-------
                              Preface
      This document was finalized in June 1989 and reviews
 information from relevant studies of O3  health and welfare
 effects  and of  exposure and risk analysis through early 1989.
 The assessment  contained in this staff paper  reflects  information
 in the documents  "Air  Quality Criteria for Ozone  and Other
 Photochemical Oxidants"  (EPA-600/8-84-020F) and "Summary  of
 Selected New .Information on  Effects of Ozone on Health and
Vegetation:  Supplement to Air Quality Criteria for Ozone and
Other Photochemical Oxidants" . (EPA-600/8-88/l-5a).

-------

-------
                          Acknowledgements
       This staff paper is the product of the Office of Air Quality
  Planning and Standards (OAQPS).   Tables and Figures .not otherwise
  cited are original to this report.   The principal authors include
  Dr.  David J.  McKee,  Ms.  Pamela M. Johnson,  Mr.  Thomas R.  McCurdy,
  and  Mr.  Harvey  M.  Richmond.   This report has been improved by
  comments  from other  staff  within OAQPS,  the Office of Research
  and  Development, the  Office of Policy and Program Evaluation, and
  the  Office of General Counsel within EPA.   Three  drafts were     '
  formally reviewed by the clean Air Scientific Advisory Committee
 and comments incorporated.  Particularly important in the final
 review of this staff paper was the technical and editorial
 support provided by Ms; victoria  Atw.ll and the clerical and
 editorial support of Mrs. Patricia R.  Crabtree  and Mrs. Barbara
 Miles.
      Helpful  comments  and suggestions  were  also  submitted by a
 number  of  independent  scientists,  by officials from  the  state
 environmental agencies of  Illinois, Minnesota, California and
 Texas,  by  the Department of the Havy,  and the Department  of
 Energy, and by environmental and industrial groups including  the
Natural Resources Defense Council, the American Lung Association,
the Chemical Manufacturers Association, the American Petroleum
institute,  and the Motor Vehicle Manufacturers Association.

-------
                         Project Team For
  Review of the National Ambient Air Quality Standards for Ozone
Dr. David J. McKee, Project Manager and Author  of  Chapters I
through III and VI through VI11
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards  (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C.  27711

Ms. Pamela M. Johnson, Author of Chapters IX through XI
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards  (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C.  27711

Mr. Thomas R. McCurdy, Author of Chapters IV and V and Appendix A
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards  (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C.  27711

Mr. Harvey M. Richmond, Author of Section VII.B,,
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C.  27711

-------
                                 iii

                U.S.  Environment,il  Protection Agency
                        Science Advisory Board
               Clean Air Scientific Advisory Committee
                                    -OH
  Chairman

  Dr.  Roger  O.  McClellan
  CUT
  Post Office Box  12137
  Research Triangle  Park,  NC  27709


  Members

  Dr.  Eileen G. Brennan
  Department of Plant Pathology
  Martin Hall, Room  213
  Lipman Drive
  Cook College-NJAES, Rutgers Univ.
  P.O. Box 231
 New Brunswick, New Jersey  08903

 Dr. Edward D.  Crandall
 Division of Pulmonary Medicine
 Starr Pavilion 505
 Cornell Medical College
 1300 York Avenue
 New York, New York  10021

 Dr. James D.  Crapo
 Box 3177
 Duke University Medical Center
 Durham, North  Carolina   27711

 Dr.  Robert  Frank
 Professor of Environmental Health
 Sciences
               School of Hygiene

 615 N. Wolfe Street
 Baltimore, Maryland 21205

 Prof. A. Myrick Freeman, in
 Department of Economics
Bowdoin College
Brunswick, Maine  04011

-------
                                 IV

 Dr. Jay S. Jacobson
 Plant Physiologist
 Boyce Thompson Institute
 Tower Road
 Ithaca, New York  14853

 Dr. Jane Q. Koenig
 Research Associate Professor
 Department of Environmental
 Health SC-34
 University of Washington
 Seattle,  Washington  98195

 Dr. Timothy Larson
 Environmental Engineering and
 Science Program
 Department of Civil Engineerincr
 FX-10
 University of Washington
 Seattle,  Washington  98195

 Dr.  Morton Lippmann,  Professor
 Institute of Environmental Medj cine
 NYU Medical Center
 Tuxedo, New York   10987

 Prof. M.  Granger Morgan
 Head, Department of  Engineering
 and Public Policy
 Carnegie-Mellon University
 Pittsburgh,  Pennsylvania   15253

 Dr.  D. Warner  North,  Principal
 Decision  Focus, Inc.
 Los  Altos  Office Center
 Suite 200
 4984 El Camino Real
 Los  Altos,  California  94022

 Dr.  Gilbert S. Omenn,
 Professor  and  Dean
 School of  Public Health and
 Community  Medicine SC-30
 University of Washington
 Seattle, Washington  98195

 Dr. Robert D. Rowe
 Energy and Resource Consultants
P.O. Drawer 0
Boulder, Colorado  80306

-------
 Dr. Marc B. Schenker,  Director
 Occupational and Environmental
 Health Unit
 University of California
 Davis, California  95616

 Mr. Stephen Smallwood
 Air Pollution Control Program
 Manager
 Bureau of Air Quality Management
 Florida Department of Environmental
 Regulation
 Twin Towers Office Bldg.
 2600 Blair Stone Road
 Tallahassee,  Florida 32301

 Dr.  George Taylor
 Environmental  Sciences Division
 P.O.  Box  X
 Oak Ridge National  Laboratory
 Oak Ridge,  Tennessee   37831

 Dr.  Mark  J. Utell
 Pulmonary Unit - Box  692
 Strong Memorial Hospital
 Rochester, New York   14642

 Dr. Jerry Wesolowski
 1176 Shattuck Avenue
 Berkeley, California  94704

 Dr. George T. Wolff
 Senior Staff Research Scientist
 General Motors Research Labs
 Environmental Science Department
Warren, Michigan  48090

-------
                                 vi

                           EPA Reviewers

 Mr. Allen C. Basala  (MD-12)
 ?TfeiC™?f Air QualitY Planning and Standards, OAR
 U.S. EPA
 RTF, NC 27711

 Mr. Frank L. Bunyard (MD-12)
 Office of Air Quality Planning and Standards, OAR
 U * o * EPA
 RTF, NC 27711

 Dr. Thomas c. Curran (MD-14)
 Office of Air Quality Planning and Standards, OAR
 U.S. EPA
 RTF, NC 27711

.Mr. Robert Fegley (PM-221)
 Office of Policy Analysis,  OPPE
 U.S.  EPA
 Waterside Mall
 401 M Street,  SW
 Washington,  DC   20460

 Mr. Lewis Felleisen
 Air Programs &  Engineering  Branch
 U.S.  EPA,  Region III
 Curtis  Building
 6th & Walnut Streets
 Philadelphia, PA  19106

Mr.  Robert A. Flaak (A-107F)
 Science Advisory Board,  OA
U.S. EPA
Waterside Mall
401 M Street, SW
Washington,  DC   20460

Dr. J.H.B. Garner  (MD-52)
Environmental Criteria and Assessment Office, ORD
U.S. EPA
RTF, NC 27711

Mr. Gerald K. Gleason (LE-132A)
Office of General Counsel
U.S. EPA  '
Waterside Mall
401 M Street, SW
Washington, DC  20460

-------
                                  vii
  Dr. Judith A. Graham  (MD-52)
  Environmental Criteria and Assessment  office,  ORD
  RTF, NC 27711

  Dr. Lester D. Grant (MD-52)
  U?l!rEPASntal Criteria and Assessment  Office,  ORD
  RTF, NC 27711

  Dr. Carl G.  Hayes (MD-55)
  Sealt™i?ffects Research Laboratory, ORD
  U. o. EPA
  RTF, NC 27711

  Dr. Donald H.  Horstman (MD-58)
  ne?lt:5™ffects Research Laboratory, ORD
  u. b.  EPA
  RTF,  NC 27711

  Mr.  William F. Hunt  (MD-14)
  SfI!CEPAf  Alr  QUality  Planning  and  Standards,  OAR
  RTF, NC  27711

 Mr. Michael H. Jones (MD-12)
 S!b-!CEPAf Alr QUality  Planning  and  Standards,  OAR
 RTP, NC 27711

 Mr. Bruce c.  Jordan (MD-12)
 Sfs!CEPAf Alr QUality Panning and Standards, OAR
 RTF, NC 27711

 Mr.  Bruce Madariaga (MD-12)
 U?S*°SpAf Air  Quality Planning and Standards, OAR
 RTF, NC 27711

 Dr.  William F.  McDonnell  (MD-58)
 U?S   EPA"*0*3  Research Laboratory,  ORD
 RTF,.NC  27711

 Mr.   Thomas  B. McMullen  (MD-52)
 Environmental Criteria  and  Assessment Office, ORD
RTF, NC  27711

-------
                                viii

 Dr. Edwin L. Meyer (MD-14)
 Office of Air Quality Planning and Standards,
 U.S. EPA
 RTP, NC 27711

 Dr. John J. O'Neil (MD-58)
 Health Effects Research Laboratory, ORD
 U.S. EPA
 RTP, NC 27711

 Mr. Norman C. Possiel (MD-14)
 Office of Air Quality Planning and Standards,
 U.S. EPA
 RTP, NC 27711

 Mr. James A.  Raub (MD-52)
 Environmental Criteria and Assessment Office  ORD
 U.S. EPA
 RTP, NC 27711

 Mr. Robert Rose  (ANR-443)
 Office of Policy,  Planning,  and Evaluation
 U.S.  EPA
 Waterside Mall
 401 M Street, SW
 Washington, DC   20460

 Mr.  Joel  Scheraga  (PM-221)
 Office of  Policy Analysis,  OPPE
 U.S.  EPA
 Waterside  Mall
 401 M Street, SW
 Washington, DC   20460

 Mr.  William P. Smith  (PM-223)
 Office  of  Stds.  & Regulations,  OPPE
 U.S.  EPA
 Waterside  Mall
 401  M  Street, SW
 Washington, DC   20460

 Dr. Joseph Sommers
 Emission Control Technology Division
 Office of Mobile Sources, OAR
Ann Arbor, MI  48105

Ms. Beverly E. Tilton  (MD-52)
 Environmental Criteria and Assessment Office, ORD
U.S. EPA
RTP, NC 27711

-------
                                IX
Dr. Dave T. Tingey
Environmental Research
Laboratory—Corvallis/ORD
200 S.W. 35th Street
Corvallis,  OR  97333

-------

-------
                                 X


                         Table of Contents
                                      •

                                                             Page

 Acknowledgements	



                                National -"«*Air
                                                             11

        e* Scientific Advisory Committee Subcommittee

          	  iii

 EPA Reviewers	
                      	  vi

 Table of Contents	
                             	  x

 List of Figures	
                      	  xv

 List of Tables	
                         	  xviii

 Executive Summary	
                     	  xxi

 I.    Purpose	
                       	  1-1

 II.   Background....
                    	  II-l

 III.  Approach-	
                       	  III-l

 IV.   Ambient Ozone Concentrations in Urban and Rural Areas,  iv-l

      A.  Urban Areas....
                        	 IV-1

      B.  Non-MSA Areas...
                         	 IV-2

      C.  Natural Ozone Background	           Iv_


V.    Ozone Exposure Analysis	


     A.  Overview of the  Ozone NAAQS Exposure Model	V-l


     B.  Air Quality Concentrations  in Microenvironments... v-2


     C.  Simulation of  Population Movement	 v_4


     °*  Model.ArSaS  M°deled in 020ne NAAQS Exposure
              	 V-4

     E.  Exercise Modeling in  Ozone  NAAQS
         Exposure Model...
                            	 V-7

     F.   Eight-Area Aggregated Estimates of Population
         Exposure to Alternative Ozone Standards.??"?	v_8

-------
                                XI


                                                             Page

     G.  Caveats and  Limitations	v-13

VI.  Factors Relevant to Review of the Primary  Standard(s)
     for Ozone	' ^  vi-1


     A.  Ozone Absorption and Mechanisms of Effects	..  vi-l

     B.  Factors Affecting Susceptibility to Ozone	  vi-3

          1 •  Age	  VI_4

          2.  Sex	  VI_5

          3.  Smoking Status	  VI-6

          4.  Nutritional Status	  VI-7

          5.  Environmental Stresses	  VI-8

          6.  Exercise	  VI-8

     C.   Potentially Susceptible Groups	VI-9

          1.  Individuals Having Preexisting Disease.".	  VI-9

          2.  Exercising Individuals	  VI-13

VII. Assessment of Health Effects and Related Health Issues
     Considered in Selecting Primary Standard(s) for
     Ozone	  VII-i


     A.   Health Effects of Concern	  VII-l

          1.    Alterations in Pulmonary Function	  VII-2

          2.    Symptomatic Effects	  VII-15

          3.    Exercise Performance	VII-20

          4.    Bronchial Reactivity and Inflammation	VII-22

          5.    Aggravation of Existing Respiratory Disease  VII-24

          6.    Morphological Effects	„. viI-28

          7.    Effects of Ozone on Host Defense Mechanisms
                in Experimental Animals	 VII-32

          8.    Extrapulmonary Effects	VII-35

-------
                                  xii


                                                             Page

      B.   Pulmonary Function and Symptom Health Risk*
           Assessment	
                       	 VII-37

           1.   Overview of Lung Function and Symptom
                Risk Assessment	. _	VII-37


           2.   Benchmark Risk Results	 Vli-40

           3.   Headcount Risk Results	                  T7TT .c
                                               ••••••••	 VXJ.—45

           4.   Caveats and Limitations	 Vll-so


      C.    Related  Health Effects Issues	 VII-52


           1.    Adverse Respiratory  Health  Effects of
                Acute Ozone Exposure	 VII-53


           2.    Attenuation of Acute Pulmonary  Effects	 VII-56

           3.    Relationship Between Acute  and  Chronic
                Effects	
                       	 VII-58

           4.    Effects, of  Other Photochemical Oxidants...1.  viI-62


           5.    Interactions with other Pollutants!	  VII-63

VIII.      staff Conclusions and Recommendations for Ozone
          Primary Standard(s)	   .       VIII-


          A.   Pollutant Indicator	    VIIl-i


          B.   Form of the Standard	  VIH-4

          C.   Averaging Time(s)	 VIIl-5


          D.   Level of the Primary Standard (s)	 VIII-9
          E.    Summary of Staff Recommendations
                                                 	  VIII-20

                             Review of the Secondary

                             	  IX-l
    A.   Mechanisms of Action  for Vegetation.
         1.   Biochemical Response..,
         2.   Physiological Response.
                                                            IX-l
    B.   Factors Affecting Plant Response	


                                   	 IX-6
                                  	 IX-5

1.    Biological Factors....

-------
                                xiii
                                                             Page

               a.   Plant Genetics	  IX-6
               b.   Developmental Factors ..."....!!!!!!]****  ix-7
               c.   Pathogen and Pest Interactions	
                      with Ozone	  IX-7

          2.   Physical Factors	   IX-8

          3.   Chemical Factors	  IX-9

               a.    Multiple Pollutants	  IX_9
               b.    Chemical Sprays	!!!!!!""  ix-li
               c.    Heavy Metals	'.'.'.'.'.'.'.'.'.I',  ix-11

X.   Assessment of Welfare Effects and Related Welfare
     Issues Considered in Selecting Secondary Standard(s)
     for Ozone	                        v
                                ************ *•* •••••••»•••••  x—i

     A.    Vegetation Effects	     x_2

          1.    Types of Exposure Effects		  X-2

               a.    Visible Foliar Injury Effects	   x-3
               b.    Growth and Yield Effects	]  X-6

                    1.    Open Top  Chamber Studies	 x-7
                   .2.    Greenhouse  and  Controlled
                         Environment Studies	 X-13
                    3.    Ambient Air Exposure  Studies...... X-14

          2.    Related  Vegetation  Issues	 X-20

               a.    Empirical  Models  Used to Develop
                    Exposure  Response Relationships	 X-20
               b.    Statistics  Used  to Characterize
                    Ozone  Exposures	 X-21
               c.    Exposure  and Response to
                    Peroxyacetyl Nitrate	  X-23
               d.    Economic Assessments  of  Agriculture....  X-24

