pollutants is provided
for in the Reasonable Potential  section of  the Guidance.

      Comment:  Some commenters were confused about the last  two requirements
in proposed procedure 2.C  regarding:  (i)  the demonstration that
antidegradation requirements had been met,  and  (ii) the demonstration  of  the
extent of increased risk to human health  and the environment.

-------
                              Section Vm.B:  Variances                           241
      Response:  EPA agrees that the proposal was confusing and has clarified
this in the final Guidance.  The term "demonstrates" as used in the proposal
was not intended to be synonymous with the "antidegradation demonstration"
required in appendix E.  The "demonstration" in procedure 2 was intended to be
a showing that either water quality would not be lowered in the waterbody  (the
requirement at procedure 2.F.I requiring dischargers to maintain the level of
treatment achieved under the previous permit would almost always prevent a
discharger from being granted a variances that would result in an actual
lowering of water quality) or, if water quality would be lowered, that the
antidegradation requirements of appendix E were met.  For these reasons, EPA
anticipates that an "antidegradation demonstration" pursuant to appendix E
would rarely be required when granting a variance.  To avoid further confusion
in this section, the term  "demonstrates" has been changed to "shows" in the
final guidance.

      In addition, the two requirements have been re-formatted at procedure
2.C.2 to indicate that they are indeed separate requirements, as was the
intent of the proposal, and are not associated with the social/economic
justification at procedure 2.C.l.f.

      Comment:  Some commenters thought it important to integrate the variance
procedures with the antidegradation procedures while others objected to this
requirement.  Many of those objecting were concerned with the amount of effort
required to conduct an antidegradation demonstration for every variance
request.  Commenters also stated that this language was necessary to ensure
that granting a variance does not have a downstream impact on high quality
waters.

      Response:  EPA believes that many of the objections to the
antidegradation requirement stem from the confusing terminology discussed in
the previous comment/response.  The terminology change in the final guidance,
as discussed above, addresses these objections.

      This antidegradation requirement in the variance procedures was intended
to prevent a variance that would result in a lowering of actual water quality
for any pollutant where water quality for that pollutant does not support
either the designated or existing uses or in any water constituting an
outstanding national resource (ONRW) at part 132, appendix E, section I.C as
well as to prevent dischargers from avoiding the proposed requirements of part
132, appendix E, section I.B in high quality waters by requesting a variance
rather than conducting an antidegradation demonstration.  The requirement at
procedure 2.F.I requiring dischargers to maintain the level of treatment
achieved under the previous permit is expected to prevent a discharger from
being granted a variance that would result in a significant lowering of water
quality, and an appendix E antidegradation demonstration would not normally be
necessary.  An appendix E antidegradation demonstration would only be
necessary if required by the language in appendix E.  The antidegradation
showing here would most often demonstrate to the State and public that a
concurrent antidegradation question is not at issue or that, if one is, the
regulatory provisions for antidegradation are being met.

      Comment:  Comments were also received on several of the specific
conditions for granting a variance at procedure 2.C.  Some of these comments
requested more specific guidance on some of the conditions.

      Response;  It is EPA's intent to provide the Great Lakes States and
Tribes reasonable discretion in defining, characterizing and developing
decision criteria for the conditions for granting a variance, subject to EPA
review and approval.  It is EPA's position that by maintaining this
discretion,  States and Tribes will be able to address specific concerns in the
State or Tribal procedures developed pursuant to procedure 2.

      c.    Final Guidance:  EPA has maintained procedure 2.C, including the
requirements on antidegradation and on demonstrating that there will be no

-------
242    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

increased risk to human health and the environment from granting the variance,
essentially unchanged in the final Guidance.  EPA has made a terminology
change, discussed above,  to clarify the intent that an antidegradation
demonstration pursuant to appendix E will apply only if the variance results
in significant lowering of water quality in a high quality waterbody, a
circumstance which is not expected to occur under procedure 2.  In addition,
the two "showing" requirements have been re-formatted at procedure 2.C.2 to
indicate that they are separate requirements, and not associated with the
social/economic justification at procedure 2.C.l.f.

4.     Submittal of Variance Application

      a.    Proposal:  The proposed Guidance required permittees to submit a
variance application to the State or Tribe no later than 60 days after the
regulatory authority reissued or modified the permit.  The application was
required to provide information necessary to evaluate whether  the conditions
for granting a variance,  as set forth in procedure 2.C of the proposal, were
met.

      b.    Discussion of Significant Comments

      Comment:  Some commenters stated that the 60 day time frame was too
short and/or that the variance application should be allowed at other times
 (e.g., during the permit application process or any time during the permit
cycle).  Several commenters pointed out that this language would require all
permits for which a variance is granted to be reopened and modified and
suggested that processing a variance request during the permit application
process would substantially reduce the administrative burden on the discharger
as well as the permitting authority.  Many similar comments were received
regarding the timeframe for application for a variance renewal at procedure
2.H.

      Response:  EPA agrees and has modified procedure 2.D to allow States and
Tribes to determine the most appropriate variance application schedule to suit
their administrative procedures for both first-time variances at procedure 2.D
and for variance renewal at procedure 2.H.

      c.    Final Guidance;  In the final guidance EPA has removed both the 60
day limit and the requirement that a variance be applied for after a final
NPDES permit is issued.  EPA expects States and Tribes to specify in their
variance procedures the timing that best suits their own regulations and
permitting procedures.



      a.    Proposal:  The proposed Guidance provided for public notice and
opportunity to comment on a variance request  (procedure 2.E) and on the draft
modified NPDES permit  (procedure  2.G.).  In addition, the requirement at
procedure 2.J that variances be appended to State water quality standards
rules ensured that the public is made aware of which variances have been
granted.

      b.    Discussion of Significant Comments

      Comment;  Commenter stated  the proposed public notice requirements  for
variance procedures^were not adequate to allow full public involvement.

      Response;  EPA  does not agree.  Procedure 2  does not modify the  normal
public participation  requirements for adopting WQS, but rather adds  a
requirement that the  notice also  present the State or Tribe's preliminary
decision.

      The following is a summary  of the elements that EPA expects to be made
available to the public in order  to meet the public notification requirements

-------
                              Section Vm.B:  Variances                           243
of the water quality standards regulation  (40 CFR I3l.20(b)).   if these items
are not included in detail in the public notice, they must be in the public
record and the public must be made aware of their existence and of how and
where they may be obtained.

      (i)   A statement that the action complies with the State's or Tribe's
variance procedures- and description of those procedures;

      (ii)  The permittee's demonstration, including the rationale for the
requested variance, and the extent of any increased risk to human health and
the environment associated with granting the variance; and

      (iii) The public notice for any draft NPDES permit, the public comments
and public hearing records pursuant to procedure 2.E, and the State approval.
This public notice can be combined with the public notice for a draft NPDES
permit,  as long as the variance is identified and all the necessary
information pertaining to the variance is included.

      Comment;  Some commenters stated that a timeframe should be required for
the public notice and/or comment period.

      Response:  EPA does not agree that timeframes for these activities need
to be specified in the Guidance because the various State and Tribal
administrative procedures are different.  EPA does, however, expect the States
and Tribes to include appropriate timeframes in their variance procedures,
pursuant to their administrative procedures, to provide meaningful opportunity
for public comment as well as orderly and timely State or Tribal action.

      c.    Final Guidance:  For the reasons stated above, EPA has retained
the language contained in the proposal, with the clarification that the public
notice can be combined with the public notice for a draft NPDES permit, as
long as the variance is identified and all the necessary information
pertaining to the variance is included.

6.    Final Decision on Variance Request

      a.    Proposal:  The proposed Guidance required the State or Tribe to
issue a final decision on a variance request within 90 days of the expiration
of the public comment period.  The proposed Guidance also required that this
decision specify all NPDES permit conditions needed to implement the variance.
These conditions were to assure that:  the permittee minimizes the water
quality standards exceedance by implementing the level of treatment currently
achievable (i.e., conditions requiring effluent limitations at least as
stringent as those achieved under the previous permit); the permittee makes
reasonable progress toward attaining the water quality standards; and effluent
limits sufficient to protect water quality standards are in effect upon
expiration of the variance.  States/Tribes were to deny a requested variance
if the permittee failed to make the demonstrations required under section C of
the proposed procedure.

      b.    Discussion of Significant Comments

      Comment;  Some commenters objected to the permit condition requiring
"reasonable progress" be made toward attaining water quality standards,
stating for example, that in most cases, the permittee receiving a variance
will not contribute to the conditions that prevent the receiving waters from
attaining water quality standards and therefore will be powerless to make
"reasonable progress" toward attaining water quality standards for the
waterbody.

      Response:  EPA does not agree that a discharger is always "powerless" to
affect pollution sources outside its immediate contribution.  Dischargers can,
for example,  participate in total maximum daily load  (TMDL) development and/or
engage in pollution trading schemes to reduce load from outside its facility.

-------
244    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      Comment:  Some commenters were unclear about what EPA intends by
"reasonable progress."