     B.    Natural  Ecosystem Effects	  X-26
              Forest Ecosystems	  X-
                                               27

a.   Effects on Plant Processes	 X-29
b.   Effects on Growth	.* ] * ] x-31
c.   Ecosystem Responses:  The San
     Bernardino Study		, x-38

Interrelated Ecosystems	 X-40

a.   Aquatic Ecosystems	 X-40
b.   Agricultural Ecosystems	'.'.'.'. X-40

-------
                                  xiv
                                                              Page
       C.    Materials Damage ...............
                                           ***••*              ™
            1 .    Elastomers
            2 .    Textile Fibers and Dyes
                                                              x_43
                                                              X-45
            4.    Conclusion ............                        „ fc
                                       ...................... X-45
      D.    Effects  on  Personal  Comfort  and Well Being ....... X-45
      E.    Related  Welfare Effects  Information and Issues... x-46
            1 .   Air Quality Analyses ......................    x_49
            2 .   Crop Loss  Estimates ...........                v cn
                                             **•*•••••••••••• A~*D(J
            3 .   Averaging  Times ........
                                                              v _ .
                                                              X-54
                a.   NCLAN/CERL Reanalysis
                b.   New Studies ......... ;
           4 .    Forest Risk Assessment ...................... x_66
                             	 XI-1
      A.    Pollutant Indicator	
                                       •••••••••«•••••»..... X-L~"1
      B.    Form of  the Standard and Averaging Time(s)	 XI-3
      C.    Level of Standard	
                                      ••••«•••••••«......... XI—10
      D.    Summary  of  Conclusions.	                   VT ,
                                        ••••••«•«..„..»„.... XI—16
Appendix A.  Air Quality	
                                            • ••••••*••».»»., A**l
Appendix B.  Glossary of Pulmonary Terms  and  Symbols	 B-l
Appendix C.  CASAC Closure  Letter	
References

-------

-------
                                   XV

                              List of Figures
                                                              Page

  VII-i     croup Mean Decrements in 1-sec Forced Expiratory
            Volume During 2-hour Ozone Exposures with Different
            Levels of Intermittent Exercise	 .. .  viI-4

  VII-2     Fraction of Heavily Exercising Population
            Experiencing > 10% and > 20% change in ?-sec
            Levels EXplratory Volume Due to Various Ozone
                 s	*	  vii-n

  Vll-3     Fraction of Heavily Exercising Population
            Experiencing Mild and Moderate Symptoms
            Due to Various Ozone Levels	    VII-17

  VII~4     1™°^°^*^***™^ Population Ex-
                             Respiratory  Symptoms  Due to Various
                             	  VII-18

 VII-5      Benchmark Risk in St.  Louis for 1-sec  Forced
            Expiratory Volume Decrements  of > 10%  and
           U?!/._Undet Heavy Exercise,  for~Three
                                              Kulle,  and
                                             	  VII-43

 VII-6     Benchmark Risk in St. Louis for Chest Discomfort
           bymptoms (any and moderate/severe),  under Heavy
            (*v^1S*:,^°r Th5Se ExP°sure-Response Data Sets
            (Avol, Kulle, and McDonnell)	  VII-44

 VII-7     Expected Headcount (pulmonary function)

           S^™??"1?!:^ ?? 9 '3 Million Dumberlth *
                                        responding during
                                       1		VII-47

 VH-8     Expected Headcount (chest discomfort) Aggregated
           for Eight U.S.  Urban Areas With a  Total Popula-
           tion of 9.3  Million (number of heavily     P
           eV^T-r"! 01 net «-^^_1_  v.— — —   o-    -•  .     JT
                             responding during  the ozone
                             	 VII-48

X-l        Examples  of the Effects  of  Ozone on  the Yield of
           Soybean  and Wheat  Cultivars	              v
                                           *******•"••••*•• X~9
X-2
                                                             C-10
X-3
                                           Across Expert
                                           	 X-73

-------
                                  xvi

                                                             Page

 A-l    '  Correlations Among Short- and Long-Term Air Quality
           Indicators in MSAs (Using 2nd High)	 A-10

 A-2       Correlations Among Short- and Long-Term Air Quality
           Indicators in MSAs (Using ExEx)	 A-12

 A-3       Proportion (In Percent) of Urban Sites Exceeding
           Expected Number of days with an  8-Hour Daily Maximum
           Average  > .08 ppm for Five 1-Hour Daily Maximum
           Standards	     A-13

 A-4       Proportion (In Percent) of Urban Sites Exceeding
           Expected Number of Days with an  8-Hour Daily
           Maximum  Average > .06 ppm for Four 1-Hour  Daily
           Maximum  Standards		t m m A-14

 A-5       Proportion (In Percent) of Urban Areas Exceeding
           Expected Number of Days with an  8-Hour Daily
           Maximum  Average > .10 ppm for Three  1-Hour
           Daily Maximum Standards		 A-15

 A-6       Generalized  Relationships of the Current Ozone
           NAAQS and Three Alternative  8-Hour Averages	A-19

 A-7       Cumulative Frequency  Distribution of Three Peak
           Air Quality  Indicators	  A-28

 A-8    .    Correlations  Among Short-Term, Multiple-Peak,
           and Longer-Term Air Quality  Indicators  in  Non-
           Urban Areas	  A-30

 A-9        Proportion (In  Percent) of Rural/Remote  Sites
           Exceeding  Specified Expected Number of  8-Hour
           Daily Maximum Averages >  .08 ppm  for Three 1-Hour
           Daily Maximum Standards	  A-31

 A-10       Proportion (In  Percent) of Rural/Remote  Sites
           Exceeding  Specified Maximum Monthly 1-Hour Daily
           Maximum Values  for Three  1-Hour Daily Maximum
           NAAQS	  A_32

A-ll       Proportion (In  Percent) of Rural/Remote  Sites
           Exceeding Specified Three Month 8-Hour Averages
           Daily Maximum Three Month 8-Hour Averages For
           Three 1-Hour Daily Maximum NAAQS	  A-33

A-12       Proportion (In Percent) of Rural/Remote  Sites
           Exceeding Specified Second High 1-Hour Daily
          Maximum Values  for Three 8-Hour Daily Maximum
          Averages > . 08 ppm Standards	  A-34

-------
                       xvii


                                                   Page
Pvo-  ,.  ~  *—-7	/ of Rural/Remote  Sites
Exceeding Specified Number of Second High  1-Hour
Daily Maximum Values fnr TK>-«« «,,.,•	„_ _^, ,
Daily Maximum Values for Three Maximum Monthly Mean
1-Hour Daily Maximum Standards	?. . .V? A-35

-------

-------
                                 xviii

                             List of Tables
 Table                          Title
                                                             Page
 V-l        Study  Areas  Modeled  in Ozone-National  Exposure
           Model	  V_5

 v~2        Estimate  of  the  Cumulative  Number  of Heavy
           Exercisers in  the  8-Area Aggregation Population
           Exposed to One-Hour  Average Ozone  Concentration
           During the Ozone Season at  Heavy Exercise Under
           Alternative  Air  Quality Scenarios	  V-12

 V-3        Estimate  of  the  Cumulative  Number  of Person-
           Occurrences  of Heavy Exercise  in the 8-Area
           Aggregation  Population Exposed to  One-Hour Average
           Ozone  During the Ozone Season  at Very  Heavy
           Exercise  Under Alternative  Air Quality
           Scenarios	•	  V-14

 VI-1       Estimated Values of  Oxygen  Consumption and
           Minute Ventilation Associated  with Representative
           Types  of  Exercise	  Vl-io

 Vl-2       Prevalence of  Chronic  Respiratory  Conditions by
         •  Sex and Age  for  1979	  VI-12

 Vll-i      Key Human Studies Near  the  Current 1-Hour National
           Ambient Air  Quality  Standard for Ozone	  VII-7

 VII-2      Morphological  Effects  of Ozone in  Experimental
           Animals	•	VII—29

 VII-3      Effects of Ozone on Host Defense Mechanisms
           in Experimental Animals	  VII-34

VII-4      Percent of Heavy Exercisers Responding Under
          Alternative Air Quality Scenarios	VII-49

VII-5     Gradation of Response for Healthy Individuals
          Acutely Exposed to Ozone	  VII-55

IX-1      Effect of Ozone on Photosynthesis	  IX-4

X-l       Ozone Concentrations for Short-term Exposure that
          Produce 5 or 20 Percent Injury to Vegetation
          Growth Under Sensitive Conditions	 X-4

-------
                                  XIX

                       LIST OF TABLES (continued)


                               Htla                         Pace

 X-2       Summary of Ozone Concentrations Predicted to
           Cause 10 Percent and 30 Percent Yield Losses
           and Summary of Yield Losses  Predicted to Occur
           at 7-hour Seasonal Mean Ozone Concentrations of
           0.04 and 0.06 ppm.	   X-ll

 X-3       Ozone Concentrations at Which Significant Yield
           Losses Have Been Noted for a  Variety of Plant
           Species Exposed Under Various Experimental
           Conditions	   X-15

 X-4       Effects of Ambient Air in Open-Top Chambers,
           Outdoor CSTR Chambers,  or Growth and Yield of
           Selected Crops	  X-17

 X-5       Effects of Ozone on Crop  Yield as Determined
           by the Use of Chemical Protectartts	  X-19

 x~6       Continuum of Characteristic Ecosystem
           Responses to Pollutant Stress	  X-28

 X-7       Effects of Ozone Added to Filtered Air on the
           Yield of Selected Tree Crops	  X-34

 X-8       Potential Ambient Ozone Standards that would
           Limit Soybean Crop Reduction  to  5,  10,  15,  or
           20  Percent	  X-52

 x~9        Percentiles  and  Mean  Predicted Relative Yield
           Losses  Associated  with  Various Levels  of  the  Four
           Exposure Indices,  HDM2, M7, SUM06,-and SUM07,  for
           the  16  NCLAN  Cases	J  x_59

 X-10       Exposure Levels  Associated with  Predicted
           Relative Yield Losses of  5 to 30%  for  the  Four
           Exposure Indices, HDM2, M7, SUM06,  and SUM07,
           for the  16 NCLAN studies  	           X-60
X-ll
Forest Response Experts	 X-71
XI-1      U.S. Agricultural Welfare Benefits from Reducing -
          Rural Ambient Ozone (7-hr seasonal means) to 60,
          45, and 30 ppb for Three Alternative Benefit
          Measures	 XI-14

A-l       Cumulative Frequency Descriptive Statistics
          Associated with Peak and Multiple-Hour Ozone
          Air Quality Indicators in Urban Areas	 A-5

A-2       Cumulative Frequency Descriptive Statistics
          Associated with Various 8-Hour Ozone
          Air Quality Indicators in Urban Areas	 A-7

-------
                                   XX



                       LIST OF TABLES (continued)
                      Fre2uen°y Descriptive Statistics

           Associated with Longer-Term Ozone Air Quality
           Indicators in Urban Areas .........     uuanty
                                           ***•
                           Exceedin9  the  Current  Ozone  NAAQS

                                            Dailv Maximum
                                                            A-21



                                             Episodes  b*
A-6       Descriptive Cumulative Frequency Statistics

          Associated with Peak Ozone Air Quality
          Indicators .......
                                                            A-25


A-7       Descriptive Cumulative Frequency Statistics


                                                  Indicators

                                                            A-27

-------

-------
                                 XXI
                          Executive Summary

       This revised staff paper evaluates and interprets the
  available scientific and technical information that the EPA staff
  believe  is most  relevant to the review of primary (health)  and
  secondary (welfare)  national ambient  air quality standards
  (NAAQS)  for ozone (O3) and  presents staff  recommendations on
  alternative approaches to revising the  standards.   Periodic
  review of  the NAAQS  is a process  instituted to ensure  the
  scientific  adequacy  of air  quality standards and  is required by
  section 109 of the 1977 Clean Air Act Amendments.  The assessment
  in this staff paper  is intended to help build a bridge between
 the scientific review contained in the EPA O3  criteria document
  (hereafter referred to as CD) (U.S. EPA, 1986)., and the CD
 Supplement  (hereafter referred'to as CDS) (U.S. EPA, 1988)
 prepared  by the  Environmental Criteria and Assessment office
 (ECAO)  and the judgments  required of  the Administrator in setting
 ambient standards for O3.  Therefore,  the staff paper  is  an
 important element in  the  standards review process  and  provides  an
 opportunity for review by the clean Air  Scientific Advisory
 Committee  (CASAC)  and the general  public  on proposed staff
 recommendations before they  are presented to the Administrator.
 This staff paper has  been revised based upon comments received
 from CASAC and the public and upon staff analyses which are
available  for public review.

-------
                                xxii



      Ozone is a trace constituent formed in the atmosphere as a



 result of a series of complex chemical reactions involving both



 anthropogenic and natural hydrocarbons and nitrogen oxides,



 oxygen and sunlight.   At ambient concentrations often measured



 during warmer months,  03 can adversely affect human health,



 agricultural crops,  forests, ecosystems,  and materials.



 Interactions of 03 with  nitrogen oxides and  sulfur  oxides  may



 also  contribute to the formation of  acidic vapors and aerosols



 which might have direct  effects on human  health and welfare,  as



 well  as  indirect effects following their  deposition on surfaces.



 It  should  be noted that  new  evidence indicates  that co-exposure



 to  acidic  aerosols can potentiate response to O,.
                                           •     «J


      Annual average background  surface  O3 concentrations in the



 northern hemisphere generally range  between  0.03 and  0.05  ppm but



 are as low as  0.015 to 0.020 ppm in  the tropics  (U.S.  EPA,  1986,



 p.  3-80).   Stratospheric  intrusion is recognized as causing



 locally high 03 levels for periods lasting from minutes to  hours,



 but these  intrusions are usually worse in spring, fall, and



 winter.  In contrast,  during the photochemically active summer



 months intrusion is less common and  less severe.  Summertime



 hourly O3 levels have recently been reported to be as high as



 0.35 ppm in one of the nation's most heavily populated



metropolitan areas.   Daily daylight seasonal averages of 03 in



some rural areas have been reported to be 0.06 ppm and higher.

-------
                               xxiii
 Primary Standard
     The staff  reviewed scientific and technical information on
 the known and potential health effects of  O3 cited  in the  CD and
 the CDS.  The information  includes studies of  respiratory  tract
 absorption and  deposition  of O3, studies of mechanisms of  03
 toxicity, and controlled human  exposure, field,  epidemiological
 and animal toxicology studies of effects of exposure to O3 as
well as air quality information.  On the basis of this review,
the staff derives the following conclusions.
     1)    Inhaled O3 may pose health risks  as a result of  (a)
          penetration of 03 into various regions of  the
          respiratory tract and absorption  of O3 in  this  tract
          (b) provocation of pulmonary response resulting from
          chemical  interactions of  O3  along the respiratory
          tract, and (c)  extrapulmonary effects caused  indirectly
          by reaction  of O3  in the  lungs.
    2)    The risks  of adverse  effects  associated with absorption
          of 03  in the tracheobronchial and alveolar regions  of
          the respiratory tract  are much greater  than for
          absorption in the extrathoracic region  (head).
          increased exercise levels are generally associated with
         higher ventilation rates and increased oronasal or oral
          (mouth) breathing.  Greater O3 penetration  and exposure
         of  sensitive lung tissue occurs when individuals are
         heavily exercising.

-------
                          xx iv
 3)    Factors which have been demonstrated to affect
      susceptibility to 03  exposure are activity level and
      environmental stress  (e.g.,  humidity,  high
      temperature).  Those  factors which either have not been
      adequately tested or  remain  uncertain  include age, sex,
      preexisting disease,  nutrition,  and smoking status.
 4)    Major  subgroups of the population that may be at
      greater risk  to the effects  of 03  include:  (a)  any
      individual  exercising heavily during exposure to 03,
      particularly  those who are otherwise healthy
      individuals who may experience significantly greater
      than group  mean lung function response  to  O3  exposure,
      and  (b)  individuals with preexisting respiratory
      disease  (e.g.,  asthmatics and persons with  allergies).
      The data base identifying exercising individuals as
      being at greater risk to 03 exposure is much stronger
      and more quantitative than that for  individuals  with
     preexisting respiratory disease.   This  is due to  the
      large number of clinical studies investigating effects
     of O3 on exercising persons.
5)    The major effects categories  of concern associated with
     exposures to O3 include:
      (a)  alterations in pulmonary function
      (b)  symptomatic effects  (e.g., cough,  throat
          irritation)
     (c)  effects on work or  athletic  performance

-------
                     XXV
(d)
          aggravation  of preexisting  respiratory  disease
      (e)  morphological effects  (lung structure damage)
      (f)  altered host defense systems  (e.g., increased
          susceptibility to respiratory infection)
      (g)  extrapulmonary effects  (e.g., effects on blood
          enzymes, central nervous system, liver, endocrine
          system).
6)    An important source of applicable exposure-response
     information for a short-term standard is controlled
     human exposure  and field studies, which provide
     concentration-response relationships  between
     alterations in  pulmonary function and O3  exposure
     concentrations,  other important  sources  of  information
     for standard setting are  epidemiological  and
     toxicological studies.  Epidemiology  has  provided
     associations between ambient  O3 exposures and lung
     function decrements and aggravation of existing
     respiratory disease, but with greater uncertainties
     about the exposures involved than with controlled human
     exposure and field studies.  Animal toxicology data
    provide acute and chronic exposure effects information
    on increased susceptibility to respiratory infection,
    lung structure damage,  and extrapulmonary effects.
    Although human exposure,  epidemiology, and animal
    toxicology  studies  all  have limitations in assessing
    adverse  effects  and risk,  it  is the  weight of  evidence

-------
                               xxvi
     and integration of findings from all three disciplines which
     should be used in assessing health effects associated with
     exposure to 03.
     Based on scientific and technical reviews, CASAC comments,
and policy considerations, the staff makes the following
recommendations with respect to primary O3 standards:
     1)    Ozone should remain as the surrogate for controlling
          ambient concentrations of photochemical oxidants.
     2)    The existing form of the standard should be retained
          (i.e.,  that the NAAQS is attained when the expected
        -  number  of days per calendar year with maximum 1-hour
          average concentrations above the level of the standard
          is equal to or less than one).
     3)    The 1-hr averaging time of the  standard should be
          retained.
     4)    The range of  1-hour average 03  levels  of  concern  for
          standard-setting purposes is  0.08  to 0.12 ppm in
          concordance with CASAC comments  (CASAC,  1986,  1987,
          1988) comments.   This range is  based solely on 1-2 hour
          exposure data.
     5)    Because  there  is a  good health  effects data base
          available  on 1-2  hour exposures, the staff  concurs with
          the  CASAC  conclusion  (McClellan, 1989) that review of
         the  scientific basis  for  the l-hr  03 primary standard
         be closed  out.   With  this portion  of the  review
         complete,  and after considering CASAC1s views  on all

-------
                           xxvii
       issues,  the Administrator will be in a position to make
       a  regulatory decision  on how and when to best act on
       the  1-hour  standard.
 6)    In response to suggestions made  by  CASAC (1986,  1987,
       1988), staff  investigated the  potential  need  and basis
       for a longer-term  (6-8 hour) primary  standard.
      Although an emerging data base reporting significant
      lung function decrements and symptoms in subjects
      exposed to O3 for 6 to  8  hours has provided some
      evidence of effects below 0.12 ppm 03, staff concurs
      with  CASAC.s conclusion that "...  sucn information
      can better be considered  in the next review of the
      ozone standards." (McClellan,  1989).  'it is recommended  .
      that  EPA. continue review  of  scientific information on
      health effects of prolonged exposure  to  03.  Once these
      studies have been more  completely evaluated during the
      next CD review, the Administrator will be able to
      assess the need for development of a longer-term  O3
     primary standard.
7)    Further review and analysis also will be necessary
     before fully assessing the need for a separate standard
     to protect against chronic effects of 03.   Data on
     nasopharyngeal removal,  dosimetry modeling and health
     effects based on and chronic exposure of animals  will
     be used for future animal  extrapolation and  risk
     assessment of chronic O3 exposures.

-------
                              xxviii
Secondary Stands -ret
     The staff has reviewed the scientific and technical
information on the known and potential welfare effects of O3
cited in the CD and the CDS.  This information includes impacts
on vegetation, natural ecosystems, materials, and symptomatic
effects on humans.  Based on this review, the staff derives the
following conclusions:
     1)    The mechanisms by which O3 may injure plants and plant
          communities include (a)  absorption of O3  into leaf
          through stomata,  followed by diffusion through the cell
          wall and membrane,  (b)  alteratipn of cell structure and
          function as well  as critical plant processes,  resulting
          from the chemical interaction' of  O3  with  cellular
          components,  and  (c)  occurrence of secondary  effects
          including reduced photosynthesis  and growth  and  yield
          and  altered carbon  allocation.
    2)    The  magnitude  of  the 03-induced effects depends upon
          the  physical and  chemical  environment of  the plant, as
         well as  on  various biological  factors  (including
         genetic  potential, developmental age of plant, and
         interaction with  plant pests).
    3)   The weight of the recent evidence seems to suggest that
         long-term averages, such as the 7-hour seasonal mean,
         may not  be adequate indicators for relating O3 exposure
         and plant response.

-------
                           xxix
 4)   Repeated peak concentrations are the most critical
      element in determining plant response.  Exposure
      indicators which emphasize peak concentrations and
      accumulate concentrations over time probably provide
      the best biological basis for standard setting (See
      staff paper,  p.  x-50).
 5)    There is currently a lack of exposure-response
      information on  forest tree effects.   in addition,  there
      is  a  broad  range  of uncertainty  among scientists
      regarding O3 effects on forest trees.   Consequently
      there  is no consensus on  the most important averaging
      time for perennials or on  the precise role of  O3 vs.
      other pollutants  in causing forest decline.  Therefore/
      the staff concludes that a separate secondary  standard
     based on protection of forest trees is not warranted at
     this time.
6)    There appears  to be no threshold level below which
     materials damage will not  occur;  exposure of sensitive
     materials to any non-zero  concentration of O3
     (including natural background levels)  can produce
     effects if the exposure duration  is  sufficiently long.
     However,  the slight acceleration  of  aging processes of

-------
                                XXX
           materials which occurs at the level of the NAAQS is not
           judged to be significant or adverse.   Consequently, the
           staff concludes that materials data should not be used
           as  a  basis for  adequately defining an averaging time or
           concentration level  for the secondary standard and that
           the secondary standard should be  based, on  protection of
           vegetation.
     7)    Effects on personal  comfort and well-being,  as defined
           by  human  symptomatic effects,  have  been observed  in
           clinical  studies at  O3  levels  in the range of  0.12-0.16
           for 1-2 hour  exposures  and  at  somewhat lower levels in
           extended  exposure clinical  and epidemiological  studies.
           CASAC recommended that these effects be considered
          health effects in developing a basis for the primary
          standard  for 03.
     Based on scientific and technical reviews, CASAC comments,
and policy considerations, the staff makes the following
recommendations  with respect to secondary standards:
     1)    In consideration of the large base of welfare
          information attributing effects to O3 exposure  and the
          limited evidence which demonstrates welfare effects
          from exposure to ambient levels of non-O3 photochemical
          oxidants,  there  appears to be little evidence to
          suggest a  change in chemical designation from 03 to
          photochemical oxidants.

-------
                                 xxxi
       2)   Given the lack of effects data on forests and the
            preliminary nature of the Lee et al. (I988c) results
            regarding selection of the appropriate exposure
            statistic for crops,  the EPA staff concludes that it
            may be premature at this point in time to change the
            form of the  standard  and the averaging time.  it is our
            judgment that a  1-hr  averaging time  standard in  the
            range of 0.06-0.12  ppm represents the  best  staff
            recommendation that could be made to the  Administrator
            at  this  time  to close  out the review of the  scientific
            data.  This is consistent with CASAC comments  (CASAC,
            1987,  1988) urging EPA to  consider a l-hr averaging
           time and to act on the existing state of science rather
           than extend the review until a more exhaustive
           assessment is made of alternative averaging times.
           With this portion of the review complete,  and after
           considering CASAC's  views on all issues,  the
           Administrator  will be  in a position to  make  a
           regulatory decision  on how and when to  best  act on  the
           1-hr standard.
     Alternatively,  EPA  could continue  the standard  review  until
the information  on  alternative exposure  indicators has  matured.
Additional  time  for review and revision  of Lee  et al.  (M88c)
would allow the scientific community the opportunity to review
the alternative indicators and move toward a consensus regarding
selection of the most appropriate exposure indicator.  The

-------
                              xxxii
liability of this alternative is that it postpones action on the
secondary standard and thus fails to utilize new and existing
information to assess the most appropriate exposure statistic or
the protection afforded by the current l-hr standard.

-------
                               1-1

  Review of the National Ambient  Air Quality Standards for Ozone
        Assessment of Scientific  and Technical Information
                           Staff Paper

I.  Purpose
     The purpose of this staff paper is to evaluate and interpret
scientific information contained in the CD and the CDS and to
identify critical elements to be considered in selecting primary
(health based) and secondary (welfare based) national ambient air
quality standards (NAAQS) for ozone (03).   Staff conclusions and
recommendations will -integrate critical elements of the review of
standards with other factors such as averaging times,  form of
standards, and margin of safety considerations.

-------

-------
                                II-l
 II •   Background
      Since 1970 the Clean Air Act as amended has provided
 authority and guidance for the listing of certain ambient air
 pollutants which may endanger public health or welfare and the
 setting and revising of NAAQS for those pollutants.   Primary
 standards must be based on health effects criteria and provide an
 adeguate margin of safety to ensure protection of public health.
 As several judicial decisions have made clear,  the economic and
 technological feasibility of attaining primary standards are not
 to be considered in setting them,  although such factors may be
 considered in the development of  state plans to implement the
 standards (Lead Industries Association v.  EPA,  1980;  American
 Petroleum Institute v.  EPA,  1981).   Further guidance  provided in
 the  legislative history of the Act indicates that the standards
 should be set at "the maximum permissible  ambient air level .
 which will protect the health of  any (sensitive)  group of the
 population" (U.S.  Senate,  1974).   Also,  margins of safety are to
 be provided such that the  standards  will afford "a reasonable •
 degree of protection .  .  .  against hazards  which research has not
 yet  identified"  (U.S.  Senate,  1974).   In the final analysis,  the
 EPA  Administrator  must  make  a policy decision'in setting the
 primary standard based  on  a  judgment regarding  the implications
 of all the health  effects  evidence and the  requirement that the
 standards  provide  an adequate margin of  safety.
      Secondary  ambient  air quality standards  must  be  based  on
 welfare  effects  and  must be  adequate to  protect  the public
 welfare  from  any known  or anticipated adverse effects  associated
 with the presence  of  a  listed ambient air pollutant.   Welfare
 effects, which are defined in  section 302(h)  of  the Act,  include
 effects on vegetation, visibility, water, crops, man-made
materials, animals,  economic values  and personal comfort and
well-being.
     On April 30,  1971, the Environmental Protection Agency (EPA)
published in the Federal Register (36 FR 8186) primary and

-------
                               II-2
 secondary national ambient air quality standards (NAAQS)  for
 photochemical oxidants.   Both standards were set at an hourly
 average level of 0.08 ppm not to be exceeded more than once per
 year.
      In accordance with  the provisions of  sections 108 and 109 of
 the  Clean Air Act as  amended,  EPA reviewed and revised the
 criteria upon which the  original photochemical oxidants NAAQS
 were based.   On  February 8,  1979,  revised  standards were
 published (44 FR 8202) with the  following  changes:  (1)  changing
 the  primary  and  secondary standards to 0.12  ppm,  (2)  changing the
 chemical designation  of  the standards  from photochemical  oxidants
 to ozone (03), (3)  changing  to a daily maximum  standard rather
 than an "all-hours" standard,  and  (4)  changing the  definition of
 the  point at which the primary and secondary standards  are
 attained to  "when  the expected number  of days  per calendar year
 with maximum hourly average  concentrations above 0.12 ppm is
 equal to or  less than one."
     Several factors  were  cited  as a basis for  reivising the
 primary standard for  O3  in 1979.  Among these were:   (1)  an
 adverse health effect threshold  for oxidants could  not  be
 identified with certainty,  (2) 03  is a pulmonary irritant  that
 affects  respiratory mucous membranes as well as other lung tissue
 and  impairs  respiratory  function at levels as  low as 0.15  ppm for
 exercising persons  (Delucia  and Adams, 1977),  (3) evidence
 suggests  an  elevated  number  of asthma attacks when peak hourly
 oxidant  concentrations reach about 0.25 ppm  (Schoettlin and
 Landau,  1961), (4)  several studies show increased susceptibility
 to bacterial  infection in  laboratory animals exposed to 03 plus a
 bacterial challenge,  (5)  premature aging symptoms reported in
 animals and  (6) apparent synergistic effects on pulmonary
 function  from exposure to 0.37 ppm O3 and 0.37 ppm sulfur
dioxide.  EPA concluded that a primary standard of 0.12 ppm (one-
hour average) not to be exceeded more than one day per  year on

-------
                                II-3
 average would protect public health with an adequate margin of
 safety.
      EPA based its decision on the secondary standard on the
 limited information that was available on growth and yield
 reduction in commercially important crops and indigenous
 vegetation exposed to O3  under  field conditions.  These  studies
 indicated that growth and yield responses were related to long-
 term  (growing season)  exposure  of  plants.   On the basis  of this
 information and available air quality  data,  EPA concluded that
 there was no evidence indicating that  a significant  decrease in
 growth  and  yield would result from  long-term  average
 concentrations  expected to occur when  the primary standard was
 attained.   Consequently,  a secondary standard more stringent than
 the primary standard was deemed unnecessary on the basis  of
03-related yield reduction effects on vegetation.

-------

-------
                               III-l

 III.  Approach
     This  is  the  final  staff  paper provided  during current review
 of the NAAQS  for  O3; judgments contained herein are based  on
 scientific evidence reviewed  in  the  CD  and in  the  CDS  prepared by
 the ECAO.   This staff paper incorporates the results of  a  health
 risk assessment and includes  the results of  an exposure  analysis
 for alternative NAAQS.
     "Critical elements" have been identified  which the  staff
 believe should be considered  in  a review of  the primary  and
 secondary  standards.  Particular attention is  drawn to those
 judgments  that must be  based  on  careful interpretation of
 incomplete or uncertain evidence.  In such instances,  the  staff
 paper states the staff's evaluation  of evidence as it  relates  to
 a specific judgment, sets forth  alternatives that the  staff
 believe should be considered, and recommends a course  of action.
 To facilitate review, this paper is  organized  into sections as
 outlined below.
     Section IV provides an overview of ambient levels of  O3
 currently  being experienced in various portions of the United
 States.   This section is intended to set the stage for the
 remaining  discussion by identifying  the present air quality
 situation  so the reader can relate available health and welfare
 information to O3  levels occurring in the real  world.
     Section V summarizes results of the 03  exposure analysis.
The NAAQS Exposure Model (NEM) was used to estimate nationwide
human exposure to O3 given attainment of alternative standards.
     Section VI deals with elements related to the health effects
evidence examined in reaching conclusions regarding the primary
standards;  these include the following:
     •     most probable mechanisms of toxicity by which health
          effects  occur,

-------
                               III-2

      •    discussion  of  factors potentially affecting
          susceptibility to 03 exposure,
      •    description of the most sensitive population  groups  and
          estimates of the size of those groups.
      Section VII is a preliminary assessment of health  effects
and related health issues.  The section:
      •    identifies  health effects which have been attributed to
          O3 and other photochemical oxidant exposures,
      •    discusses health effects evidence used to develop staff
          judgments concerning which effects are most important
          for the Administrator to consider in reviewing and
          setting primary standard(s),
      •    describes the  health risk assessment for acute O3
          exposures,
      •    discusses issues related to health effects attributed
          to O3 and other photochemical oxidants.
      Drawing on discussions in Sections IV through VII, Section
VIII  identifies and assesses factors that the staff believes
should be considered  in  selecting a pollutant indicator,
averaging time, form,  and level of primary standards,  including
margin of safety considerations.   Staff recommendations also are
presented in this section.
      In Section IX, the  effects of O3  and other photochemical
oxidants on vegetation,  personal comfort, natural ecosystems,  and
man-made materials are identified.   The section:
      •    describes physiological and biochemical alterations
          associated with welfare effects which result from
          exposure to 03  and  other  photochemical  oxidants,
     •    identifies welfare effects of O3  and  other
          photochemical oxidants,
     •    discusses factors affecting plant response.
     Section X is a discussion of welfare effects to be
considered in selecting secondary standard(s).   The section:

-------
                               III-3
      •    describes the existing scientific evidence on welfare
           effects attributed to O3 and other photochemical
           oxidant exposures,
      •    describes new studies and analyses related to the issue
           of averaging times,  and
           identifies and evaluates scientific uncertainties
           related to welfare effects evidence in  addition to
           staff  judgments concerning which  welfare  effects are
           important  for the  Administrator to  consider  in
           reviewing  and setting  secondary standard(s).
     Drawing on discussions  in Sections IX  and X, Section  XI
identifies and assesses  the  factors  that the  staff  believes
should be considered  in  selecting averaging time(s), level(s) and
form of secondary standards,  staff recommendations also are
presented in this section.