      EPA agrees that "reasonable progress" is imprecise, but believes this is
necessary to provide flexibility to the State or Tribe to require activities
appropriate to the State or Tribal programmatic objectives and the individual
situation subject to EPA review and approval.  In addition to progress
reducing loads outside its immediate contribution discussed above, reasonable
progress can be made in improving the quality of the discharge through
conditions such as: (i)  the establishment of a capital improvements fund; and
(ii)  continued investigations of treatment technologies, process changes,
pollution prevention,  wastewater reuse and/or other techniques that will
reduce the level of the pollutant or result in compliance by the permittee
with the WQS and submission of reports on the investigations at such time
specified by the State.   It would be difficult for EPA to define what
"reasonable progress" is without considering site-specific information on the
facility, pollutant and receiving water for a given situation.  In instances
where a discharger is indeed "powerless" to affect any change in conditions,
the State or Tribe may interpret existing practices as "reasonable."

      Comment;  Procedure 2.F.3 proposed requiring, upon expiration of a
variance, compliance with the effluent limitation in effect prior to the
granting of the variance.  Commenters thought this requirement confusing.

      Response:  Because the final Guidance no longer requires that a final
NPDES permit be issued prior to a variance application, there may well be no
pre-existing effluent limitation implementing the underlying WQS with which to
comply.  EPA has, therefore, modified this requirement to provide that
effluent limitations be in compliance with applicable water quality standards
upon expiration of the variance.

      c.    Final Guidance:  For the reasons stated above, EPA has retained
procedure 2.F essentially unchanged.  Because of the changes to procedure 2.B
which extend the allowable term for a variance beyond the 3 years in the
proposal  (see discussion at B.2.c, above) new language has been added at
procedure 2.F.4 requiring a permit condition that allows the permitting
authority to reopen and modify a permit which is based on a WQS variance if
that variance has been rescinded or modified pursuant to a State or Tribal
triennial standards revision.

7.    Incorporating'Variances into NPDES Permits

      a.    Proposal:  Once a variance is granted, the proposed Guidance
required the State or Tribe to modify the NPDES permit to incorporate all
NPDES permit conditions determined to be necessary to implement the variance.

      b.    Discussion of Significant Comments

      Comment:  Some commenters stated that the Guidance should not routinely
require that a permit be modified in order to implement a variance.

      Response:  EPA agrees.  EPA has removed the requirement in procedure  2.D
that a variance may be applied for only after a final NPDES permit is issued.
By allowing variance applications any time during the permit renewal process,
EPA anticipates that most permits that receive variances will have the
provisions of the variance incorporated into the permit when it is reissued.

      c.    Final Guidance:  EPA has removed the requirement in procedure 2.D
that a variance may be  applied for only after a final NPDES permit is issued.
The final Guidance allows  the permitting authority to incorporate the
provisions of  the variance into the permit at a time, and in a manner which is
appropriate, given the  circumstances  of the variance.

-------
                              Section Vm.B: Variances                           245
8.    Renewal of Variances

      a.    Proposal:  The proposed Guidance required a permittee  to  re-apply
for a variance no later than the submission of a permit application or 60  days
prior to the expiration of the variance.  The permittee was required  to make a
new showing of justification; variances would not be automatically renewed.
As part of the renewal application, the permittee would be required to
demonstrate that it has met the NPDES permit conditions implementing  the
existing variance.  The same public notice requirements for the  initial
issuance of a variance applied to the renewal.  Permittees not demonstrating
compliance with these conditions would not be eligible for a variance renewal.

      b.    Discussion of Significant Comments

      Comment^  Several commenters stated that an application for  variance
renewal should not be required until the permit renewal becomes  final.  Others
stated that the variance reapplication process should occur in conjunction
with the permit renewal process.

      Response:  EPA agrees that such flexibility is reasonable  and has
modified procedure 2.H to allow States and Tribes to determine the most
appropriate variance re-application schedule to suit their administrative
procedures.

      Comment;  Some commenters stated that discharges that violate conditions
of a variance should not necessarily be ineligible for a renewed variance
pointing out that EPA and the states already have ample authority  to  impose
enforcement sanctions on dischargers violating their permit conditions,
including variance conditions, and that EPA need not require states to impose
a second penalty on those dischargers.  Other commenters also stated  that EPA
proposes to treat all violations of variance conditions the same,  and
therefore, even de minimis violations would result in the inability to renew a
variance.

      Response;    The permit conditions are intended to implement the basic
principles of water.quality standards variances that (i) the non-attainment of
standards be minimized, (ii) progress toward attaining standards be achieved
where possible, and  (iii)  the discharger must meet the standard upon
expiration of the variance  (see discussion of current EPA policy in the
preamble to the proposed guidance at 58 PR 20921.)  Dischargers  that  fail to
meet permit conditions designed to achieve these basic principles  should not
be allowed to continue the practice through a variance renewal.  EPA  intends,
however, that States and Tribes should have discretion in determining whether
a particular non-compliance with variance conditions in NPDES permit  would
warrant denial of a variance renewal request, subject to EPA review and
approval.

      c.    Final Guidance;  Based on the rationale above, EPA has retained
procedure 2.H but has modified it to allow States and Tribes to determine the
most appropriate variance re-application schedule to suit their administrative
procedures.   EPA has also modified the final Guidance consistent with 40 CFR
122.64 to allow the permitting authority to deny a variance renewal where the
permittee has not fully complied with the permit conditions applicable to the
variance.

9.     EPA Approval
                   *
      a.    Proposal;  The proposed guidance listed the information to be
submitted to EPA for review and approval, including the permit application,
public comments.  State or Tribal approval, and NPDES permit; and gave
timeframes for those submittals.

-------
246    Water Quality Guidance for the Great Lakes System - Supplementary Information Document

      b.    Discussion of Significant Comments

      Commentj  Some commenters stated that a time requirement  for EPA review
of WQS variances was needed.

      Response:  EPA does not agree that it is necessary  for  this  Guidance to
specify an EPA deadline for review of variances.  Variances are modifications
of State or Tribal water quality standards and are, therefore,  subject to EPA
review and approval.  Like other water quality standards  changes,  variances
are effective when adopted  (under the terms of the adoption), whether or not
EPA review is complete.  In addition, there are already statutory  and
regulatory timeframes, at §303(c)(2)(B) of the CWA and  40 CFR 131.21(a)
respectively, that govern EPA approval/disapproval decisions  on all State or
Tribal water quality standards revisions including variances.

      c.    Final Guidance:  Based on the rationale above, EPA  has retained
procedure 2.1 as proposed.  Editorial changes have been made  for clarity.

10.   State or Tribal Water Quality Standards Revisions

      a.    Proposal;  The proposed Guidance  required the State or Tribe to
append the State- or Tribal-adopted variances to the State or Tribe's water
quality standards.

      b.    Discussion of Significant Comments

      Comment:  EPA requested comment on whether a timeframe  was necessary for
the State or Tribe to append variances to their standards.  Commenters stated
that a timeframe is not necessary.

      Response;  EPA agrees and expects States and Tribes to  append variances
to their standards in as timely a fashion as  possible,  consistent  with their
administrative procedures,  so that the public is fully  informed on the
variance that have been adopted.  The appended information should  include at a
minimum:  the discharger receiving the variance, the term (beginning  and
ending dates) of the variance, the waterbody  or waterbodies affected, the
pollutant(s) affected by the variance, and the modified allowable  ambient
concentration value(s) for  those pollutants.

      c.    Final Guidance  EPA has retained  procedure  2.J as proposed.

-------
                               Section Vm.C: TMDLs                           247
C.    Total Maximum Daily Loads

1.    Background

      Section 303(d) of the Clean Water Act requires the establishment of
total maximum daily loads  (TMDLs), in accordance with priority rankings, for
waters, that are failing to meet or not expected to meet applicable water*
quality standards despite implementation of technology-based and other
existing controls.  See 40 CFR 130.7 and existing EPA guidance including
"Guidance for Water Quality-based Decisions: The TMDL Process," EPA 440/1-91-
001, April 1991.

      TMDLs quantify the maximum allowable loading of a pollutant to a water
body, and allocate this loading capacity to contributing point and nonpoint
sources (including natural background, in-place contaminants, direct wet and
dry deposition, groundwater inflow, and overland runoff) such that water
quality standards will be attained.  A TMDL must incorporate a margin of
safety  (MOS) that accounts for uncertainty about the relationship between
pollutant loads and water quality.  TMDLs may involve a single pollutant
source or multiple sources (e.g., both point sources and nonpoint sources).
Current regulations specify that TMDLs need to take into account critical
conditions for stream flow, loading, and water quality parameters  (see 40  CFR
130.7(c)(1)).  Site-specific factors are thus to be reflected in the TMDL  even
though the TMDL process may be used to ensure that water quality goals are
achieved for a waterbody segment, whole waterbody or watershed.

      Under the CWA, States and Tribes are primarily responsible for
developing TMDLs.  EPA is required to review and approve or disapprove TMDLs
developed and submitted by States and Tribes.  If EPA disapproves a State  or
Tribal TMDL, EPA must establish such TMDL {CWA section 303(d)(2)and 40 CFR
130.7(d)}.

      When applicable water quality standards cannot be attained through the
implementation of controls on point sources, within the time period specified
in the applicable standards or implementing regulations, States and Tribes may
choose to develop TMDLs using a phased approach.  The phased approach to TMDL
development is intended to achieve load reductions capable of ensuring the
attainment and maintenance of water quality standards.  EPA expects the
allocations within phased TMDLs to be based on a reasonable expectation that
water quality standards will be met in a reasonable period of time.