-------

-------
                                IV- 1
  IV.
                                             and Rui
       This  Section provides a summary of ambient O3  air  quality in
 urban and  non-urban  areas.   More information on the topic appears
 in Appendix A.  The  data  base used  generally is 1985-1987 Storage
 and Retrieval of  Aerometric  Data (SAROAD) air quality data,* but
 data  are presented for earlier time  periods  if  1985-1987  data  are
 not available.  Urban areas  are  interpreted  to  be Metropolitan
 Statistical Areas  (MSAs, as  defined  by the U.S. Bureau  of Census
      T*as section a^o Briefly discusses the  concept of "natural
 03 background, -  and provides a general estimate of its ambient
 concentrations for different averaging times at ground-level.

 A.  Urban Areas
      There are 331 MSAs in the 50 states.  EPA staff in the
 Ambient standards  Branch has identified 224  MSAs (68%)  as having
 enough O3 air quality data  to ascertain  whether  or not  they
 exceed the  NAAQS.   of these,  101  areas (45%,  have more  than one
 expected  exceedance per year of the "current  03 standard  of o  „
 ppm.   Thus, slightly  less  than then  one-half  of  MSAs with
 sufficient  data exceed the  standard.   Approximately  loo  minion
 people less than  one-half  of the total  U.S.  population, Uve in
 these  101 metropolitan areas.   (However,  this  does not .ean that
 everyone ln these  areas is exposed to O3 concentrations at  or
above the standard.  See Section  V.)
     About 10% of the MSAs with sufficient data have a
characteristic highest concentration  (cue)' of 0.16  ppm o3  or
than
using more recent dat~fo
result in more exceedances and
here (and in Appendix *7   IS
incorporate these new data as

labelsTp?aced on
infrequently.  o»
                                          t™ coun*;1T   Thus,
                                         •      period— would
                                       ^       " tha"  resented
                                                     specific

-------
                               IV-2
 higher; over 5% (11)  of MSAs have a CHC above 0.18 ppm 03.   There
 is no clear temporal trend in 03  CHCs in most MSAs around the
 country,  although the 1984-1985 time period generally had lower
 levels than previous years and 1987-1988 had higher levels.  The
 trend in  expected exceedances in MSAs likewise is indefinite.
 Over the  1980s,  expected exceedances were relatively high in
 1980,  dropped in 1981-1982,  and jumped radically in 1983.
 Expected  exceedances  for 1984 and 1985 were like those in 1981-
 1982,  but increased again in 1987 and 1988.
      Generally,  maximum monthly mean concentrations for 1-  and 8-
 hour daily maximum averages  are in the range of  0.050-0.085 ppm.
 The  maximum 3-month mean of  8-hour daily maximum averages is in
 the  range of 0.045-0.065 in  most  urban areas of-the country.
 There  are statistically significant relationships among peak and
 longer-term mean indices of  O3  air  quality  in urban  areas.   These
 relationships  can  be  used to estimate  the impact that  alternative
 short-  term peak O3 standards will  have on  long-term average
 levels  of O3.
                            •

 B.   Non-MSA Areas
     Air  quality data indicate  that non-MSAs have  a  lower CHC
 than do MSA areas  and that the  CHC  often  is  associated  with  O3
 transported  into non-MSA  areas  after 4 pm.   On average,  a daily
represented by the second-highest 1-hour daily maximum O,
concentration monitored during an area's O3 season (generally
April through October).  Another CHC is the characteristic
largest daily maximum (CLDM),  usually represented by a
concentration from a mathematical distribution fitted to actual
air quality data.  For 03 air  quality data, a distribution often
used is the two-parameter Weibull.  Generally, the CLDM is the
1/n value from the Weibull, where n = 03 season length (in days).
     All three labels (CHC, design value, and CLDM)  are used in
this report, especially in Appendix A.  The CHC mentioned above
in the text is the "design value" version.  See also the footnote
on p. V-4.

-------
                                 IV-3
  maximum 8-hour average of 0.08 ppm is exceeded on an average of
  11 days during the O3 season in non-MSA areas.  This
  "exceedance rate" drops to 1 for a 0. 10 Ppm "outpoint '•
  concentration and slightly less than 1 for a 0.12 ppm cutpoint
  concentration.
       Longer-term daily maximum O3 averages in  non-MSAs usually
  are lower than those seen in MSA areas.   The maximum monthly
  means for l- and 8-hour daily maximum averages generally are
  0.067 ppm and 0.058  ppm,  respectively.   Means  as  high as o 118
  pprn and  0.092 ppm for these  two  averaging times have been seen  in
  non-urban areas,  however.  High  monthly  means  for either
  averaging time occur most often  in  May,  June,  and July.
       Longer-term  (>  month) air quality indicators are more
  closely associated with each  other  than  with short-term
  indicators,   in fact, correlations  among  long-term air quality
  indicators are higher than correlations among short-term
  indicators.  This is to be expected, since long averaging  time
 statistical measures are relatively less variable than short
 averaging time measures.  (Figure A-8 in Appendix A provides
 estimates of correlations among selected air quality indicators
 in non-urban areas . )
      Relationships can be developed among air quality indicators
 xn non-MSA areas that can be  used to estimate the  impact that
 attaining a standard  for one  averaging time will have on other
                                       - these are presented in
C.  Natural Ozone Background
     Ozone is a trace constituent of the atmosphere.  There is
controversy regarding ho* much of ambient 03 Monitored at ground-
level 1S natural and how much is produced from man-made
precursors.  Estimates of the natural component of 03 vary widely
« the literature,  and there is no standardized

-------
                                IV-4
 terminology regarding the concept of natural O3 background.3
 Even when a numerical estimate of background (however labeled)  is
 provided, rarely is the averaging time provided for the estimate.
      Based on a thorough review of this literature, it is obvious
 that natural 03 background .is a multidimensional and complex
 concept.  Natural O3 background concentrations  vary by geographic
 location, altitude and season.
      A working definition of natural O3 background is:
      n*-n,r-F?rna Par^iculaJ geographic area and  averaging time,
      natural 03 background is that constituent  of  "background o,»
      (0-j that cannot be  affected by manipulating anthropogenic
      emissions in an area)  which arises solely  from
      photochemically-reacted biogenic precursors and from
      stratospheric O-j transported downward into the area  (either
      directly or indirectly).
      Note that this definition is predicated upon  the concept
 that natural background  O3 cannot be  affected by manipulating
 anthropogenic emission sources.   Note also that the definition
 does not include some constituents of natural background  that are
 considered by many to be background 03.  Excluded  constituents   •
 are  (I)  trapped anthropogenic-based O3  downwashed  into an area
 due  to  breakup of  a morning  inversion,  (2)  nocturnal  O3 maxima
 due  to  downward mixing of O3-rich  air from  above the  inversion
 layer,  (3) transported O3 in  an urban or stationary source plume,
 and  (4)  anthropogenic-based O3 that is  formed and/or stored  in
 the  troposphere and subsequently  downwashed.
u  i.     ^act' a survey of the available literature that mentions
background 03 or natural O3 background-approximately 50 mentl°ns
articles—did not uncover a single rigorous definition of either
term!  Even the appellations used for the concepts vary greatly
in the relevant literature.  Examples include:  "baselinJ 0, "
"clean air background," "global background," "North American'
background," "Urban background," and "regional surface
background."  In addition, twelve other labels were used—all
without being defined.

-------
                               IV-5
      A reasonable estimate of the natural O3  background
 concentration near sea-level in the U.S.  for  an annual average is
 0.020-0.035  ppm.   This  includes a 0.010_0.015 ppm contribution
 (averaged  over time)  from the stratosphere and a 0.01  ppm
 contribution from photochemically-affected biogenic  non-methane
 hydrocarbons,   m addition,  another o.oi  ppm  is  possible  from  the
 photochemical  reaction  of  biogenic  methane.
     A reasonable  estimate of natural 03 background concentration
 for a l-hour daily maximum at sea-level in the U.S. during the
 summer is on the order of 0.03-0.05 ppm.
     These estimates are synthesized from the available
 literature, but rely most heavily on Altshuller  (1986)  and Kelly
et al. (1984) .                                                 *

-------

-------
                                V-l
  v-  Ozone EXPOSUT-O
       Analysis of population exposure under alternative NAAQSs
  requires that significant factors contributing to total human
  exposure be taken into account.  These factors include the
  temporal and spatial distribution of people and 03 concentrations
  throughout an urban area as people go through their typical daily
  pattern of life.   To the maximum extent possible with existing
  data,  this has been done in the NAAQS Exposure Model (HEM)   a
  simulation model  designed to estimate human exposure in selected
  urbanized areas under user- specified regulatory scenarios
  (Biller et al., 1981).   This chapter is a  summary of information
  provided in Paul  et .al.  (1986) ,  which has  been used  in
  development of. the risk  assessment described  in chapter VII.

      A.  overview of  the Ozone NAAQS  Exposure  Model
      The 03 HEM model partitions all land within a selected urban
 area into  large "exposure districts"  (Paul et  al., 1986) .  There
 are between three and fifteen exposure districts identified in
 the ten urban areas used in the 03 NEM analysis.  The number of
 districts identified is directly related to the number of
 monitors having valid air quality data in a study area
      People living within each exposure district, as  estimated by
 the U.S. census in 1980,  are assigned to a  single discrete point
 the population centroid.   The air guality !evel uithin each
 exposure district  is  represented by air quality at the population
 oentroid,  which is estimated for each hour  of  the year l^^
 monitoring data from nearby  monitoring sites.   Because pollutants
 ,n  the ambient air are generally  modified considerably when

 ''"
aluste                                             -
adjusted to account for five different microenvironments :
indoors at home, indoors other  (all other indoor locations)
inside a transportation vehicle, outdoors near a roadway  and all
other outdoor locations.  The air quality adjustments ar  le by
us ng microenvironmental transformation factors, as explained

-------
                               V-2
      Because degree of exposure and/or susceptibility to effects
 of pollution may vary with age, occupation, and intensity of
 exercise, the total population of each study area is divided into
 age-occupation (A-O)  groups.  Each A-o group is further
 subdivided into three or more subgroups.   A typical pattern of
 activity through the five inicroenvironments is established for
 each subgroup and an exercise level (high,  medium,  or low)  for
 each is also specified.
      Units of population analyzed by NEM  are called cohorts.
 Each cohort is identified by exposure district of residence,  by
 exposure district of  employment, or school attendance,  by A-O
 group,  and by activity-pattern subgroup.   During each ten minute
 period  of the day,  each  cohort is assigned  to a particular
 exposure district and a  particular microenvironment.   The
 assignment is based upon (1)  data regarding human activity
 patterns gathered by  EPA and university "time budget"
 researchers,  and  (2)  home-to-work transportation data  gathered  by
 the  U.S.  Census Bureau.   Using these  sources of information,  NEM
 simulates  ten minute  movements of cohorts through different
 districts  of  the  urban area  and through different
 microenvironments,  combines  the movements with  hourly  averaged
 air  quality data,  and accumulates  the resulting exposures over
 the  "O3 season."1

 B.   Air Quality Concentrations  in Microenvironments
     NEM requires that hourly  air quality estimates be available
 for  each microenvironment that  every cohort passes through.  The
 estimates are obtained by using  a simple linear  function relating
 outdoor air quality concentrations to microenvironment levels via
 a "transformation" relating outdoor-to-indoor levels and then
      The 03 season is that part of  the year for which o,
monitoring is undertaken as required by EPA regulations for
implementation of the O3 NAAQS (40 CFR 58,  Appendix D).  For
three urban areas that were modeled it is the entire year- for
the remaining 7 areas, it is April through October.

-------
                                V-3
  adding to thxs the pollutant concentration due to sources located
  in the mxcroenvironment itself.   For 03/  the  transformation  is
  essentially a multiplicative ratio derived from a review of  the
  03 exposure literature (Ferdo, 1985) .
  The relationship is:
       where     Vt = air quality in microenvironment m during
                       hour t
                 am,t  - hourly-averaged pollutant concentration'
                       due to sources located in the
                       microenvironment
                 bm   *  multiplicative ratio of the
                       microenvironment concentration value to the
                       monitored air quality value
                xt ^  -  monitor-derived air quality value for time


 Because no significant sources  of O3  were  identified  for any of
 the mxcroenvironments, am,t = o for all environments.  Estimated
 al ^1985 °btalned fr°m  the UteratUre are Ascribed  in. Paul  et

      The xt values are actual monitored values for the current
 (or "as is")  situation.  These monitored values are adjusted
  n  th                                               *- ^l
 in  the  study area  Dust meets the 03 NAAQS being analyzed.   By
 Is tLT^3 NAAQVS attained when a11 monitors in -
 less than one expected exceedance of the standard concentration
 value (currently it is 0.12 Ppm) in a year.  The
 analyse ls based on a "just attains" scenario,  where air  quality
 evels at the monitor currently having the highest number  oT   *
 expected exceedances are reduced mathematically  to where that
     "
  ta a  ot                                      '       «»
data at other monxtors in the study area  are  adjusted using a
non-lmear approach described in Paul  et  al.  (1986) .

-------
                                V-4
       C.  Simulation of Population Movement
            Population movement in NEM is based upon information
  gathered by the U.S. Census Bureau regarding householders' home-
  work commuting patterns  (Bureau of the Census, 1982).  The
  information includes MSA- specific data on the census tract
  level, which itself is based upon actual location information
  regarding the sampled population's home and workplace.  This
  census tract information is aggregated for exposure districts
  used in the NEM analysis to obtain district-to-district trip
  information for those cohorts that work.   (Other-wise, cohorts are
  assumed to stay in their home districts.)   Because of lack of
  travel data for non-work activities,  we assume that all shopping
  is done in each cohort's residential district.   The same
  assumption also is made regarding school-related activities.
       Three one-way commuting times are used to represent non-
  household-worker commute times:   20,  30,  and 40 minutes.   Most
.  workers fall into the 20 minute  commute time (representing the
  rather large interval of 0 to 24 minutes),  since the  average
  commuting time in the United States  is  about. 20 minutes.
  Housewives/househusbands are assumed  to have no commuting time.
       The number of different cohorts -explicitly modeled in an
  area is equal  to the  product of  54 cohort groups  times the square
  of the number  of districts identified  in each study area.   Thus,
  the number  of  cohort  groups  explicitly  modeled  varies between 486
  and 12,150  in  the ten urban  area sample.  (See Table V-l.)
       There  is  no trip information available  from  the  U.S.  Census
  Bureau regarding weekend  inter-district travel  in SMSAs.   (There
  is information available  regarding weekend recreational travel,
  but it  is not  locationally specific nor is it systematic.)
  Consequently,  all  cohorts are assigned  to their home  districts on
  weekends  in NEM analysis.