      The phased approach to TMDL development is an iterative process that
provides for pollution reduction while the regulatory agency collects and  uses
new monitoring data and the demonstrated performance of existing controls  to
evaluate the TMDL and revise it as necessary.  TMDLs established using the
phased approach are based on best available information, sound professional
judgment,  and a margin of safety to account for uncertainty in available data
and the anticipated relationship between controls, loading reductions and
predicted changes in water quality.  Such TMDLs require a monitoring plan, a
schedule for installation of controls, collection of monitoring data to verify
point and nonpoint source load reductions, assessment of water quality
standards attainment and additional modelling, where appropriate.  If
standards are not attained after implementation of controls recommended by the
TMDL, the data obtained through the monitoring program should be used to
revise the TMDL.
                   *
      The phased approach to TMDL development recognizes that water quality
standards cannot be attained immediately, but TMDLs developed on this basis
nevertheless must reflect reasonable assurances that water quality standards
will be attained in a reasonable period of time.  When developing a TMDL using
the phased approach,  all known sources of pollution are considered, although
specific controls on those sources may be implemented in stages.  The time
period associated with these stages of implementation ultimately determines
when water quality standards will be met for a particular waterbody.  The

-------
248    Water Quality Guidance for the Great Lakes System - Supplementary Information Document

phased approach may provide a scheduled time frame in which to implement
controls' recommended by the TMDL and achieve water quality standards and may
be particularly appropriate when addressing difficult water quality problems
in cases when data,-models and predictive tools are generally less well-
developed than for water quality problems associated primarily with the
discharge of a few point source pollutants into small watersheds.  Determining
the reasonable period of time in which water quality standards will be met is
a case-specific determination.  This determination depends upon a number of
factors, including, but not limited to, receiving water characteristics,
persistence, behavior and ubiquity of pollutants of concern, type of
remediation activities necessary, available regulatory and non-regulatory
controls and individual State requirements for the implementation of water
quality standards.

      TMDLs established using the phased approach are the preferred approach
for developing schedules of how and when water quality standards will be met
in cases when data, models, and predictive tools are not yet adequate to
address complex water quality situations characterized by persistent,
ubiquitous pollutants and water quality impacts resulting from nonpoint
sources of pollution.  EPA believes that it is reasonable and appropriate in
these circumstances to establish TMDLs which schedule implementation
activities over a period of time.  This would result in some sources achieving
load allocations prior to other sources, provided that progress is being made
in achieving water quality standards in accordance with the schedule
established by the TMDL.  Thus, for example, EPA believes it is reasonable to
consider expected nonpoint source load reductions if they will result from the
implementation of specific voluntary or non-voluntary controls, are specific
to the pollutant of concern and the waterbody for which the TMDL is being
developed.  In some cases, for example, water quality standards may reasonably
be expected to be met within one NPDES five-year permit cycle.  In other cases
the reasonable expectation of meeting water quality standards could be  twenty
years, following the implementation of controls on nonpoint sources such as
sediment.  In still other cases, the reasonable expectation of meeting water
quality standards could be keyed to the implementation of other controls,
(e.g. air quality standards.)
                   *
      The final Guidance is not intended to comprehensively address all
aspects of TMDL development and implementation of CWA section 303(d).   Rather,
for specific matters not addressed by the final Guidance, national regulations
and guidance for the TMDL program will continue to apply to States and  Tribes
in the Great Lakes System  (see 40 CFR 130.7 and existing EPA guidance
documents such as the Technical Support Document for Water Quality-based Toxic
Control,  (TSD) EPA 505/2-90'-001, March 1991, and Guidance for Water Quality-
based Decisions:  The TMDL Process, EPA 440/1-91-001, April 1991, both
available in the docket).

      The final Guidance does not include specific provisions for deriving
nonpoint source load allocations and implementing nonpoint source controls.
While general guidance on how TMDLs should consider nonpoint source loadings
is provided, EPA regulations and technical guidance should be consulted for
more specific information.   (See, e.g., Guidance Specifying Management
Measures for Sources of Nonpoint Pollution in Coastal Waters, EPA 840-B-92-
002, January 1992, for a discussion of best management practices for nonpoint
sources; and Technical Guidance for Estimating Total Maximum Daily Loads
(TMDLs): Integrating Steady-State and Episodic Point and Nonpoint Sources,
draft June, 1994, both available in the docket).

      Procedure 3 specifies procedures for establishing total maximum daily
loads, wasteload allocations, and load allocations.  Portions of this
procedure also apply to wasteload allocations calculated in the  absence of a
TMDLs, and to preliminary wasteload allocations for the purpose  of determining
the need for Water Quality Based Effluent Limits  (WQBELs) under  procedure  5 of
appendix F.  See procedure 5.A.2 and  5.F.2 of appendix F and corresponding

-------
                               Section Vm.C: TMDLs                           249
discussion at section VIII.E.2.A and VIII.E.2.H of this document for further
information.

2.    Overview of Proposed Procedures 3A and 3B

      a.    Proposal:  The proposed guidance included two distinct approaches
for developing TMDLs: procedure 3A  (Option A) and procedure 3B  (Option B).
Both options contained the same eleven general conditions applicable to TMDL
development and special conditions regarding control of bioaccumulative
chemicals of concern  (BCCs).  Options A and B were also essentially the same
with respect to the development of TMDLs for open waters and connecting
channels of the Great Lakes System as defined at section 132.2 of the proposed
Guidance.

      The main differences between the two options existed in the development
of TMDLs for discharges to tributaries. These differences reflected the
process by which such TMDLs were developed and the degree of specificity
contained in the particular procedure. A TMDL developed for discharges to
tributaries under Option A was to be based on evaluation of the basin as a
whole, followed by site-by-site adjustments.  In contrast, Option B focused
initially on evaluating limits needed for individual point sources, with
supplemental emphasis on basin-wide considerations as necessary.  Specific
components of Options A and B are discussed in greater detail below, and in
the preamble to the proposed guidance.  Readers are encouraged to review the
preamble to the proposed guidance for more detailed information on the
proposed Options A and B  (58 PR 20928) .

      EPA sought comments on all aspects of both options, including the
overall technical and programmatic approaches set out in each option, the
consistency of each option with regard to existing national policy and program
approaches, and the degree to which each option allows for integrated
development of effective point and nonpoint source controls.  EPA requested
comments on how the options should be incorporated into the final
implementation procedure and specifically asked whether all States and Tribes
in the Great Lakes System should be required to adopt either Option A or B, or
whether States and Tribes should be allowed to choose an approach that is
consistent with one of the proposed options depending on the situation at
hand.  EPA also solicited comments on the option of not providing specific
TMDL provisions in the final Guidance and instead relying on existing TMDL
regulations and guidance.

      b.    Comments: Several commenters claimed that the proposed TMDL
procedures were confusing, fragmented and provided insufficient guidance on
how water quality-based permit limits would be calculated.  For example, many
commenters found the formulas specified in Option B confusing and some
suggested that certain components of the formulas were inaccurate or
inappropriate.

       Most commenters expressed no clear preference for either option.  Many
commenters advocated that the final Guidance allow States to choose either
procedure 3A or 3B.  Some expressed preferences for particular elements of an
option.  For example, one commenter suggested that the State or Tribe should
be given discretion to deviate from Option A's basin-wide approach and use an
area-specific approach if appropriate in a particular circumstance.

      A number of commenters suggested that, although both options had merit,
and/or limitations,, only one should be adopted to ensure consistency
throughout the Great Lakes System.  Many commenters,  including a number of
States, preferred Option B and maintained that if a single procedure is
adopted in the final Guidance,  it should be Option B.  These commenters
believed that Option B would provide greater consistency among the States than
Option A. Several commenters preferred Option B but suggested that stronger
elements from Option A should be incorporated into a revised Option B.  A
number of commenters suggested that Option A was too burdensome.

-------
250    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

      Several commenters supported the watershed-based approach reflected in
Option A.  Other commenters preferred Option A but recommended specific
modifications.  Among the recommended changes to Option A were establishing
specific requirements for mixing zones for non-BCCs.  Some commenters
suggested including,specific formulas for calculating nonpoint source
loadings.

      c.    Final Guidance:  In response to these comments, EPA simplified the
TMDL procedure in the final Guidance and clarified a number of provisions.
EPA includes only one TMDL procedure in the final Guidance in response to
concerns that the TMDL procedure promote consistency throughout the Great
Lakes System.  The final procedure 3 combines aspects of both Options A and B,
and, in response to comments, includes some of the more specific provisions of
both options A and B.  For example, in order to promote consistency among the
Great Lakes States and Tribes, EPA is retaining, with some modifications,
certain mixing zone'provisions for non-BCCs from option B.  EPA eliminated
some of the more burdensome and confusing aspects of both the proposed
options.  For example, in the final Guidance, the formulas in Option B are no
longer included.

      The final Guidance provides a greater degree of flexibility than
afforded by either proposed procedure 3A or 3B by allowing States and Tribes
to choose different implementation approaches while at the same time ensuring
a level of consistency by requiring implementation of specific components of
the procedure.  For example, the final Guidance does not specify whether a
State must adopt a basin-wide approach such as that in proposed Option A, or
an approach like proposed Option B, which would focus initially on evaluating
limits needed for individual point sources.

      The final Guidance also retains the flexibility provided in the
proposal.  For example, although the final Guidance specifies that States and
Tribes consider nonpoint source loadings, EPA has not adopted the commenters'
suggestion to specify a formula to calculate nonpoint source contributions.
Rather, States and Tribes are provided flexibility to evaluate such
contributions and to address nonpoint source contributions through existing
programs.

      The final Guidance retains the eleven general conditions and the
separate provisions for open waters of the Great Lakes System and tributaries,
with certain modifications.  Like both the proposed options 3A and 3B, the
final Guidance requires the elimination of mixing zones for BCCs; however, the
final Guidance adds a procedure to grant an exception for existing discharges
of BCCs in limited circumstances.  The general conditions of application and
specific provisions of the final procedure 3 are discussed in detail below.

      In addition, procedure 3 has been revised to include new language
(section A), which authorizes the use of certain assessment and remediation
plans in lieu of TMDLs whenever, in the final Guidance, a TMDL would be used
as the basis for a wasteload allocation.  Specifically, these assessment and
remediation plans could be used in lieu of TMDLs when deriving wasteload
allocations under procedures other than the  "baseline" procedures in procedure
5.F.2 of appendix F, when establishing mixing zones for existing discharges of
BCCs in waters not attaining water quality standards under procedure 3.C.6 of
appendix F, or as an alternative to TMDLs and the intake pollutant procedures
in procedures 5.D-E of appendix F when adjusting point source controls to
account for intake pollutants as provided in procedure 5.D.1.C of appendix F.
Thus, for example, when developing a WLA for a particular pollutant and point
source, a State or Tribe would rely upon the applicable WLA established  in an
approved TMDL or assessment and remediation plan.  If no such TMDL or
assessment and remediation plan exists, the WLA would be derived using
procedure 5.E.2.a or 5.F.2 of appendix F as appropriate.

      Under procedure 3.A of appendix F, assessment and remediation plans may
be used in lieu of TMDLs if they meet all the requirements of procedure  3,

-------
                               Section Vm.C: TMDLs                           251
satisfy the public participation requirements applicable to TMDLs, and are
approved by EPA under 40 C.F.R. § 130.6 as meeting these requirements.  Once
approved by EPA, the assessment and remediation plans will function as updates
to State or Tribal continuing planning processes, which may include, among
other things, TMDLs and areawide waste management plans under section 208.
When seeking EPA approval of these assessment and remediation plans, States
and Tribes must certify that the requirements of procedure 3 are met.
Procedure 3.A also authorizes the use of qualifying assessment and remediation
plans, such as Remedial Action Plans  (RAPs) and Lakewide Management Plans
(LaMPs), under section 118(c)(3) &  (4) of the CWA.

      The TMDL process is an important planning tool that helps identify water
quality problems and recommends solutions that link the development and
implementation of control actions to the attainment of water quality
standards.  The objective of a TMDL is to allocate allowable loads of a
particular pollutant among difference sources of that pollutant so that the
appropriate control actions can be taken and water quality standards achieved.
As discussed in section VIII.C.I above, when water quality standards cannot be
attained immediately, TMDLs may be developed under a phased approach if
appropriate.  While TMDLs are the preferred mechanism for addressing water
quality impairments, particularly where nonpoint source contributions are
significant, EPA recognizes that other mechanisms can employ the same type of
analysis and obtain the same results as formal TMDLs.  EPA also acknowledges
the comments, particularly of States, that identify comparable planning tools.
In particular, as described in section I.D.4 of this document, the States and
EPA Regional offices in the Great Lakes basin have undertaken significant
assessment and remediation planning efforts through the development of RAPs
and LaMPs.  Some States may undertake similar efforts through water quality
management plans under sections 208 of the CWA.  Accordingly, the final
Guidance specifically recognizes that assessment and remediation plans other
than TMDLs can be used with comparable water quality effect, provided that
they contain certain basic elements.  In other words, EPA expects that
assessment and remediation plans developed and approved under procedure 3.A of
appendix F can function in lieu of a TMDL for water quality decisionmaking in
the Great Lakes System because such plans, at a minimum, will assess the
sources causing or contributing to a particular water quality impairment,
identify remediation activities that are reasonably expected to result in
nonpoint source load reductions as necessary, in combination with point source
controls, to achieve water quality standards within a reasonable period of
time, incorporate a'margin of safety, and establish wasteload allocations for
point sources that are consistent with these water quality objectives.
Procedure 3.A also provides that any part of an assessment and remediation
plan that also satisfies one or more requirements under CWA section 303(d) or
implementing regulations may be incorporated by reference into a TMDL as
appropriate.  If a State or Tribe submits for EPA approval an assessment and
remediation plan under procedure 3.A that fully satisfies the requirements for
a TMDL, EPA may also approve that plan under section 303(d).

3.    General Conditions of Application

      As proposed,  Options A and B both contained the same eleven general
conditions of application for every TMDL established under the GLWQI to assure
that TMDLs employed consistent methodologies, analytical approaches and
assumptions.  Commenters overwhelmingly supported the proposal to include a
set of general conditions applicable to all aspects of TMDL development.

      Language is added in the final Guidance to clarify that the general
conditions also apply,  where indicated, to wasteload allocations (WLAs)
calculated in the absence of TMDLs and preliminary WLAs for purposes of
determining the need for WQBELs under procedure 5 of the final Guidance.

-------
252    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

a.    General Condition 1 - TMDLs Required

      i.    Proposal:  General condition 1 described the circumstances under
which a TMDL would be required upon State or Tribal adoption or EPA
promulgation of the Guidance.  In the proposal, general condition 1 specified
that, at a minimum, TMDLs were to be established for each pollutant for which
it was.determined that there is reasonable potential that a discharge will
cause or contribute to an exceedance of water quality standards as determined
pursuant to proposed procedure 5.  As proposed, such TMDLs would need to be
established in advance of the issuance of any new or revised permit for the
discharge of the pollutant, unless it was determined pursuant to the proposed
procedures that a TMDL is not needed.

      Proposed procedure 5 specified that the State or Tribe was to include a
water quality-based effluent limit in an NPDES permit whenever a pollutant is
or may be discharged into the Great Lakes System at a level that will cause,
have the reasonable,potential to cause, or contribute to an excursion above
any Tier I criterion or Tier II value.  Dnder procedure 5, as proposed. States
or Tribes would have been required to develop preliminary effluent limitations
to determine if all Tier I criteria and Tier II values would be met after
discharge, where there was data to develop such criteria or values.
Preliminary effluent limitations were to be derived from preliminary wasteload
allocations, which in turn were to be based upon and consistent with the
wasteload allocation procedures defined in the proposed procedure 3.  As
proposed, the procedure 3 requirement that a TMDL be developed whenever
reasonable potential to cause or contribute to an exceedance of water quality
standards was found, and the requirement that preliminary effluent limits
based on preliminary WLAs developed under procedure 3 be used to determine if
there is reasonable potential, were confusing and implied a circular logic.
The proposal provided, in effect, that TMDLs be developed when reasonable
potential was demonstrated, and that reasonable potential be demonstrated on
the basis of preliminary effluent limits, normally derived from TMDLs.

      ii.   Comments;  EPA received numerous comments on general condition 1.
A number of commenters were concerned with the perceived burden associated
with developing TMDLs under the proposal and, in particular, many were
concerned with the' burden associated with requiring a TMDL for a waterbody in
advance of issuing any new or revised permits.  These commenters asserted that
general condition l'would essentially prohibit any point source discharges of
a particular pollutant in the absence of a TMDL for that pollutant.  The
commenters contended that the effect of the prohibition would be to require
TMDLs on the basis of a single discharger's  "reasonable potential" to exceed
standards even in waters where TMDLs would have minimal environmental benefit,
and thus would be inefficient.  Several commenters claimed that the existing
national TMDL program requirements for identifying waters not meeting
standards, and for setting priorities to develop TMDLs, are sufficient  to
ensure  that TMDLs are developed for those waterbodies most in need.

      Numerous commenters pointed out the ambiguity between proposed procedure
3 and proposed procedure 5 relating to determination of reasonable potential.
Under proposed procedure 3, a TMDL was required when there was a finding of
reasonable potential.  However, under proposed procedure 5, a finding of
reasonable potential would be based on a  "preliminary wasteload allocation"
prepared using the procedures set forth in  the TMDL procedure.  The proposal
did not define "preliminary WLA."

      A number of  commenters suggested other circumstances that should  trigger
the requirement  to establish a TMDL.  Several  commenters  suggested that TMDLs
be required  for  pollutants that have  caused fish  consumption advisories on  the
premise that waters' subject  to fish  consumption advisories are  exceeding
narrative water  quality  criteria, if  not numeric  criteria.

      iii.   Final  Guidance;  General  condition 1  in  the final Guidance  no
longer  specifies that a  TMDL would need  to  be  developed for each pollutant  for

-------
                               Section Vm.C: TMDLs                           253
which reasonable potential is found.  Instead, TMDLs shall be established in
accordance with the waterbody listing and prioritization process outlined in
CWA section 303(d), 40 CFR 130.7 and existing EPA guidance.  Under existing
law, if existing required controls are not sufficient to attain and maintain
applicable water quality standards, the waterbody must be included on the
303(d) list, which, under the regulations, is to be submitted to EPA for
review.and approval or disapproval.  The list must include a priority ranking
of listed waters and must identify those waters targeted for TMDL development
as required by CWA section 303(d)(1)(A) and 40 CFR 130.7(b)(4).  EPA makes
this change in response to commenters1 concerns about the proposal.  First,
the final Guidance refers back to the national TMDL program rather than
creating a new trigger for TMDL development based on a finding of reasonable
potential under procedure 5.  This should minimize the confusion created by
the proposal.  The final Guidance,  by referring to existing TMDL regulations,
should also minimize concerns about the additional burden that might have
occurred under the proposal  {e.g.,  permitting subject to TMDL development).
Changes to the proposal were also made to address concerns about the use of
limited resources to develop TMDLs in waters presently attaining water quality
standards but where1 the discharge of a particular pollutant has the reasonable
potential to cause or contribute to an excursion above those water quality
standards.  While TMDLs for waters currently attaining water quality standards
are important tools to ensure that such standards are maintained, EPA
recognizes that many States and Tribes may choose to place a higher priority
on restoring impaired or threatened waters and will choose to use their
limited resources for that purpose.  This change is intended to preserve State
and Tribal discretion in establishing priorities for TMDL development and
implementation.