       D.   Study  Areas Modeled in  Ozone National Exposure Model
       As mentioned,  ten urban areas are  used to model  O3 exposures
  explicitly.  Results  from these  ten areas will be extrapolated to

-------
            Table V-l

STUDY AREAS MODELED IN OZONE-NEM
Study
Area
Name
•"•' • i - —
Chicago
Denver
Houston
Los
Angeles
Miami
New York
Philadelphia
St. Louis
Tacoma
Washington,
DC
MSAs Included Study Area
(in whole or Population
	 ln part) Modeled (ii
Aurora-Elgin, IL
Chicago, IL
Gary. IN
Joliet, IL
Lake County. I L
Boulder, CO
Denver, CO
Houston, TX
Anaheim, CA
Los Angeles, CA
Riverside, CA
San Bernardino, CA
Fort Lauderdale. FL
Miami, FL
Middlesex, NJ
Nassau-Sulfolk, NY
Newark. NJ
New York, NY
Stamford. CT
Philadelphia. PA
St. Louis. MO
Tacoroa. WA
Washington, OC
7.48
1.54
2.54
10.22
1.55
13.62
4.36
2.20
0.58
2.88
   Total
 Population
 of Included
JtSAs
1 -'  —•—..

      7.82
      1.62


     2.74

    10.97




     2.65


   13.85
    4.72

    2.38

    0.46

    3.25
                        2nd Highest
                       Daily Maximum
                      03 Concentration
                             0.17
                            0.14


                            0.19

                            0.37
                           0.12


                           0.25
                          0.20

                          0.18

                          0.14

                          0.14
Number
of
Exposure
Districts
9
6
7
15
4
10
7
8
3
8
Number
of EPA
Cohorts Region
Modeled Number^
4.374 5
1.944 a
2.646 6
12.150 g
864 4
5.400 it2
2.646 3
3.456 5.7
486 10
3.456 3
                                                                                            r

-------
                               V-6
 the nation's urban population as a whole to obtain aggregated
 national exposure estimates.   The study areas  and their base
 population,  along with other  pertinent area information,  appear
 in Table v-1.
      The ten areas vary greatly in geographical  location,  03
 "design  value",2  population size (both modeled and total MSA),
 and number of exposure districts included.   The  areas  were
 selected to  obtain as  widely  representative a modeling data base
 as possible  given the  overall need for monitoring data
 completeness in an area.
      For instance,  study area populations modeled vary from
 Tacoma,  with a population of  almost 580,000, to  New York,  with  a
 population of  about 13.6 million people.  Design value
 concentrations vary from the  current standard level of 0.12  ppm
 to 0.37  ppm—the  highest design  value  included in EPA's
 regulatory data base within the  last five years.
      The number of  exposure districts  in study areas varies
 between  3—the minimum  thought necessary to  adequately capture
 the variation  in  03 air quality for an,area—and  15.  The number
 of exposure  districts used in an  area  directly translates  into
 the number of cohort groups in the area that are  tracked through
 time  and space in 03-NEM.  Each one of these cohort groups
 experiences  a different pattern of air quality as  its weekday
 activities are simulated.  Thus,  in Los Angeles  over 12,000
 different patterns were modeled.
      One limitation of the exposure assessment is that  it does
 not provide  good coverage of the New England area  (only Stamford,
 CT is included).  This is a limitation worthy of note since  the
     «
      A design value is that measured air quality concentration
value in a MSA that must be reduced to the O3 standard level to
ensure that the area meets the current 03 NAAQS formulation of <1
expected exceedances of 0.12 ppm daily maximum 1-hour average.
The value shown in Table V-1 is the second-highest 1-hour daily
maximum concentration in the O3 air quality data base for the
year modeled.  See the footnote on p. IV-2 for additional
information regarding design values (and other CHCs).

-------
                               V-7
 CT is included).   This is a limitation worthy of note since the
 New England area has high population densities and high 03
 levels.   The staff was unable to adequately model the New England
 area because of (1) the adequacy of monitoring data suited for
 exposure assessment and (2)  U.S. Census Bureau data—both
 population and transportation—are difficult to use in New
 England  due to the political/spatial classification system used
 there.   No other  MSA in New England could be explicitly modeled
 for these two reasons.

 E.   Exercise Modeling in Ozone NAAQS Exposure Modeling
      Because dose received by a person exposed to an air
 pollutant is highly dependent upon her or his ventilation rate,
 exercise level is an important consideration in exposure
 modeling.   In O3-NEM, four exercise  levels  are considered.  They
 are listed here,  along  with  their associated ventilation rates
 (in units of liters per minute,  or 1/min):
      1".   low exercise,  25  L/min or less
      2.   medium exercise,  26-43  L/min
      3.   heavy exercise, 44-63  L/min
      4.   very heavy exercise,  64  1/min or higher
 Two broad  cohort  group  distinctions  are made in 03-NEM with
 respect  to exercise:  exercisers  and  non-exercisers.   Exercisers
 are further divided into those who exercise  heavily  and  those  who
 exercise very heavily.  Exercisers are those cohorts  who
 participate in any  activity that  requires a  ventilation  rate
 greater than  43 L/min—i.e., the  last  two exercise categories.
 Heavy exercisers participate in an activity  level requiring more
 than 63 L/min  of ventilation.
     Exercise  level designations used  in NEM are roughly adjusted
 for age and body size.  For example, a baby's  "thrashing around"
 is heavy exercise for the  0 to 6 month old cohort, but the same
type of activity in an adult cohort  is low exercise.  Exercising
cohorts are emphasized in 03-NEM because exercisers,  per se,  may
be "sensitive"  (react more) to 03 exposure.   In fact, 35 of the

-------
                               V-8
 54 cohort groups used in NEM are exercisers.   In most of the ten
 study areas modeled,  however,  these 35 exercising cohorts
 constitute less than  40% of a MSA's population.   Twenty cohort
 groups (included in the 35)  participate in very  heavy exercise,
 but contain at maximum approximately 7% of any MSA's population.
 Thus,  both exercising and cohorts are over-sampled in O3-NEM.
      The amount of  time spent  in exercise  for any cohort is quite
 small.   The largest portion of any cohort  group  undertaking heavy
 exercise is 6.2% for  cohort #3,  which consists of children 1 to  2
 year old.   (This cohort constitutes only about 1.0-1.5% of a
 MSA's  total population.)   The  proportion of time spent in heavy
 exercise by people  in cohorts  that undertake  such exercise is
 less than 2%,  on average.
     The largest portion of  any  cohort  group  undertaking very
 heavy  exercise is 100%  for cohorts #52-54,  consisting  of male
 outdoor  workers.  Less  than  2% of  any study area's population  is
 included in this occupational  grouping.  The  usual proportion  of
 exercising  cohort population participating  in very heavy exercise
 is  less  than 1%.
     The  combination  of  small  exercising fractions and  small
 amounts  of  time  spent in heavy or  very heavy  exercise results  in
 a very small fraction of total population-time devoted  to
 exercise.   For one modeled study area, New  York,  the amount of
 total population-time spent in exercise is:
     1.   0.9% for heavy exercise  (undertaking exercise  at a
          level  of 44-63 L/min).
     2.   0.002% for very heavy exercise (undertaking exercise at
          a level of 64 or higher L/min).
The other modeled MSAs have similar, but not  identical,  fractions
of total population-time spent in exercise.

F.   Eight-Area Aggregated Estimates of Population Exposure to
     Alternative Ozone Standards
     Although ten urban areas were modeled using  O3-NEM, only
eight areas were modeled for the "headcount" risk assessment that

-------
                                V-9
  x.  described  in  Section VII-B.3  The total population  included
  in  the  aggregation  is  25.9  million people, and  the  total  number
  of  one-hour exposure occurrences for this population is 125
  billion person-hours annually.
      Since the lung function/symptom health risk assessment
  currently focuses on heavily exercising individuals as a
  "sensitive population" with respect to O3 exposure,  we present
  exposure estimates for only that group.  Thus, the exposure
  estimates that follow are for people who are exercising at a
 ventilation rate of between 44 and 63 L/min.   Because only a
 small portion of total population- time is spent in this exercise
 category,  the number of people in the aggregation drops to 9 3
 million and the one-hour exposure occurrences  drops  to  2.6
 billion person-hours.   These are the  "base numbers"  for the
 exposure estimates contained later  in this chapter.
      In  any NEM analysis,  three  different  indicators are used to
 estimate exposure of people  to various  levels  of air pollution
 One  unit is  "occurrences of  exposure:"  the number of times a
 given level of pollution is  experienced by one individual.   if 30
 people experience a  pollutant level of l ppm which remains  steady
 over a 3-hour  period, population exposure can be expressed  as  30
 occurrences of exposure  for a 3-hour averaging time or  90
 occurrences for a  1-hour averaging time.  A second indicator of
 population exposure is "people-exposed."  This is simply the
 number of people who experience a given level of air pollution
or higher, at least one time during the time period of analysis.
The third indicator of exposure used in NEM is  "people at peak
exposure."  This is the number of people who experience  their

-------
                               V-10
 highest pollutant level within a given concentration interval.
 Examples of the first two output measures will be provided in
 this Section.
      Four alternative air quality scenarios are modeled.  The
 first is the "as is" situation,  which uses recent monitored air
 quality data to represent O3  concentrations in the outdoor-other
 microenvironment.  (This concentration is then modified by the
 microenvironmental factor to  estimate 03 levels in the other
 microenvironments.)   Air quality for the "as is" case comes from
 the 1983-1985 time period,  using whichever year has the most
 complete data base.
      The remaining air quality scenarios represent the
 hypothetical situation when a national ambient air quality
 standard for 03  is  just  attained  in  an urban  area.   This
 situation is simulated by adjusting  current air quality data  so
 that  the worst monitor in the urban  area has  a CHC4  equal  to  the
 NAAQS concentration  level.  The  adjustment procedure is complex
 and nonlinear.   (For instance, peak  hourly concentrations  are
 adjusted more-absolutely and  relatively—than those  near the  mean
 of  the  "as  is" distribution.)  It utilizes  regression  analyses  of
 parameters  of  the Weibull distribution fit  to each valid monitor
 in  the  urban area.   For  more  information  regarding the  air
 quality  adjustment procedure  used to simulate a  just-attaining
 situation see  Paul et  al.  (1986).
      The CHC in this case is the characteristic largest dailv
maximum  (CLDM) version.  The worst monitor in most MSAs is
located  in a downwind suburb of a central city.  Most of the
people in the MSA never experience the high O3 levels seen at
that monitor, since they may not travel into that exposure
district as they go about their daily activities.  Even people
living in that district do not experience the high concentrations
seen at the monitor, because (l) high O3 values often occur in
the mid- to late-afternoon when many district residents are at
work in other less polluted districts, and (2) most people spend
most of their time indoors, and do not directly experience hiah
outdoor 03 concentrations.   When the worst monitor in an area
attains an alternative O3 NAAQS, all other monitors experience
better air quality bv definition.                     v

-------
                                V-il
       Three alternative O3 NAAQS are simulated:  0.08, o.io  and
  0.12 ppm.  The latter value, of course, is the current standard
  level of the O3 NAAQS.   It should be recognized that we are not
  concerned in our exposure analyses about how or when an
  alternative 03 NAAQS is attained.   That is the concern of other
  analyses which OAQPS and other EPA offices undertake: especially
  the regulatory and benefits analyses.   For O3  exposure analyses
  purposes,  it is sufficient to simulate  the just-attaining
  situation without  being concerned  about how, when,  or even if
  that situation will  occur.
       Estimates appear in Table  V-2  of the  cumulative  numbers  of
  people  (in millions) who experience increasing l-hour O3
  exposures as they  engage in heavy exercise.  These  estimates  are
  based upon using "best  estimate" microenvironment factors   The
  first column shows the  03 concentration that is equaled or
  exceeded, and the other  four columns show cumulative  exposure
 distributions for the four air quality scenarios previously
 discussed.  Because all people experience the same number of
 exposures at or above o.O ppm, the entire aggregated population
 base of 9.3 million exercise is shown for all scenarios.
      The overall pattern of data shown in Table V-2  is cordon to
 most of the exposure results that follow.  Basically,  that
 pattern is:
      1.    there is  a long "tail" to the  current situation
           scenario,  in that there are tabular entries  for the
           higher O3 concentration "cutpoints,"  even  up to  0.34
           ppm.
      2.    There is  a  short tail  to the remaining scenario
          distributions,  with very small  tabular entries at  or
          above  the O3 concentration used for the scenario
          standard.
What this pattern implies  is that attainment of  any alternative
NAAQS investigated results in a dranatic reduction in  peak 0,
eZsTr ,F°VnStanCe' the -«»*• «* th. number of persons
exposed dunng heavy exercise to 03  levels 40.12 pp.  drops to 0 1

-------
                                   V-12
                               Table V-2




    iS2X5nf BX!!, ^H!^II!(!._NUMBER  OF HEAVY EXERCISERS
                          (millions of people)
03 Cone.
Equaled
or Exceeded
(ppm)

0.361
0.341
0.321
0.301
0.281
0.261
0.241
0.221
0.201
0.181
0.161
0.141
0.121
0.101
0.081
0.061
0.041 -
0.021
0.001
0.000
• *ii \4
The
"As Is"
Situation

n
\J
*
*
0.1
0.1
0.3
0.5
0.7
1.3
1.6
2.6
3.3
4.0
5.2
6.9
8.6
9.1
9.3
9.3
9.3
ua i ( ujr oueridrios
Alternative
0.12



0
0
0
0
0
0
0
0
0
0.1
0 3

NAAQS
0.10

0
0
0
0
0
0
0
0
0
0
0
0
0
+
w * *J
2.2 0.9
6.5 4
8.7 7
9.3 9
9.3 9
9.3 9
.9
.9
.3
.3
.3

(in ppm)
0.08

0
0
0
0
0
0
0
0
0
0
0
0
0
0
*
2.0
6.9
9.1
9.2
9.3
—— — i^— _ ___ _

*Fewer than 50,000 people.

-------
                                V-13
  million with attainment of a 0.12 ppm NAAQS,  as compared to 4.0
  million people under "as is" O3  air quality conditions.  Attaining
  a tighter 03  NAAQS/  either 0.10  or  0.08  Ppin/  reduces  the number
  of people- occurrences >0.12 ppm to zero.
      A  different  type of exposure statistic is  depicted  in Table
  V-3.  The subject of the Table is the estimated number of people-
  occurrences of exposure to various  O3  levels  as  they  undertake
  heavy exercise.   The universe for this aggregation  is 2.6 billion
  person-occurrences.   AS can be seen,  the number  of  heavy
  exercising person-occurrences of  exposure >.12 ppm  drops
  radically from the "as  is"  situation with attainment  of  one of
  the alternative standards,   it should  be realized,  however, that
 the relative proportion of heavy  exercising person-occurrences
 estimated for O3 exposure levels  at or above 0.12 ppm is fairly
 small,  even for the "as is" situation.  For example,  it  is 1.7%
 of total heavy exercising person  occurrences at >.i2 ppm and is
 0.2% at >o.20 ppm.  With attainment of' a 0.12  ppm 03 NAAQS in all
 areas,  the proportion of heavy exercising person occurrences of
 03 exposure above  that concentration would  only  be 0.003%.