      In response tp comments advocating that fish consumption advisories
should trigger TMDL development EPA is developing guidance to clarify the
relationship between fish advisories and section 303(d) lists.  (Draft memo
dated July, 1994, available in the docket). EPA believes that, absent
information to the contrary, it should be presumed that fish consumption
advisories demonstrate use impairments for waters designated for the uses
specified in section 101(a) of the Clean Water Act, when defined by a State or
Tribe to include fishing.  The listing of such waterbodies on section 303(d)
lists is consistent with the purpose and intent of the Clean Water Act.

      General condition 1 also provides that, when water quality standards
cannot be attained immediately,  the TMDL must reflect reasonable assurances
that they will be achieved in a reasonable period of time.  For a more
thorough discussion of this concept, see section VIII.C.I above.

b.    General Condition 2 - Attainment of Water Quality Standards

      i.    Proposal:  In the proposal, general condition 2 discussed the load
reductions that should be achieved through TMDLs.  The first sentence of
general condition 2 supplemented the provisions of proposed general condition
1 by specifying that TMDLs would also need to be developed whenever the sum of
existing point source and nonpoint source (including natural background)
loadings of a particular pollutant exceeds the loading capacity of the water
for that particular pollutant, minus any margin of safety and minus any
capacity reserved for future growth.  As proposed, general condition 2 also
established that a TMDL for a given pollutant must implement all criteria for
that pollutant that are applicable to the waterbody in question.

      ii.   Final Guidance:  EPA did not receive significant comments on
general condition 2 as proposed.  However, EPA has reorganized proposed
general conditions 2 (Load Reductions), 3 (WLA Values) and 9  (TMDL
Allocations) in order to present the same material in a sequence that more
closely tracks the requirements of CWA section 303(d).  General condition 2 is
now entitled "Attainment of Water Quality Standards" and consists of a single
sentence on that subject drawn from proposed general condition two.  The final
guidance specifies that a TMDL must ensure attainment of water quality

-------
254    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

standards, including all numeric and narrative criteria, Tier I criteria, and
Tier II values where applicable for each pollutant or pollutants for which a
TMDL is established.  By including a specific reference to water quality
standards in addition to criteria, the final Guidance clarifies that, under
section 303(d), TMDLs must provide for the attainment of water quality
standards in their entirety, and not just their criteria components.

      The third sentence of proposed general condition 2 has been incorporated
into general condition 3 (now entitled "TMDL Allocations"), of the final
Guidance.  The final Guidance does not include the proposed language
specifying that TMDLs be prepared if the sum of existing point source and
nonpoint source loadings exceeds the loading capacity minus any specified
margin of safety for a substance.  This sentence was intended merely to
restate existing requirements under section 303(d) of the Clean Water Act and
TMDL regulations at 40 CFR 130.7, and therefore is unnecessary.  In EPA's
view, these provisions and other applicable requirements of the Guidance are
sufficient to ensure that TMDLs developed under this final Guidance will
provide for attainment of water quality standards.

c.    General Condition 3 - TMDL Allocations

      i.    Proposal:  This general condition was numbered as general
condition 9 in the proposal.  As proposed, this condition provided that
nonpoint source load allocations must be based on existing loading rates or on
anticipated increased loading rates unless a lower loading rate is expected to
occur within a reasonable period of time as a result of implementation of best
management practices or other control measures.  It also provided that the
portion of the loading capacity not assigned to nonpoint sources, or to an
MOS, or reserved for future growth is allocated to point sources.  Finally, it
stated that, upon reissuance, NPDES permits for these point sources must
include limitations' consistent with the WLAs in EPA-approved or EPA-
established TMDLs.

      ii.   Comments:  Some commenters advocated that the final Guidance only
allow the incorporation of nonpoint source reductions where such reductions
are required by legally enforceable mechanisms to ensure that reductions from
nonpoint sources are "reasonably expected to occur" within relevant time
frames.  Furthermore, the commenters suggested that a reasonable period for
such reductions would be eight years.  Another commenter supported the phased
approach for load allocations because it allowed an iterative process for
implementing nonpoint and point source controls.

      iii.  Final Guidance:   As part of its reorganization of the general
conditions in the final Guidance, EPA has renumbered proposed general
condition 9 to become general condition 3 in the final Guidance.  As part of
that reorganization, EPA has also incorporated into new general condition 3
the language proposed under the heading "Load Reductions" that defines the
elements of a TMDL.  EPA has also established subparagraphs within general
condition 3 of the final Guidance to correspond to the discussion in general
condition 3 of the elements of a TMDL, nonpoint source load allocations, point
source wasteload allocations, and monitoring.

      Specifically, EPA has added as subparagraph  (a) the statement from
proposed general condition 2 that TMDLs shall include wasteload allocations
and load allocations for nonpoint sources,  including natural background, such
that the sum of these allocations is not greater  than the loading capacity of
the water for the pollutant addressed by the TMDL, minus the sum of a
specified margin of safety and any capacity reserved for future growth.  EPA
has made only minor changes to the proposed language to clarify that the
nonpoint source load allocations  include natural  background conditions and to
link loading capacity to the pollutant for  which  the TMDL is being  developed.

      Subparagraph  (b)  comprises  the portions of  proposed general condition  9
pertaining  to nonpoint  sources.   These provisions were modified  in  general

-------
                               Section Vm.C: TMDLs                           255
condition 3 only to make consistent use of the term loadings and to clarify
that expectations regarding decreased loadings from nonpoint sources must be
based on a reasonableness standard.  The only significant comments EPA
received on proposed general condition 9 addressed nonpoint source issues.
EPA disagrees with the commenter's suggestion that nonpoint source reductions
be considered only when such controls are required by legally enforceable
mechanisms.  EPA suggests that means other than legally enforceable mechanisms
are available to ensure that nonpoint source reductions that are "reasonably
expected to occur" within a specified time frame actually do occur.  For
example, funding nonpoint source controls and using the monitoring component
of the phased approach to TMDL development, as described earlier in this
document, are means to assure that anticipated load reductions are actually
occurring.  Although EPA supports the use of a phased approach to TMDL
development where appropriate, EPA stresses that smaller load allocations to
nonpoint sources can be used to justify larger WLAs to point sources only when
the anticipated reductions in nonpoint source loadings are reasonably expected
to occur.

      EPA agrees with the comment that a TMDL can consider anticipated
nonpoint source loading reductions.  TMDLs developed using the phased approach
are based on the reasonable expectation that water quality standards will be
met in a reasonable period of time and that specific controls may be
implemented in stages.  What constitutes a reasonable period of time will vary
depending upon the situation.  Therefore, EPA will not specify any particular
period, such as eight years. The time period associated with these stages of
implementation will ultimately determine when water quality standards will be
met for a particular waterbody.  To the extent consistent with other
applicable law concerning schedules of compliance, permits issued after the
completion of a TMDL should be consistent with implementation schedules
established by the TMDL.

      Placed within'new subparagraph (c) are the provisions in proposed
general condition 9 pertaining to point source wasteload allocations and their
effect on NPDES permits.  Apart from including a reference to natural
background in connection with nonpoint sources, these provisions are unchanged
from the proposal.

      In the final Guidance, EPA added subparagraph (d) to address the
monitoring issues encompassed within the proposal's discussion of anticipated
decreases in pollutant loadings from nonpoint sources.  Subparagraph (d)
provides that, for load allocations established on the basis of (a) (iii) of
general condition 3,  monitoring data shall be collected and analyzed in order
to validate the TMDL's assumptions, to verify the anticipated load reductions,
to evaluate the effectiveness of controls being used to implement the TMDL,
and to revise the WLAs and load allocations as necessary to ensure that water
quality standards will be achieved within the time period established in the
TMDL.  This monitoring can be performed as part of the water monitoring
program established by the State (or at its election by the Tribe)  under 40
CFR 130.4, which specifies development and review of TMDLs, wasteload
allocations and load allocations as among the uses for such monitoring data.

d.    General Condition 4 - WLA Values

      i.    Proposal:  This general condition was numbered as general
condition 3 in the proposal.  As proposed, this condition specified that point
sources be regulated to ensure attainment of all downstream water quality
standards.  Proposed general condition 3 also recognized that TMDLs developed
for a particular waterbody may include WLAs for sources already covered by a
TMDL of a different geographic scope.  For example, a source-specific TMDL may
already be in place when a basin-wide TMDL is developed.  General condition 3,
as proposed,  provided that water quality-based effluent limits (WQBELs) in
NPDES permits for a particular pollutant be consistent with the most stringent
of the WLAs for that pollutant and point source included in any EPA-approved
or EPA-established TMDLs.  This provision was intended to assure that water

-------
256    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

quality standards will be met throughout a drainage basin, including in
downstream waters. .

      ii.   Final Guidance:  EPA did not receive significant comments on
proposed general condition 3.  Thus, the final Guidance retains  the substance
of the general condition with slight modifications, but EPA has  renumbered it
as general condition 4 in the final Guidance to reflect EPA's decision to move
proposed general condition 9 (TMDL Allocations) up to become new general
condition 3 in the final Guidance.