 G.  Caveats and Limitations
      A number  of caveats must be  acknowledged  concerning  the o -
 NEM results.   These must be recognized by the  reader in her or
 his interpretation of exposure estimates.  Probably  the most
 important  caveat is that there is  considerable uncertainty
 concerning a number of important  inputs to 03-NEM, especially
 those regarding human activity and exercise  patterns
 Uncertainty also exists  regarding  (!) predicted O3 concentrations
upon attainment of  alternative O3  NAAQS, (2)  microenvironment
 factors relating outdoor-to-indoor concentrations, and  (3)
extrapolation of urban area-specific exposure estimates to the
nation as a whole.
     Some of these uncertainties are addressed by use of the
"low, "best,', and "high" estimates, but not all sources of

-------
                             Table V-14
      Estimate of the Cumulative number of Person-Occurrences
  of Heavy Exercise  in the 8-Area Aggregation Population Exposed
     to One-Hour Average  Ozone During the Ozone  Season at Verv
      Heavy Exercise Under Alternative Air Quality Scenarios

                  (millions of person-occurrences)
 O3 Cone
                               Air Quality Scenario
Equaled
or Exceeded
(ppm)
0.361
0.341

0.321

0.301
0.281
0.261
0.241
0.221
0.201
0.181
0.161
0.141
0.121
0.101
0.081
0.061
0.041
0.021
0.001
0.000
The
"As Is"
Situation
0
*

*

0.1
0.3
0.6
1.4
2.7
4.3
6.8
14.4
24.9.
45.2
67.4
160.9
343.5
758.0
1,622.4
2,348.3
2,610.3
Alternative NAAQS
0.12 0.10
0




0
0
0
0
0
0
0
0
0
0.1
0.7
10.5 1
77.8 39
439.2 352
1,655.7 1,617
2,348.3 2,348
2,610.3 2,610
0




0
0
0
0
0
0
0 '
0
0
Ci
*
.7
.0
.6
.0
.3
.3
(in ppm)
0.08
n
w



n
\J
0
o
n
w
n
\J
o
w
0
o
o
V
o
o
*
8.4
209.0
1,544.9
2,326.1
2,610.3

*Fewer than 50,000 person-occurrences

Source:  Summarized from O3 NEM Results

-------
                               V-15
 uncertainty are  so addressed.   The Ambient Standards Branch (ASB)
 will attempt to  explicitly analyze the uncertainties mentioned
 above  in  future  O3-NEM modeling efforts, if contract funds  are
 available for that purpose,  in addition,  due  to  increasing
 interest  in  multiple-hour  03 exposures  it  is very likely that  the
 current O3-NEM will not be' used in the  future as it  only provides
 hourly exposure  estimates.  ASB's plan  is  to revert  back to the
 CO-NEM and update  it to handle 10-minute or "probabilistic"
 activity patterns.  Thus,  future O3 exposure estimates may be
very different than those reported here.

-------

-------
                               VI-1
       Factors Relevant to Review
       Ozone
       Of primary concern in this section are the health effects
  associated with levels of O3 and other photochemical oxidants
  that are observed in the ambient air of the United states,  of
  the photochemical oxidants, only 03 has been reported' to exist in
  cities at sufficiently high concentrations to be of significant
  concern for human health,   since other photochemical oxidants
  such as hydrogen peroxide  (H2O2) and perdxyacetyl nitrate  (PAN)
  are known to produce health effects only at concentrations much
  higher than those found in the ambient air,  this section- will
  focus  on the mechanisms of toxicity for 03  and documented
  evidence of pulmonary and  extrapulmonary effects of O3 .  This
  approach will be  based on  both available human and  animal  data.
  Mechanistic studies  are used to support the possibility of human
  health   ffects
                                                       ^
 data.  in addition, factors affecting susceptibility and
 potentially susceptible groups will be discussed.

      A.  Ozone Absorption and Mechanisms of Effects
      Ozone enters the human body through the respiratory system
 During inhalation and exhalation, 03 is  removed from the air by'
 airway surfaces through a process of dissolution and chemical
 reaction    The rate,  amount,  and sites of O3 uptake and  removal
 largely determine the extent  to which exposure of an individual
 to a particular concentration of O3  will  evoke adverse health
 streets.
      Numerous  studies  help  to  explain  the quantity and sites of
 03 uptake ln mammalian respiratory tracts.   Nasopharyngeal
 removal studies  reviewed  in the  CD  (p. 9.4)  suggest that: !,  the
 fraction of 03 uptake depends inversely on the flow rate  2
 tracheal and exposure chamber concentrations of 03 are positively
 correlated (Yokoyama and Frank, 1972; Moorman, et al., 197 ; Ind
o   >"    f"d          3> ™*°*™^ S° percent of inh   d
°3 0 72, for dogs and > 50% for rabbits)  im removed in the

-------
                              VI-2
 nasopharyngeal region of animals exposed to between 0.1 and 0.2
 ppm 03, thereby indicating the role of the nasopharynx in
 removing 03 before it reaches the more sensitive lung tissues.
 Only one study measured 03 uptake in the lower respiratory tract
 and reported 80 to 87% uptake by the lower respiratory tract of
 dogs (Yokoyama and Frank, 1972).  There are, however,  no
 published data currently available for human nasopharyngeal or
 lower respiratory tract absorption.
      Inhaled O3 not absorbed in the nasopharyngeal region can  be
 deposited along the entire respiratory tract.   Penetration into
 the lower respiratory tract increases in rats  as inhaled O3
 concentrations increase from 0.2 to 0.8 ppm (Dungworth et al.,
 1975).   Models have been developed to estimate lower respiratory
 uptake  (Miller et al.,  1978b,  1985; McJilton et al.  1972).  One O3
 deposition model  has predicted  that similar patterns of O3
 deposition occur  in humans,  rabbits,  and guinea pigs and that  the
 junction of the conducting  airways  and  gas  exchange  region
 receives the maximal dose (Miller  et  al., I978b;  1985).   This
 prediction is consistent with location  of O3-induced lesions as
 indicated by pathology data  from several  species  of  03-exposed
 animals.   Recently  the Miller dosimetry model  has  been  used to
 estimate sensitivity to  lower respiratory tract secretions  and
 exercise of  the uptake of O3 in the human lung  (Miller et al.,
 1985).   The  Miller  dosimetry model  is also  an  important tool in
 assessing  animal toxicology data.   Chapter  9 of the CD provides a
 detailed discussion  of the use and  comparison of O3 dosimetry
 models.
     Biologically important functional groups which react
 relatively rapidly with 03 include the alkenic  groups (carbon-
carbon double bonds), amino groups, and sulfhydryl groups.
Alkenic bonds are found in essential fatty acids and
polyunsaturated fatty acids (PUFA)  which are important components
of the lipids in cell membranes.  Oxidation of  amino and
sulfhydryl groups can result in protein denaturation with

-------
                               VI-3
  concomitant loss of structural and functional integrity for the
  affected protein.
       Although there is general agreement that the oxidative
  properties of O3 cause toxic effects,  the precise molecular
  mechanism of toxicity remains unclear.  Several theories have
  been proposed:
       1.   Oxidation of polyunsaturated  lipids contained mainly in
  cell membranes;
       2.   Oxidation of sulfhydryl,  alcohol,  aldehyde,  or amine
  groups in low molecular weight compounds  or proteins;
       3.   Formation of  toxic  compounds  (ozonides  and peroxides)
  through reaction with  polyunsaturated  lipids;
       4.   Formation  of  free radicals, either directly or
  indirectly, through lipid peroxidation; and
       5.   injury  mediated by  some pharmacologic action,  such as
 via a neurohormonal mechanism or release of histamine  (CD, p.
 9"~ I'l 3) .
      Molecular targets  (e.g., carbon - carbon double bond
 sulfhydryl groups)  are shared across all species.  Therefore  if
 03  initiates mechanisms such  as those  listed above in  an animal
 Which ultimately result in lung structura! damage, there is
 reason for concern that the  sane mechanisms may be activated in
 humans with similar outcomes.  This assumption is limited by
 differences in human and animal dosimetry  which produce different
 doses of 03  for equivalent exposures.   Also  defense and repair
 mechanisms are likeiy  to provide quantitative differences in the
 display of toxicity resulting from  equivalent doses
 Nonetheless, mechanistic information strongly supports  the
 hypothesis that equivalent effects  may  occur in humans  and
 animals, albeit not  necessarily  at  the  same  concentrations.

B.  Factors Affecting Susceptibility to Ozone
     There are numerous factors which could affect susceptibility
to 03 exposure and to alter physiological responsiveness   These
factors include such individual characteristics as age   sex

-------
                              vr-4
 smoking status, and nutritional status.  In addition,
 environmental stresses and exercise level during exposure can
 influence the extent and level of an individual's response to O3
 by increasing the volume of inhaled 03 and by promoting
 penetration deeper into the lungs.

      1.   Age
      It  has been postulated that age-related changes in lung
 growth and development are among the factors responsible for
 altering individual susceptibility to 03;  thus,  extent of damage
 may depend on the stage of lung development.   Support for this
 hypothesis has come from several human studies,  but definitive
 evidence is unavailable at this time.
      Several clinical,  field,  and epidemiology studies have
 provided evidence of pulmonary function and/or symptomatic
 response in younger persons exposed to less  than 0.15 ppm O3
.(McDonnell et al.,  I985b,c;  Avol  et al.,  I985a,b;  Lebowitz  et
 al.,  1982,  1983;  Lebowitz,  1984;  Lippmann  et  al.,  1983;  Lioy et
 al.,  1985;  Bock  et  al.,'l985;  Spektor  et al.,  19S8a).   In
 addition,  symptoms  have  been reported  for  children  exposed to
 oxidant  concentrations of  0.10  ppm  and  higher  (Makino and
 Mizoguchi,  1975).   Although  it  has  been suggested that older
 individuals  may  be  less  susceptible  to  O3 exposure than younger
 persons,  age differences in response to O3 have not yet been
 fully investigated  in controlled human  studies.  Because  children
 tend to  exercise outdoors  in the summer more often than adults,
 there may be more reason for concern about effects of  O3 in
 children.
     Reisenauer et al. (1988) find that only female subjects  show
 a rise in total respiratory resistance  (R,,) when exposed to 0.3
ppm 03 during exercise.  Drechsler-Parks et al. (1984) also
report no significant differences between older males  and females
exposed to higher O3 levels (0.45 ppm)  for 2 hours during
 intermittent exercise (25 L/min) and suggested that older
individuals may be less responsive to 03 than younger persons.

-------
                               VI-5
  Bedi et al. (1988) substantiate these results in finding that
  older subjects, as a group, are less responsive to 03 than
  younger subjects.   Thus, while there continues to be no
  controlled human experimental evidence of additive effects from
  exposure to other  pollutants combined with O3  or of differential
  response due to sex,  preliminary evidence suggests possible age-
  related differential  response.
       The few animal studies that have addressed  the potential
  susceptibility  of  the young have shown a  few differences  between
  neonates and young adult animals.  Generally,  however,  the  very
  young rats were not more responsive  than  the young  adult  rats
  (CD, p.  12-36).
      Conclusions regarding  age as a  susceptibility  factor in   '
 humans will remain uncertain until the results of additional
 research are available (CD, p. 12-35).  while more  information is
 needed to fully understand age-related differences, it should be
 noted that recent studies by Reisenauer et al.  (1988) and
 Drechsler-Parks et al. (1984) provide evidence of reduced
 responsiveness in older adults.
      2.   Sex
      Due to the small  number of  female subjects in controlled
 human studies,  no  definitive conclusions  can  be drawn  regarding
 sex  differences in response  to o3 exposures.  Those human studies
 which have  evaluated lung  function differences  report  that  forced

             0      f°r  1— °
                                                            more
            ,
           than in males for similar exposure concentrations and
exercise levels, but the results are not conclusive. Only four
studies gave enough information for limited comparative
evaluation (Horvath et al., 1979; diner et al., 1983; Delucia et
al., 1983,  Launtzen and Adams, 1985).   Results of these studies
suggest that lung function of women may be affected more than
et
et
                                                        '
              report no differences in male and female response

-------
                              VI -6
 when exposed to 0.3 ppm, 0.6 ppm, N02 separately or in
 combination.
      Most 03 research on animals has been performed using only
 males.  When females were used, sex comparisons were not usually
 made.  An exception is the study by Graham et al. (1981) which
 reports increased pentobarbital-induced sleeping time in all
 females but not in male mice or rats after 5 hour exposure to l.o
 ppm 03.   Although reasons for sex differences have not yet been
 elucidated,  these differences have been suggested in some
 research and thus further investigation is needed (CD, p. 12-3?).
 Because an inadequate data base prevents drawing definitive
 conclusions  about sex differences,  staff assumes at this time
 that there are no differences between sexes regarding response to
 03  exposure.

      3.   Smoking  Status
      Results  of studies comparing the effects of 03  on smokers
 versus  non-smokers remain somewhat  inconclusive,  although below
 0.5 ppm  03 smokers  appear to  be somewhat less  responsive.  Two
 studies  reported greater pulmonary  function changes  for  non-
 smokers  than  for smokers at rest  during  exposure to  03 levels of
 0.37  to  0.5 ppm; this  finding was reversed  for hdgher  03  levels
 (0.75 ppm), with-smokers showing greater response  (Hazucha et
 al.,  1973; Bates and Hazucha, L973).  Greater  pulmonary  function
 response was  also  reported in non-smokers than in smokers at 0.5
 ppm (Kerr et  al.,  1975)  and at 0.3 ppm  (Delucia  et al.,  1983).
 Effects of 03 were greater for non-smokers compared to smokers at
 0.5 and at 0.3 ppm  (Kagawa and Tsuru, I979a);  in a subsequent
 study exercising non-smokers  showed a greater  response  (SGaw)
 than exercising smokers at 0.15 ppm O3 (Kagawa, 1983).
 Exercising smokers also showed slower and smaller spirometric
variable changes than non-smokers exposed 2 hours at 0.5 and 0.75
ppm 03 (Shephard et al., 1983).   Although none of these studies
has examined the effects of different amounts of smoking,
available data supports the contention that smokers are less

-------
                               VI-7
  responsive to O3 than non-smokers, at least for lower O3
  exposures.  Why smokers appear to be less responsive to O3 than
  non-smokers is not clear, though it has been suggested that the
  altered lung function and increased mucus production experienced
  by smokers could influence O3 deposition in the lungs (CD, p  12.
  37).
       4.   Nutritional status
       Results of studies investigating nutritional status as a
  factor affecting susceptibility to 03 are  inconclusive.   Some
  inconsistencies between human and animal data are apparent but
  might be  explained  based on different experimental approaches
       Human  subjects receiving 4 to 8  times the recommended daily
  allowance of  vitamin E  showed no statistically significant
  differences  in  blood biochemistry or  agglutination compared to
  unsupplemented  subjects  after 2-hour  exposures to  0.5 ppm  0,
  (Posin et. al.,  1979; Hamburger  et. al., I979a,b). Animal  studies
  show that vitamin E  deficiency makes  rats more  susceptible  to  o -
  induced enzymatic changes (Chow  et. al., 1981;  Piopper et.  al  '
  1979; Chow and Tappel, 1972) and that vitamin E alters the  rate
 and extent of O3 toxicity but not the lesion in the centriacinar
 region of the lungs  (Stephens et al.,  1974; Schwartz et al
 1976).  Lesions have been reported to be worse in vitamin E'
 deflclent rats or in rats marginally supplemented with vitamin E
 when compared to highly  supplemented rats (Piopper et al.,  1979-
 Chow et al.,  1981);  this supports results of mortality (Donovan'
 et  al  1977>  and biochemical studies  suggesting that vitamin E
 is  protective in rats (CD, p.  12-38).
      Since redistribution of vitamin E from extrapulmonary  stores
 in  humans  to  the lungs is slow,  it is  possible that such  I
 protective effect does not occur  over  short-term exposures  to C,
 in  human studies.  Another possible  explanation  is  that an  effect
 is  only seen when subjects are vitamin E deficient.  Animal
 studies in which effects have been reported were conducted with a
vitamin E deficient group for comparison and for long-term

-------
                              VI-8
 exposures (Sato et al., 1976, 1978, 1980),  neither of which is
 possible in human studies.  Although laboratory animal studies of
 vitamin E support the hypothesis that lipid peroxidation is
 involved in 03  toxicity,  benefits to be derived from dietary
 vitamin E supplementation in connection with O3 exposure have not
 yet been demonstrated in humans.

      5.   Environmental Stresses
      Subjective symptoms and physiological  impairment caused by
 03  exposure  can be  substantially worsened by environmental
 stresses such as heat and high relative humidity (rh).   High
 temperatures (31 to 40°C)  and/or humid  conditions  (85%  rh)  when
 combined with exercise during O3 exposure have  been  shown to
 reduce  FEV-L  more than  similar  O3  exposures at a more moderate
 temperature  (25°C)  and humidity  (50% rh) (Folinsbee  et.  al.,
 1977a,b;  Gibbons and Adams,  1984).   Enhanced effects of O3  caused.
 by  high  temperatures and  humidity may be the result  of  higher
 ventilation  which increases  the  volume  of inhaled  O3 and  promotes
 deeper penetration  into the  lungs.   There may also be an
 independent  effect  of  elevated body  temperature  on pulmonary
 function  (CD, p.  12-35).   Since,  many urban areas around the U.S.
 tend to  experience  the highest O3  levels during periods of  high
 temperatures  and humidity, environmental stress  should  be
 considered a  factor  of concern in assessing  potential effects
 from 03 exposure.

     6.  Exercise
     Exercise stresses  respiratory, cardiovascular, and
musculoskeletal  systems and  increases ventilation  and,  therefore,
total dose to the lungs.  Exercise thus should be  considered an
important susceptibility factor when 03 exposure occurs
concurrently.  The most apparent  and well-studied  effects of O3
during exercise occur  in the respiratory system.   In particular,
pulmonary function decrements and respiratory symptoms caused by
03 exposure are  increased by a greater work  load, which is

-------
                               VI-9
  characterized by increased frequency and depth of breathing
  (Folinsbee et al., 1979, 1984; McDonnell et al., 1983, 1985a,b,c;
  Avol et al., 1984, 1985a,b).  Representative activities and
  associated ventilation rates are summarized in Table Vl-i  (CD  p
  10-14) .   Higher ventilation increases total volume of O3 breathed
  and could cause deeper penetration of O3 into peripheral areas of
  the lung which tend to be most sensitive to injury.   Because most
  individuals engage in some type of outdoor exercise  where 03
  exposure is possible during summer months,  exercise  should be
  considered a susceptibility factor in assessing health effects of


  C.   Potentially Susceptible Groups
      The Clean Air Act requires EPA to set  standards which
  protect the^health of individuals who are potentially  susceptible
  to 03 exposure. 'This section identifies potentially susceptible
  groups or subpopulations and provides a rationale for selecting
 these groups.
      susceptibility to any pollutant may depend upon many  factors
 such as those previously discussed (age,  sex, smoking,  nutrition
 environment, and exercise,  but also is influenced by individual '
 sensitivity.  Greater than normal sensitivity to a particular
 pollutant »ay be conferred by numerous individual characteristics
 including (1)  predisposition to pulmonary infection,  (2)
 ar!^Stln! diSeaSe ^ nutritlonal  Deficiency,  (3)  some aspect of
 growth or dedme in  lung development,  (4,  a prior infection or
 immunological  problem,  (5)  sensitization  caused  by prior
 pollutant exposure  or  challenge with respiratory  irritants  and
 (6) genetic  variability in the population.   These  sensitivity  and
 susceptibility factors should be considered  in identifying  groups
 or subpopulations. which May be susceptible to 03 exposure

     1.   individuals Having Preexisting Disease
     The first major group identified in the CD (p. 12-73) as
appearing to be at particular ris* to O3 exposure is that group

-------
                                                                                  "nun.
Level of work
Light
Light
Moderate
Moderate
Moderate


Heavy

Heavy
Very heavy
Very heavy

Very heavy
Severe

Work
watts
25
50
75
100
125


150

175
200
225

250
300

-1 	
performed
kg-m/minb
•
150
300
vvU
icn
^9U .
fifln
uuu
750
/ wU

onn
y\j\M
ifl*ift
4U9U
1200
1350
* \/*JW
1500
A *JU\J
1800
4 U\J\J
	 	 	 	 	 , 	
02 consumption,
L/mln
._.
0.65
Of\£
.96
10 1~
.25
1C A
.54
1O1
.83

211
.12
24 ~i
.47
2.83
3 in
.19
3r r
.55
4OT
.27
.
Minute
ventilation
L/mln
•••
12-16
17-23
23-30
29-38
35-46


42-55

52-57
62-79
73-93

89-110
107-132
	 	
Representative activities
• •»•»»• •*^*i»ivivi Cd
Level walking at 2 mph; washing clothes



Level walking at 3 mph; bowling; scrubbing floors
Dancing; pushing wheelbarrow with 15-kg load-
simple construction; stacking firewooV
Easy cycling; pushing wheelbarrow with 75-ka
load; using sledgehammer 9
C 1 1 Alb*! nO «Ct" A i f*C • r\l a A *. .t
spade" 	 ' l""jr
-------
                              VI-11
 characterized as having preexisting disease.   This includes
 individuals  with asthma,  allergy,  and chronic obstructive lung
 disease  (COLD)  such as emphysema.   Asthmatics,  who experience
 variable  airway obstruction and/or reactivity and who may have
 altered immunological  states (e.g.,  atopy)  or cellular function
 (e.g., eosinophilia) may or may not be more sensitive to 03 than
 non-asthmatics.   They  have  not  been shown to  be more sensitive in
 numerous  studies, which have shown asthmatics to have equal or
 lesser responsiveness.   Asthma, however, is not a specific
 homogeneous  disease and efforts to precisely  define asthma have
 been unsuccessful.   Susceptibility to 03 of patients  with  COLD
 remains somewhat uncertain,  depending on their  clinical  and
 functional state.   Table  VI-2 provides estimates of the  number  of
 individuals  with these  conditions  (CD,  p. 12-79).
     Although several  controlled human exposure studies  suggest
 that adults  with preexisting respiratory disease may  not be more
 responsive to O3 than healthy, adults,  two considerations place
 these subgroups  at  potentially  higher risk  to 03  exposure
 according to the CD (p.  12-88).  First, due to  concern for the
 health of these  persons,  higher concentrations  of O3  and higher
 exercise levels  have not  been used during controlled  exposure
 studies,  thus making comparisons with healthy subjects difficult.
 A second consideration  is that  for individuals  with already
 compromised  respiratory systems, any  further  decrement in
 pulmonary function  may be expected to  impair  the  ability to
 perform normal activities.  The same  decrement  in  a healthy
 individual may go unnoticed.  Increases in symptoms reported for
 asthmatics and allergic individuals exposed to  O3 may also be
 expected to  interfere with normal function.    Because  recent
 studies (McDonnell et al., 1987; Eschenbacher et al., 1988; Kreit
et al.,  1988) have reported that both asthmatics and  allergic
subjects  have a greater increase in airway resistance after 03
exposure  than do healthy subjects,  the CD Supplement  (U.S.  EPA,
1988)  has  concluded that the order of airway responsiveness to O3
is normal  < allergic < asthmatic subjects.

-------
                                               TABLE VI-2
                                  PREVALENCE OF CHRONIC RESPIRATORY CONDITIONS BY SEX AND AGE FOR 1979a
Number of persons, in thousands
Condition
Chronic bronchitis
Emphysema
Asthma6
Hay fever and
other upper
respiratory
allergies
Total0
7474
2137
6402
15,620
Male
3289
1364
3113
7027
Female
4175
770
3293
8584
<17
years old
2458
12d
2225
3151
17-44
years old
2412
127d
2203
8278
45-64
years old
1547
1008
1482
3012
>65
years old
1060
990
488
1181
% of U.S.
population
3.5
1.0
3.0
7.2
aU.S.  Department of Health and Human  Services, 1981.
 Classified by type, according to  the Ninth Revision of the International Classific of Diseases (World Health
 Organization, 1977).
""Reported as actual number in thousands; remaining subsets have been calculated from percentages and are rounded off.
 Does  not meet standards of reliability or precision set by the National Center for Health Statistics (more than 30%
 relative standard error).
eWith  or without hay fever.
fWithout asthma.

 Source:   03 Criteria  Document, U.S.  EPA, 1986.

-------
                               VI -13
       2.  Exercising Individuals
       A second susceptible group potentially at higher risk to o,
  exposure is that group of. individuals including both normal
  healthy persons and those with preexisting respiratory disease
  whose regular outside activities cause increased minute
  ventilation.   As stated in the CD (p.  12-71) ,  ,,the most prominen
  modxf.er of response to O3  in  the general population is minute
  ventilation,  which  increases proportionately with increases in
  exercise workload."   The increased proportion of oral  breathing
  with  increased  exercise also contributes to the  increase in
  effects  since more O3 is delivered to the lungs.  Although
  individuals with preexisting respiratory or cardiovascular
  disease may not exercise as heavily as healthy persons, any
  increases in activity over resting levels will increase 0,
  exposure and resultant effects.
      Exercise has become recognized in numerous recent studies as
 a factor which can predispose all individuals to 03 health
 effects (McDonnell et al., 1983, 1985a,b,c;  Avol et al.,  1983
 1984,  1985a,b; Folinsbee et al., 1978,  1984;  Kulle 6t      ^
 Thus,  activities which increase minute  ventilation out-of-doors  '
 will also increase the risk associated  with  exposure  to 0, of
 exercising individuals.
     Unusual responsiveness  to  O3 has been observed in  some
 individuals.   These  individuals,  often  referred to as
 "responders," have not been  characterized, as having any
 particular medical problen but  experience signif ioantly greater
 pumonary function decrements than the average response of the

 If0tlssecon7rTy studied during °3 exposure-   zt is not *<>°™
 if these individuals are a population subgroup with a  specific
ri* factor or simply represent the upper 5 to 20 percent of the
03 response distribution.  As yet there  is  no Means of
identifying these highly responsive individuals  prior  to 0
exposure.                                         '        3
                reSP°nSiVeness *° °3 <* apparently healthy,
            groups has been shown to vary widely.  For

-------
                               VI-14
  during heavy exercise (minute ventilation, Ve/  = 45-51 L/min) and
  exposure to 0.4 ppm O3/  subjects showed FEY^ decrements ranging
  from below 10% to as much as 40% of control values for those
  subjects with an average of 26% (Haak et al., 1934; Silverman et
  al.,  1976).  Mode (oral versus nasal)  and pattern (rapid shallow
  versus slow deep)  of breathing may contribute to intersubject
  variability but cannot fully explain it.   Unsuccessful attempts
  have  been made to determine which factors are responsible for
  modifying individual responsiveness (Hazucha,  1981).   Factors
  such  as prior exposure to O3  or other  pollutants,  latent
  infections,  and nutritional deficiencies  may also contribute to
  differential response (CD,  p.  12-22).   The potential  for these
  contributions is biologically plausible,  but these have yet to be
  demonstrated in human subjects.
       Although the  specific  factors  which  modify  individual
  response  to  03  are not yet  identified,  it has been suggested that
  intersubject variability  in the magnitude  of effects  induced by
  03 is caused by  large differences in intrinsic responsiveness  to
  03 (McDonnell et al., I985a; Kulle et al., 1985).  As measured  by
  FEVi and FVC, individual responses to O3 are highly reproducible
  for at  least the 10 month period of observation and for  a given
  individual differences tend to  be much  smaller than differences
  between subjects  (McDonnell et  al., 1985a; Gliner et al., 1983).
  Hence, some  intrinsic factor appears to be responsible for
  individual responsiveness.
      Changes in FEVj_ of exercising subjects exposed to clean air
 are small and normally distributed in the subject population.  As
 03 concentrations increase this distribution widens and becomes
" skewed toward larger FEVj. decrements,  the largest changes
 representing the most responsive subjects (McDonnell et al.,
 1983;  Kulle et al., 1985).  Classification of subjects into'
 "responders" and "non-responders" has been somewhat arbitrary.
 Examples of retrospective classification include identifying
 responders by using medical history and exposure test results
  (Hackney et al. 1975)  and selecting those with greater than 10%

-------
                              VI-15
post-exposure  decrements (Horvath et al.,  1981)  or with FEV
decrements  greater  than 2  standard deviations above control (Haak
et al.,  1984).  while  there  are  no clearly established criteria
for identifying or  selecting "responders, '• it has  been suggested
that between 5 and  20% of  the healthy population may represent a
subgroup of the population which  is  more responsive and,
therefore, at higher risk  to O3 exposures  (CD, p.  12-22)'.
Further discussion of  frequency distribution  is  in section
VII.A.I.

-------

-------
                              VII-1
  VII. Assessment of H
       Considered in S
       This assessment of health effects attributed to O3 and
  related health issues is based on the review and evaluation of
  health effects research literature contained in Chapters 9 to 12
  of the CD as well as newer information discussed in the CDS.

  A.    Health  Effects  of Concern
       For purposes of this  staff  paper and  review of the NAAQS for
  03, the  staff recommends that  the  following  categories  of  health
  effects  attributed to 03 exposure  be  considered  in  developing  the
  basis  for the primary 03 standard:
       l.  Alterations  in pulmonary  function
       2.  Symptomatic  effects
       3.  Exercise performance
      4.  Bronchial Reactivity
      5.  Aggravation of existing respiratory disease
      6.  Morphological effects
      7.  Altered host defense systems
      8.  Extrapulmonary effects
 Respiratory effects provide by far the strongest data base for
 considering adverse health  effects of o3.   Extrapulmonary effects
 such as behavioral, blood chemistry,  chromosomal, cardiovascular
 reproductive,  teratological,  liver metabolism,  and endocrine    '
 system modifications  also have  been associated  with O3 exposure
 but most  are  considered to  be of  uncertain  functional important
 at  this time  and,  thus,  do  not  provide very strong evidence of
 adverse effects associated  with 03  exposures.
     Evidence for  the  above effects has come  from both human and
 animal  studies.  The strongest  and  most quantifiable data are
provided  by controlled human exposure  studies, but these studies
are limited to acute  (short-term) exposures.  Field  studies  are
similarly limited, but permit investigation of the effects of
oxidants  in ambient air and allow for better characterization of
exposure than epidemiological studies.  Although use of most

-------
                             V1I-2
 community studies has been hampered by the difficulty in
 adequately characterizing exposure and numerous confounding
 variables, these investigations provide important supporting
 evidence for effects occurring in populations.   Animal toxicology
 studies have provided evidence of some acute and chronic (long-
 term)  exposure effects which can be detected only with invasive
 procedures,  but uncertainties inherent in dosimetry and species
 sensitivity differences have limited quantitative extrapolation
 to humans.
     Assessment of health effects attributed to O3  requires
 consideration of the data base in each of the above areas of
 study.   This section integrates research  in  each of these areas
 to provide an indication of  the strength  of  the data base for
 each effect.   A more thorough review and  evaluation of individual
 studies  is available in chapters 9  to 12  of  the CD.