      This provision in the final Guidance, like the proposal, directs permit
writers to apply the most stringent of the WLAs included in any  EPA-approved
or EPA-established TMDL.  The final Guidance clarifies that this provision
applies only when more than one approved TMDL establishes a different WLA for
the same pollutant discharged by the same point source.  In addition to
renumbering this as general condition 4, EPA made one other change. The
proposed language stating that "point sources must be regulated  so as to
ensure attainment of all downstream water quality standards" has been deleted
in the final Guidance because it merely restated current law.  Specifically,
under existing CWA section 402 and 301(b)(1)(C), WQBELs in NPDES permits must
ensure attainment of all applicable water quality standards, including
downstream water quality standards.  Under 40 CFR 122.44(d)(1)(vii), such
WQBELs must be consistent with any available WLAs developed and  approved
pursuant to 40 CFR 130.7.

e.    General Condition 5  - Margin of Safety

      i.    Proposal;  This general condition was numbered as general
condition 4 in the proposal.  As proposed, this condition reiterated the
requirement in CWA section 303(d) that each TMDL include a margin of safety
(MOS) and described the manner in which the MOS is provided.  It also
reiterated EPA guidance that the MOS may be established either by setting
aside a portion of the loading capacity or by using conservative modelling
assumptions in deriving the TMDL.

      ii.   Comments:  Several commenters were concerned that it would be
inappropriate to leave determination of an MOS to the discretion of the permit
writer.  One commenter recommended that in order to facilitate basin-wide
consistency and maximum environmental protection, the Guidance should
implement an explicit MOS  factor equal to the Criterion Maximum  Concentration
(CMC) value  (which equals  one-half of the Final Acute Value  (FAV)) .  Other
commenters advocated specifying a specific  confidence level to use in modeling
a MOS.

      Several commenters believe that the MOS requirement is redundant  given
the number of conservative assumptions built into the criteria development
process and into the assumptions on fate and transport.

       Several commenters  were concerned that including uncertainties
regarding controlling pollutants from nonpoint sources into the  margin  of
safety merely shifted the  control burden to point sources without requiring
EPA, States or Tribes to regulate other sources of pollution.  They were
concerned that a larger MOS would result in a smaller WLA,  thus  requiring a
facility to discharge less and treat more while nonpoint  sources would  not be
controlled.

      iii.  Final Guidance:  Apart  from minor changes to  improve clarity and
renumbering to reflect  the overall  reorganization of procedure 3.B of appendix
F, the final Guidance is unchanged  from the proposal.  General condition 5
maintains flexibility for  the State or Tribe to consider  a  number of  factors,
including case-specific conditions  (e.g., availability and  quality of data)  in
establishing a margin of safety.  As  indicated  in 40 CFR 130.7(c)(1), the
margin of safety is intended to  account for uncertainty  in  the available  data
or in the actual effect controls  will have  on  loading  reductions and receiving

-------
                               Section Vm.C: TMDLs                           257
water quality.  EPA has determined that because of the need to reflect local
conditions and case-specific technical 'considerations, it is inappropriate to
specify a universal.MOS factor.  Although EPA recognizes the flexibility of
the State or Tribe to assess available information, EPA retains the authority
to disapprove a TMDL if EPA finds that a MOS is inadequate.

      In response to comments that the MOS has the effect of shifting the
burden of load reductions to point sources, EPA notes that the MOS requirement
does not compensate for failure to consider some sources  (e.g., nonpoint
sources as suggested by commenters) but rather is intended to account for any
technical uncertainty regarding both point and nonpoint source loading data
and the effectiveness of controls.  EPA acknowledges that the technical
uncertainty related to nonpoint sources may in fact be greater than
uncertainty regarding the effects of point sources.  EPA believes that the
phased approach to TMDL development provides, over time, an effective
mechanism for reducing technical uncertainty related to nonpoint sources.
This reduction in uncertainty will, over time, quantify and consider relative
contributions and water quality impacts and lead to appropriate levels of
control for both point and nonpoint sources.

      EPA disagrees with the commenters' suggestion that the MOS is redundant
given the conservative assumptions built into the criteria development and
into assumptions on fate and transport.  The MOS, as required by CWA section
303(d), is intended.to account for technical uncertainties regarding the
relationship between pollutant loads and water quality.  These factors are not
considered in the development of criteria and thus are not duplicative of
assumptions used in developing criteria.  Conservative assumptions in criteria
development are designed to address specific uncertainties and concerns
regarding extrapolations of toxicity data to individual or population
endpoints.  EPA also suggests that there should not be an issue of redundancy
regarding the fate and transport assumptions and the MOS.  The assumption of
no pollutant degradation for purposes of TMDL development is rebuttable when
scientifically valid field studies or other relevant information demonstrate
that degradation of the pollutant is expected to occur.

f.    General Condition 6 - More Stringent Requirements

      This general condition was numbered as general condition 4 in the
proposal.  As proposed, this condition provided that States may employ section
510 of the CWA to establish TMDLs more stringent than those developed pursuant
to procedure 3. The condition reiterated the reserved right of States to
require more stringent controls than those required under the CWA.

      EPA received no significant comments on this provision.  The proposed
language is modified slightly in the final Guidance to clarify that both
States and Tribes may employ section 510 and to correct a typographical error.
It has also been renumbered as general condition 6 to reflect the overall
reorganization of procedure 3.B of appendix F.

g.    General Condition 7 - Accumulation in Sediments

      i.    Proposal:  This general condition was numbered as general
condition 6 in the proposal.  As proposed, this condition specified that TMDLs
must be stringent enough to prevent accumulation of the pollutant of concern
in sediments to levels injurious to designated or existing uses, human health,
wildlife and aquatic life.  It also specified that TMDLs consider
contributions to the water column from sediments inside and outside applicable
mixing zones.  Although TMDLs are calculated on the basis of pollutants in the
water column, the preamble to the proposal indicated that all sources of
pollution, including sediment re-release of pollutants to the water column,
would need to be considered in establishing TMDLs.

      ii.    Comments:  EPA received numerous comments on this condition.  A
number of commenters disagreed with the proposal. Several suggested that

-------
258    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

proposed general condition 6 be deleted until EPA finalizes and  implements a
national sediment strategy.  One commenter suggested that proposed general
condition 6 be optional depending on the availability of information or, if
not, that general condition 6 should be removed entirely.

      Several commenters stressed the importance of considering  the release of
toxics from contaminated sediments, which in many instances may  result in a
failure to meet water quality standards.  Several commenters, while agreeing
with the need to consider all sources of pollutants, including sediment
release or resuspension of pollutants,  believe that methodologies do not
currently exist to accurately reflect the sediment re-release process.  One
commenter suggested that sediments should only be accounted for  by
concentrations measured in the water column and that any additional factors
would be duplicative.  Commenters recommended that EPA continue  to work on
National guidance fbr such methods and suggested that any process for
developing sediment criteria should be subject to a peer review process.

      iii.  Final Guidance:  The final Guidance retains the requirement that
TMDLs reflect processes such as re-release of pollutants from sediments,
because, as noted by many commenters, contaminated sediments are often a
source of pollutant loading to the water column and thus may cause or
contribute to an exceedance of water quality standards.  However, EPA has
modified this provision to clarify that such contributions should be
considered only where appropriate and where sufficient data are  available.
EPA has renumbered this provision as general condition 7 to reflect the
overall reorganization of procedure 3.B of appendix F.

      EPA agrees with commenters that existing methodologies may not fully
reflect all aspects of the sediment re-release process.  However, EPA recently
proposed its Contaminated Sediment Management Strategy  (EPA 823-R-94-001) for
public comment, 59 FR 44880, (August 30, 1994, available in the  docket), and
is continuing to develop methodologies to evaluate the sediment  re-release
process.  The strategy proposes establishing standardized test methods to
assess whether sediments are contaminated and proposes to continue supporting
research on the re-release of pollutants from contaminated sediment.  Under
the strategy, EPA  would develop new biological methods to assess the
ecological and human health effects of sediment contaminants, sediment
wasteload allocation models, and technologies for remediation of contaminated
sediment.  EPA is also working to develop chemical-specific sediment quality
criteria.  This process will involve review from outside parties.  See 59 FR
2652, January 18, 1994 for further information.

      EPA is moving forward with many of the activities described in the draft
Contaminated Sediment Management Strategy and expects many of these activities
to be completed in time to support State and Tribal procedures under part 132.
The final Contaminated Sediment Management Strategy and associated outreach
efforts will support States and Tribes in implementing general condition 6.
Therefore, EPA disagrees with the comment that this condition needs to be
deleted until EPA finalizes the Strategy.

      Several commenters suggested that situations may exist where information
is not available to determine the nature and extent of contaminated sediments'
contributions of pollutants to the water column.  EPA has modified the final
Guidance to specify that contributions to the water column from  contaminated
sediments be included where appropriate.  It may be considered appropriate to
reflect contributions of pollutants from contaminated sediment only where data
exist regarding sediment re-release of the pollutant(s) of concern.  Where
such information does exist, however, the TMDL must account for  contributions
from contaminated sediments.

      In the final Guidance, EPA has reversed the order of the two sentences
appearing in proposed general condition 6 in order to emphasize  that
contaminated sediments can be sources of pollutants to the water column  and
that TMDLs need to account for contributions from that source.   As in the

-------
                               Section Vm.C: TMDLs                           259
proposal, in addition to specifying that sediment re-release of pollutants
shall be considered where appropriate, the final Guidance provides that TMDLs
must be sufficiently stringent so as to prevent injurious accumulation of the
pollutant of concern in sediments, because such injurious accumulations would
represent exceedances of Water quality standards  (at a minimum by impairing a
designated aquatic life use).

h.    General Condition 8 - Wet Weather Events

      i.    Proposal:  This general condition was numbered as general
condition 7 in the proposal.  As proposed, this condition recognized that some
of the TMDL development procedures may be appropriate for wet weather events
 (e.g., nonpoint sources, storm water discharges, and combined sewer
overflows).  However, the proposed TMDL implementation procedures did not
include explicit procedures detailing how to develop TMDLs to reflect wet
weather events; rather it left maximum flexibility to the States and Tribes on
how best to accomplish this.  The preamble discussion of proposed general
condition 7 interpreted that condition as providing that loadings from wet
weather events be included in establishing TMDLs, but the proposal itself was
silent on this point.

      ii.   Comments;  Several commenters suggested that proposed general
condition 7 needed to clarify that all TMDLs must include consideration of
necessary waste load allocation and load allocations for wet-weather pollutant
contributions.  Another commenter pointed out that certain POTWs face
compliance difficulties as a result of wet weather flows.  The commenter
suggested that these factors, which are beyond the control of the POTW, be
considered in developing permit limits. Several commenters asserted that wet
weather contributions cannot be accurately estimated and therefore suggested
this general condition be removed all together.

      iii.  Final Guidance:  The final Guidance retains the proposed language
on wet weather flows with minor modifications and an additional sentence for
clarification purposes.  This provision has been renumbered in the final
Guidance as general.condition 8 to reflect the overall reorganization of
procedure 3.B of appendix F.  EPA agrees with the commenter's suggestion that
this general condition should be clarified to state specifically that TMDLs
must consider pollutant loadings resulting from wet weather events, where
appropriate and where sufficient data are available.  EPA believes TMDLs
reflecting wet weather events would be appropriate where such events
contribute the pollutant(s) during the flow conditions for which the TMDL is
being developed.  For example, the TMDL for a pollutant that has an annual
averaging period (e.g., dioxin) would need to consider loadings from wet
weather events because such events can occur during the yearly averaging
period. However, a TMDL  based on a 7-day critical low flow (e.g., lead) for a
pollutant that has a 4-day averaging period would not directly consider
loadings from wet weather events because such events are unlikely to occur
during critical 7-day low flows. Contributions from previous wet weather
events would be considered through load allocations to the sediment.  In
addition,  a TMDL based on dynamic or stochastic water quality model would
include all dry and wet weather loadings from all sources.  In any case, where
the TMDL for the receiving water accounts for loadings that occur from wet
weather events, the resulting WLAs, including those for POTWs, must be
consistent with the TMDL and WLAs.  The only exception to this is where the
POTW discharge meets the definition of wet weather point source under 132.5.
The final Guidance does not regulate wet weather point sources.

      Many nonpoint sources and wet weather point sources as defined at
section 132.2 of this Guidance typically have their greatest impacts following
storm events and the influx of pollutants from these events needs to be
factored in when establishing a TMDL to ensure attainment of water quality
standards.  Accordingly, EPA has inserted language to clarify this point in
the final Guidance and has amended 132.4 (e) (1) to provide specifically that
procedure 3 applies to wet weather events, as appropriate.  Like the proposal,

-------
260    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

the final Guidance does not require a specific procedure to address wet
weather flows, but rather leaves it to the discretion of the State or Tribe to
choose the most appropriate procedure, considering ail relevant  facility
specific, .pollutant specific, and receiving water specific factors.

      In EPA's view, this clarification will not subject POTWs to any
additional burden.  Any adjustments to a POTW's permit conditions to account
for wet-weather flows should be addressed through the NPDES permitting  and
enforcement policies and procedures.  Finally, EPA disagrees with the comments
asserting that wet weather contributions cannot be accurately estimated.  A
number of models currently exist to generate loadings estimates  from a  range
of wet weather events.  EPA is working on additional guidance on assessing
pollutant loadings associated with CSOs and nonpoint sources (see "Technical
Guidance for Estimating Total Maximum Daily Loads (TMDLs):  Integrating
Steady-State Episodic Point and Nonpoint Sources, draft, June 1994, available
in the docket).

i.    General Condition 9 - Background Concentrations of Pollutants

      This general condition was numbered as general condition 8 in the
proposal.  As proposed, this condition established procedures for determining
representative background concentrations of pollutants to assure that
background concentrations are consistently considered in TMDL development
among the Great Lakes States.  The proposal included provisions  for
calculating background.  The proposal defined background, described the choice
of data set, the use of the geometric mean, and the treatment of data sets
with data points abpve and below detection.  EPA received no significant
comments on the definition of background and the proposed language is retained
in the final Guidance with only minor changes to account for the use of the
term in procedure 5.  The proposal, comments and the final Guidance for each
provision are discussed below.  EPA renumbered this provision as general
condition 9 in the final Guidance to reflect the overall reorganization of
procedure 3.B of appendix F.

i.    Choice of Data Set

      (A)   Proposal:  The proposal provided that the representative
background concentration for a pollutant shall be established as the geometric
mean of one of three possible data sets: available ambient water column data
(e.g., ambient monitoring data), representative caged fish tissue data, or
representative pollutant loading data.  When more than one data  set exists,
best professional judgment  (BPJ) would be used to determine which data  set
most accurately estimated background concentrations.  The preamble to the
proposal stated that, in general, ambient monitoring data are preferred over
other sources of data.  The preamble also recognized that there  may be
instances where other data sets may be more appropriate, such as where  ambient
data are not available, or where ambient data are not as informative or
reliable as either caged fish tissue data or pollutant loading data because of
limits in analytical detection methods.

      (B)   Comments:  Several commenters supported EPA's proposal to allow
States and Tribes to choose among data sources. Others suggested that,  by
allowing a choice of data sets, there was too much discretion allowed to the
State or Tribe in establishing background levels and suggested that EPA
provide more specific guidance on the choice of data sets.

      One commenter suggested that States and Tribes should be required, where
possible, to eliminate unrepresentative data from the data set using factual
information and statistical methods.  Commenters suggested that  more recent
data should take precedence over older data even when the more recent data  set
is smaller.  Furthermore, they believe that data more than five  years old
should not be considered.  One commenter suggested that  fish tissue and
pollutant loading calculations should be rejected as acceptable  data sets when
those calculations predict background concentrations above the criteria for

-------
                               Section vm.C: TMDLs                           261
ambient monitoring data and such concentrations were not detected by ambient
monitoring.

      Several commenters advocated that only ambient data be used to estimate
background concentrations. Other commenters wanted the Guidance to require
regulators to use ambient monitoring data to calculate background
concentrations of pollutants when such data is available.

      A number of commenters disagreed with the requirement to consider caged
fish tissue data in calculating background concentration because procedures
for the use of caged fish analysis have not been thoroughly evaluated,
validated, or standardized.  Several commenters believe that the quality data
necessary to provide accurate background data using the caged fish approach is
not available.  Commenters suggested that using  resident fish tissue as a
basis for deriving background would be more accurate.  A commenter further
suggested that EPA attempt to calibrate the fish tissue and pollutant loading
models with real data.  Commenters also requested more specific procedural and
technical information relating to use of caged fish data.

      (C)   Final Guidance: In response to comments and concerns, EPA has
added resident fish tissue data as a fourth specified data set available for
calculating background.  Apart from that, EPA retains the proposed language
with only minor modifications to ensure clarity and avoid redundancy.

      In the final Guidance, EPA has consolidated into a single section the
list of available data sets and the basis for determining what available data
is acceptable for use in calculating background.  These provisions are now
included in the subparagraph specifying calculation requirements.  The final
Guidance retains flexibility for States and Tribes to choose from among a
number of data sets, including fish tissue data, in calculating background
concentrations.  EPA concludes that because of wide variability in the
suitability of available data for a particular situation and because of site-
specific considerations, use of BPJ is appropriate to make case-by-case
determinations.  EPA recognizes that more recent data, with improved detection
or quantification levels may be more appropriate, while some older data with
poorer detection or quantification levels may be less acceptable.  However,
EPA recognizes that, in some instances, the older data may be the only data
available may be the only representative data of sufficient quality from which
to make decisions and thus is not establishing a prohibition on the use of
older data.  In the*final Guidance, the State or Tribe retains the flexibility
to use BPJ to eliminate unrepresentative data or to give greater weight to the
most recent data as suggested by commenters.  States and Tribes may also use
statistical techniques to identify and eliminate unrepresentative data.

      The final Guidance thus does not include more specific direction to
limit the use of any particular data set.  Although EPA agrees with the
commenters1 suggestion that ambient monitoring data are generally preferred
over other data sources, there may be situations where ambient data are not
available, or are not as informative or reliable as either fish tissue or
pollutant loading data because of limits in analytical detection methods.
Because of limits in existing technologies, ambient data may still yield non-
detects above criteria levels.  Fish tissue data and pollutant loading data
may be particularly useful alternatives for these situations.

      EPA recognizes that caged fish tissue studies may have limitations in
that such studies may not fully account for duration of exposure and food
chain magnification.  However, EPA has determined that such studies should be
considered with other data sources in choosing among data sets to calculate
background concentration.  Aquatic organisms can serve as valuable indicators
of whether water quality standards are being attained.  The final Guidance
also authorizes the'use of resident fish tissue data, as suggested by
commenters, because of concerns regarding food chain effects and in response
to concerns about the lack of caged fish tissue data.  Use of resident as well

-------
262    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

as caged fish tissue data is intended to provide more latitude in selecting
the appropriate data set.

      Like the proposal, the final Guidance does not provide a methodology to
use in translating fish tissue concentrations to a water column concentration,
or for evaluating their validity.  EPA agrees that care should be exercised in
determining what fish tissue data are representative of background pollutant
concentrations and encourages permitting authorities to consult EPA guidance
on this topic.  For,example, EPA recommends that when fish tissue data are
available from resident fish the geometric mean is divided by the
bioaccumulation factor pursuant to the methodology in appendix B of this final
Guidance, to yield estimated ambient concentrations.  See Assessing Human
Health Risks from Chemically Contaminated Fish and Shellfish: A Guidance
Manual (USEPA, September, 1989, EPA-503/8-89-002, available in the docket).

       EPA believes that best professional judgment should be used to
determine if caged fish tissue data is appropriate for calculating background
concentration in -a given situation.  Furthermore, the use of caged fish tissue
data is not required unless no other data exist to calculate background
pollutant concentrations.  Even in a situation where only caged fish tissue
data exists, a facility always has the option to collect alternative data that
more accurately reflects background concentrations  (e.g., ambient monitoring
data).  In addition, the final Guidance does not require that new fish tissue
"studies" be conducted in the absence of existing fish tissue data.

ii.   Geometric Mean

      (A).  Proposal;  The proposal specified that the representative
background concentration for a pollutant shall be established as the geometric
mean of one of the selected data sets described in that paragraph and the
preamble offered guidance for performing the calculations.  EPA is retaining
the proposed language in the final Guidance.  The Agency received no
significant comments on this provision.

      As the preamble to the proposal explained, a geometric mean is
calculated for the set of data chosen to represent background conditions.  The
geometric mean calculated is based on both measured concentrations and an
appropriate methodology for treating measurements below quantification levels.
For pollutant loading data, the geometric mean should be taken of pollutant
loading data from individual sources.  The individual means of each of the
individual sources should then be added to estimate total loading to the
receiving water.  Background concentration is calculated by dividing total
loadings by the volume of water available at the appropriate design flow.
Design flow will vary depending on the criterion being implemented at the
point immediately upstream of the watershed, water body or water body segment
for which the TMDL, WLA in the absence of a TMDL, or preliminary WLA for the
purpose of determining reasonable potential under procedure 5 of this Guidance
is being established.  For further discussion, see the preamble to the
proposal at 58 FR 20929.

      (B).  Final Guidance:  EPA received no significant comments on this
provision.  EPA believes that the use of the geometric mean is the best
approach for calculating a median concentration from data chosen to represent
background conditions.  An arithmetic mean would be one appropriate method for
calculating median values when a sample concentration is as likely to be above
the true average concentration as it is to be below the true average
concentration.  However, concentration measurements in fish tissue and water
are more likely to be below the true average concentration.  Under these
conditions, the geometric mean is an appropriate estimator for the median
while the arithmetic average will generally produce a value that is higher
than the median.  More explicitly, fish tissue and water concentration
measurements generally follow positively skewed probability distributions
where the median is appropriately estimated by the  geometric mean.

-------
                               Section Vm.C: TMDLs                           263
iii.  Data Points Above and Below Detection

       (A)   Proposal:  The proposal allowed the use of best professional
judgment to determine which data points are acceptable.  However, within a
given data set, some data points may indicate that the pollutant was not
present at levels capable of being detected by the analytical method used.
For these data points, the true concentration of the pollutant can be zero or
is somewhere between zero and the detection level of the analytical method.
Other data points may indicate that the pollutant was detected, but at levels
below which the analytical method is capable of reliable quantification.  For
these data points, the true concentration will be between the detection level
and the quantification level of the analytical method.  Finally, there may be
data points showing reliably quantified levels of the pollutant.        The
proposed Guidance specified that the following assumptions be used in
calculating background when, within a data set, some data points are
determined to be above and others are below the detection level.  The proposal
included the following assumptions: data points reported at levels below
detection shall be set equal to one half of the detection level; and data
points reported at levels greater than the detection level but less than the
quantification level, shall be set equal to the midpoint between the detection
level and the quantification level.  If all acceptable available data points
in a data set are reported as below the detection level for a specific
pollutant, then all the data for that data set are assumed to be zero.

      Section 132.2 of the proposed Guidance included a definition of
detection level that is identical in substance to the definition at 40 CFR
136.2 (f).  There is no similar long-established definition of the term
quantification level.  However, the proposal defined the quantification level
as the concentration at which a particular substance can be quantitatively
measured using a specified laboratory procedure.  EPA solicited comment on the
definition and on the issue of whether a particular degree of confidence
should be specified.

       (B)   Comments:  EPA received a number of comments on the proposal to
assign values equal to one-half of the detection level to data points reported
below the detection level when other data points in the data set were reported
above the detection level.

      Several commenters supported the use of one-half of the detection level
as a default.  Another commenter suggested that if 25% or more of the data
points are quantifiable, the remaining values reported as less than the
detection limit should be zero. One commenter advocated that the requirement
to use one-half of the detection level as a background concentration be
deleted and the evaluation left to best professional judgment.  Another
commenter recommended that if a large proportion of the data is reported as
non-detect,  assumptions regarding what value to assign should be left up to
the permitting agency and that such determinations need to be made on a ca.se-
by-case basis rather than through the application of a general rule.  Several
commenters wanted EPA to allow the use of appropriate statistical methods for
data sets that include a large number of values below the detection limit and
further advocated that the final Guidance cite examples of such statistical
methods.  One commenter suggested that a statistically valid sliding scale be
used to assign concentration values to any non-detect measurements.  Another
commenter expressed concern that the proposed approach will result in
unrealistically high background concentrations for data sets with a large
share of measurements below the detection level and suggested that the final
Guidance include methods presented in "NCASI Technical Bulletin No. 621."
Several commenters supported the use of one-half the detection level when
calculating means or averages from data sets that include non-detect values.

      EPA also received comments addressing situations where some of the data
is between the detection level and the quantification level.  One commenter
suggested that the final Guidance require the quantification level be used as
the default value in determining the mean for pollutants that have caused or

-------
264    Water Quality Guidance for the Great Lakes System — Supplementary Information Document

contributed to fish advisories downstream.  Another commenter suggested that
when data points are below the detection level or quantification level, zero
or a default percentage of the criteria value, when the criteria value  is also
below the level of detection, should be assumed.

      In situations where all the data points in a particular data set  are
below detection, several commenters agreed with the proposal that these data
points should be assumed to be zero.  One commenter suggested that for  a data
set of more than ten data points, the proposal should apply, but that if there
are fewer  than ten data points and all the data points are below detection,
background shall be assumed to be one-half the detection level.    Several
commenters supported EPA's definition of quantification level.  A few
commenters did not support including the quantification level definition.
Another commenter suggested that the definition of quantification level should
be the same as that used in setting the Compliance Evaluation Level  (CEL) for
determining permit compliance in proposed procedure 8.  The CEL was defined in
the proposal, as the level at which compliance with an effluent limit is
assessed.  Some commenters advocated that the term "detection level" be
changed to "method detection level" since the proposal defined detection level
the same as method detection level is defined in 40 CFR 136.

      (C)   Final Guidance;  The final Guidance recognizes the need for
flexibility when calculating background using a data set containing data
points both above and below the detection level or quantification level.  EPA
has concluded that, for these data sets, although default values of one-half
of the reported detection level for data points reported as below detection,
and the mid-point between the detection level and quantification level  for
data points reported below the quantification level and above the detection
level, are a reasonable and appropriate estimate for purposes of calculating
background concentration, they are not the only reasonable and appropriate
approach.  As many of the commenters pointed out, there are a number of
commonly accepted statistical approaches to evaluating mixed data sets  (also
known as censored data sets).  Therefore, in the final Guidance, States and
Tribes are required to use commonly accepted statistical techniques to
evaluate data sets containing values both above and below the detection level.
Commonly accepted statistical techniques can include a variety of approaches,
including the use 'of default values as proposed.  Some commonly accepted
statistical techniques are outlined in Chapter 14 of Statistical Methods for
Environmental Pollution Monitoring  (Richard 0. Gilbert; published by Van
Nostrand Reinhold) and Truncated and Censored Samples  (A.Clifford Cohen;
published by Marcel Dekker).

      Because there is no universal method to reliably quantify pollutant
concentrations below the detection level, EPA believes that using a default
value of one-half of the reported detection level is a reasonable balance of a
State and Tribes' obligation to provide dischargers with an appropriately
stringent WLA and the statutory requirement that TMDLs ensure the attainment
of water quality standards.  The same reasoning applies when calculating WLAs
in the absence of a TMDL.  Likewise, EPA has concluded that the  reasoning
above also supports-using the mid-point between the detection level and
quantification level as an acceptable,