     1.  Alterations in Pulmonary Function
     The best documented and strongest evidence of  human  health
 effects  of 03  exposure  are pulmonary  function decrements.
 Controlled human  exposure, field, epidemiology,  and  animal
 toxicology studies have provided  evidence that  exposure to  O3 can
 modify such  pulmonary measurements  as  forced  expiratory volume
 (FEV),  forced  expiratory flow  (FEF),  forced vital capacity  (FVC),
 vital capacity (VC), tidal volume  (VT) , peak expiratory flow rate
 (PEFR),  inspiratory  capacity  (1C),  total lung capacity  (TLC),
 airway resistance  (Raw)  and breathing frequency  (fB).  These and
 other terms are defined  in Appendix B which is  an abbreviated
version  of the glossary  found  in the CD.
     Early controlled experimental  studies of resting human
subjects exposed to 03 levels up to 0.75 ppm for 2 hours
demonstrated little or no change in FVC (Silverman et al., 1976;
Folinsbee et al., 1975;), FEV1(, and FRC (Silverman et al., 1976).
Flow rate variables such as FEF 25% and FEF 50%  showed up to 30%
decreases in some subjects exposed at rest to 0.75 ppm O3 (Bates
et al., 1972; Silverman et al., 1976), while only small increases

-------
                            VII-3
in R
      aw (<  17%)  were reported for > 0.5 ppm 03  exposures (Bates et
  al.f  1972; Golden et al., 1978).  More recent studies have
  reported occurrence of FEV and FEF decrements during resting
  exposures  to > 0.5 ppm O3 (Folinsbee et al.,  1978;  Horvath et
  al.,  1979);  however, no statistically significant changes in R
  and only suggestive changes in RV and TLC have been reported for
  similar exposures (Shephard et al.,  1982).  Airway  resistance is
  not generally affected in resting subjects at  these 03  levels.
  Changes in pulmonary function have not been observed in resting
  subjects exposed  to 03  levels  between  0.12  ppm (Koenig  et  al
  1985)  and  0.3 ppm (Folinsbee et  al.,  1978), though  some subjects
  exhibit  03-induced pulmonary symptoms  during resting exposures
  (Konig et  al.,  1980;  Golden  et al.,  1978).  In general,  however
  because  subjects were at  rest  in most  of the older  studies
  significant respiratory effects were not reported even  for higher
 03 exposures.
      Exercise, which causes  increased minute ventilation (V )
 enhances individual and group mean response to 03 exposure * As
 discussed in section VLB. 6. of this staff paper, exercise
 increases breathing frequency and depth of breathing resulting  in
 greater total dose of 03 inhaled  and  increased  penetration to the
 most sensitive lung tissue.  As exercise levels increase to the
 point  where Ve exceeds approximately  35 L/min,  oronasal  or  oral
 breathing tends to predominate (Niinimaa et al.,  1980);  thus at
 higher exercise levels a greater  portion of the inhaled  O3  will
 bypass the  nose and nasopharynx (Niinimaa et al.,  1981)
 individual  variability will affect the Ve at which oral  or
 oronasal  breathing predominates.   The relationship between
 exercise  and magnitude of  response  is illustrated quite  well  by
 Figure VII-i prepared as Figure 12-6  for the CD (p.  12-31)
 showing group mean decrements in FEV, caused by exposure of
 exercising  subjects to various  03 levels based on results from 25
different studies.  The curves  clearly  demonstrate that  as
exercise  levels increase for  a given 03 exposure there is a
resultant increase in the group mean FEV, decrements.

-------
    110
    100
 ui
 5
 >   90
cc

2  80
Q
UI
o
cc
o

£  70
to
    60
                                                                    '••..  LIGHT EXERCISE
VERY HEAVY
EXERCISE
            uc*t
            HEAVY ^
            EXERCISE
                        MODERATE
                          U.2
0.4
                                                                  0.6
                                        0.8
                                  OZONE CONCENTRATION, ppm
    FIGURE VI I-1.  Group mean decrements in 1 -sec forced expiratory volume during 2-hr ozone
       exposures with different levels of intermittent exercise: light (v"E ^23 L/min); moderate tip =
       24-43L/mm); heavy (VE = 44-63 L/min); and very heavy (C'g ^ 64 L/min). Concentration-
       response curves are taken from Figures 12-2 through 12-5.
       Source:  03 Criteria  Document, U.S. EPA, 1986

-------
                              VII-5
       Pulmonary function decrements have been reported in
  controlled exposure studies of healthy exercising human adults
  exposed to 03 levels between 0.12 ppm and 0.24 ppm, a range for
  which a concentration-response relationship has been established
  During 2 hours of intermittent very heavy exercise (V  = 64
  L/min.), healthy subjects experienced group mean decrements in
  FEVi  of 4.5,  6.2,  and  14.5% for O3  exposures of  0.12,  0.18   and
  0.24  ppm,  respectively (McDonnell et al.,  1983).  Positive
  associations  between group mean magnitude  of  FVC (3,  4/  and 12%)
  and FEF (7.2,  12,  and  23%)  decrements and  O3 exposures of  (.12,
  .18,  and .24  ppm)  also were reported in the McDonnell  et al
  (1983)  study;  statistically significant increases  in SR   and f
  did not occur  until  O3 was  > 0.24 ppm.  In  a separate study   *
  McDonnell et al.   (I985b)  reported  a  statistically  significant
  but small decline  (3.4%) in group mean FEV, of children  (8-11
  years)  after 2 hours of exposure to 0.12 ppm 03 during
  intermittent heavy exercise  (Ve = 39 L/min).  Furthermore,  there
 was a suggestion that the small decrements  in FEV, persisted for
 16 to 20 hours after 03 exposure ended.  Findings by Lioy et al
  (1985) provide epidemiological evidence that lung function
 decrements can persist in children up to a week following a smog
 episode in which l-hour 03  peaks were 0.135 to  0.186 ppm.
      Support for 03-induced pulmonary function  changes  also comes
 from  other controlled exposure studies.  Avol  et  al. (1984)
 report small but statistically significant  decreases (6.1%)  ln
 FEVi,  at 0.16 ppm O3  with a  larger decrements (19.1%) at  > o 24
 PP* 03.   Decrements have been reported in FEV,, FVC, and  FEF for
 distance runners and  distance  cyclists exposed  to 0.20 ppm
 and 0.21 ppm O3 respectively, during 1 hour  of continuous, heavy
 exercise (ve =  77.5 and 81 L/min)  (Adams and Schelegle, 1983-
 Folinsbee et al., 1984).  Separate studies of continuously   '
 exercising males (Ve = 61.8 L/min) and females (Ve = 46 L/min)
 exposed to 0.2 ppm O3 for about an hour showed no statistically
 significant FEV, decrements, but these negative results can  be
attributed in part to the small number of subjects (i e

-------
                             VII-6
 inadequate power),  8 male and 6 female respectively  (Adams et
 al., 1981; Lauritzen and Adams, 1985).  In. another study
 involving 03 concentrations ranging between O.lo and 0.25 ppm/
 exponential decreases in FVC, FEV^  FEF25_75/ SGaw/  IC/  and TLC
 have been reported  with exposure to increasing 03 concentration
 during very heavy exercise (V(J = 68  1/min) ;  time of exposure was
 related to linear decreases in FVC and FEVX  (Kulle et al., 1985).
      Field studies,  which contain elements of both controlled
 human exposure and  epidemiologic studies,  provide the most
 quantitatively useful human exposure data  available for ambient
 photochemical oxidants.   Results from field  studies are
 consistent with pulmonary function decrements reported in
 controlled O3  exposure studies.   For a 1-hour exposure  to mean O3
 concentrations of 0.144  ppm,  small significant group mean
 decreases  in FVC (3.3%),  FEV0.75 (4.0%), FEVl (4.2%), MMFR (3.2%)
 and  PEFR (3.9%)  relative to pre-exposure levels were reported in
 59 healthy continuously  exercising (Ve=32 L/min}  adolescents
 (Avol  et al.,  1985 a, b).   Based on  the comparison  of ambient air
 and  control  exposures in these  studies, it appears  that O3 was
 the  causative  agent  of the  pulmonary function effects.
     In  a  separate study of  50  healthy, adult,  continuously
 exercising (Ve =  53  L/min) bicyclists, mean 03 concentrations of
 0.153  ppm  for  1  hour produced statistically  significant group
 mean decreases  in FEVl (5.3%) compared to pre-exposure  (Avol  et
 al., 1984).  This study  showed  that  similar  effects  result in
 subjects exposed  to comparable O3 levels in ambient and
 controlled 03 exposures.   Small but statistically significant
 decreases  in FEVX and FVC were also reported in exercising
 healthy and asthmatic adults during exposure to mean  O3
 concentrations of 0.165 and 0.174 ppm  (Linn et al.,  1980,  1983b;
Avol et al., 1983).   Table VII-l provides a summary of  group  mean
 % changes in FEVX for controlled exposure and field studies.
     In summary, pulmonary function decrements have been reported
 in healthy adult subjects (18 to 45 years old) after  1 to  3 hours
of exposure as a  function of the level of exercise performed  and

-------
Ozone
Concentration Measurement* 'b Exposure
Mtl/ir1 ppM Method Duration
0 0.00 UV, UV
157 0.08
0 0.00 UV, UV
157 0.08
196 0.10
235 0.12
0 0.00 UV. UV
157 0.08
196 0.10
0 0.00 CHEM. UV
157 0.08
196 0.10
235 0.12
0 0.00 UV, UV
196 0.10
0 0.00 CHEN, NBKI
196 0.10
0 0.00, UV, UV
216 O.llf
0 0.00 CHEM, UV
235 0.12
0 0.00 CHEM, UV
235 0.12
0 0.00 CHEM, UV
235 0.12
0 0.00 UV
235 0.12
0 0.00 UV
235 0.12
353 0.18
Source: Supplement to Air
1 hr
2 hr
2 hr
1 & 2 hr
of 6.6 hr
study
2 hr
2 hr
1 hr
2 hr
2 hr
1 i 2 hr
of 6.6 hr
study
1 hr
(mouthpiece)
40 mfn
(mouthpiece)
Quality Criteria
Activity0
Level (V£)
CE (57)
IE (68)
IE (68)
CE (40)
IE (68)
IE (67)
CE (22)
IE (68)
IE (39)
CE (40)
R
IE (33)
30 min
ft + 10 min
exercise
for 0
Percent .
Change in FEV?
+0.6
+1,7 (ns)
(26,4 i 6.9)
+1.0
+2.4 (ns)
+1.7 (ns)
+1.0
+2.4 (ns)
+1.7 (ns)
-1.5 (1 hr) -1.0 (2 hr)
-0.4 (ns) -1.1 (ns)
-1.3 (ns) -1.3 (ns)
-0.5 (ns) -2.7 (ns)
+1.5
+1.1 (nd)
range: +10 to -4
-0.3
-2.6 (ns)
' -2.7
-2.9 (ns)
-i.o
-4.5 (p = 0.016)9
range: +7 to -16
-0.5
-3.4 (p = 0.03)
range: +5 to -22
-0.2 (1 hr) -1.2 (2 hr)
-2.6 (ns) -3.8 (ns)
-1.1
0.0 (ns)
-1.0
+1.7 (ns)
-0.3 (ns)

Number, Sex, and
Age of Subjects
42 nale
6 female
24 male
(18-33 y'r)
24 Male
(18-33 yr)
21 male
O8-3J yr)
20 mate
(25.3 ±4.1 yr)
10 male
(18-28 yr)
33 male
33 female
(8-11 yr)
22 male
(22.3 t 3.1 yr)
23 male
(8-11 yr)
10 Male
(18-33 yr)
4 Male
6 female ,
(13-18 yr) .
5 male
7 females
(11-19 yr)
(continued on the
Reference6
Avol et al.
Linn et al.
Linn et al.
Horstman et
Kulle et al.
Foltnsbee et
Avol et al.
McDonnell et
McDonnell et
FoHnsbee et
Koenig et al.
Koenig et al.
following page)
(1984)
(1986)
(1986)
al. (1988)
(1985)
al. (1978)
(1987)
al. (1983)
al. (1985)
al. (1988)
(1985)
(1987)


-------
Table  VII-1  (confd)  KEY HUMAN STUDIES NEAR THE CURRENT 1-HR NAAQS FOR OZONE
Ozone
Concentration Measurement3'
ug/M3
0
235
0
235
0
235
0
235
274
0
274
0
294
0
294
0
294
0
314
0
314
0
314
0
333
0
333
PPM Method
0.00 UV
0.12
0.00 UV, UV
0.12
0.00 UV, UV
0.12
0.00 UV, UV
0.12
0.14
0.00, UV. UV
0.14r
0.00, uv, UV
0.15r
0.00 UV, UV
0.15
0.00 UV, UV
0.15
0.00 UV, UV
0.16
0.00 UV. UV
0.16
0.00, UV. NBKI
0.161
0.00, UV, NBKI
0.17r
0.00, UV. NBKI
0.17f
Exposure
Duration
1 hr
(Mouthpiece)
1 hr
(mouthpiece)
1 hr
2 hr
1 hr
1 hr
2 hr
1 hr
(mouthpiece)
1 hr
2 hr
1 hr
1 hr
2 hr
Activity0
Level (V£)
IE (33)
CE (86)
CE (89)
IE (6fl)
CE (31)
CE( 53)
IE (68)
CE (55)
CE (57)
IE (68)
CE (38)
CE (42)
IE (2XR)
Percent .
Change in FEV?
-2.4
-0.6 (ns)
+2.4
-1.8(ns)
+4.1
-5.6 (p <0.02)
range; +10 to -29
+1.0
+2.8 (ns)
+1.6 (ns)
' -0.5
-4.2 (p <0.01)
+0.6
-5.3 (p <0.05)
+ 1.5
-0.5 (nd)
range: +3 to -9
+0.6
-4.5 (ns)
range: +3.5 to -30.6
+0.6
-6.1 (p <0.05)
+ 1.0
-2.3 (p <0.05)
range: +8.9 to -35.8
-0.1
-0.8 (ns)
-0.4
-3.4 (p <0.006)
+0.6
-2.1 (p <0.05)
Number, Sex, and
Age of Subjects
5 male
8 females
(12-17 yr)
10 male
(19-29 yr)
15 male
2 female
(24 + 3 yr)
24 male
(18-33 yr)
46 male
13 female
(12-15 yr)
42 male
8 female
(26.4 ± 6.9 yr)
20 male
(25.3 ±4.1 yr)
10 female
(22.9 ± 2.5 yr)
42 male
8 female
'26. 4 ± 6 9 "r^
24 male
(18-33 yr)
27 male
21 female
(28 + 8 yr)
45 male
15, female
(30 + 11 yr)
14 male
20 female
(29 ± 8 yr)
Reference6
Koenig et al. (1988)
Schelegle and Adams (1986)
Gong et al. (1986)
Linn efal. (1986)
Avol et al. (1985)
Avol et al. (1984)
Kulle et al. (1985)
Gibbons and Adams (1984)
Avol et a). (1984)
Linn et al. (1986)
Linn et al. (1983);
Avol et al. (1983)
Linn et al. (1983);
Avol et al. (1983)
Linn et al. (1980, 1981)
                                                                   (continued on  the following page)

-------
                                                                 KEY """*" STUD1ESNEARTHE CURRENT 1-HR
   0   0.00
 470   0.24
                                            1  hr
                                                            CE (90)
                                                            +0.3
                                                            -3.1 (ns)
                                                            range:  +6.0 to  -16.6


                                                            +4.1
                                                            -21.6 (p <0.001)
                                                            range:  +10  to -46
                                                                                  NAAQS FOR OZONE
                                                                                                             Number, Sex. and
                                                                                                              Age of Subjects

                                                                                                              20 male
                                                                                                              (23.3 ± 2.8 yr)
                                                                                                              10  male
                                                                                                              (19-29 yr)
                                                                                                                    ...
                                                                                                              20  male
                                                                                                              (25.3 ±4.1 yr)
                                                                                                                8  male
                                                                                                               13  female
                                                                                                               (18-31 yr)


                                                                                                               15  male
                                                                                                               2  female
                                                                                                               (24 + 3 yr)
         Reference6
        '•
 McDonnell et al. (1983)


 	—	_—
 Schelegle and Adams (1986)

 	——	
 Kulle et al. (1985)


 	—         	—	
 Gliner et al.  (1983)



——	—	.
 Gong et a).  (1986)
                                                                                       10 male
                                                                                       (24 + 4 yr)
                                                                                                                                   Adams  and Schelegle  (1983)
                                                                                  -6.0 (p <0..05)
                                                                                                             8  male
                                                                                                             (22-46 yr)
                                                                                                                                    Adams et al.  (1981)

                                                                                                                                    	—	—	
                                                                                                                                    Lauritzen and Adams (1985)  ^
                                           1 hr
                                       (mouthpiece)
                                                                                        6 female
                                                                                       (22-29 yr)
                                                                                                                                  Folinsbee et al.  (1984)
UV, UV
                     1 hr
                                      CE  (60)
                                                                                  -1.0
                                                                                  -14.5  (p <0.005)
                                                                                  range: -1 to -36
                                                                                 	___
                                                                                 +0.6
                                                                                 -19.1  (p <0.05)
                                                                                                            20 male
                                                                                                            (22.9 ± 2.9 yr)
                                                                                                            42 male
                                                                                                             8 female
                                                                                                            (26.4 ± 6.9 yr)
                                                                                                            McDonnell et al. (1983)
                                                                                                            Avol et al. (1984)
                                                                                                                                  Schelegle  and Adams  (1986)

                                                                                                                                  	—	
                                                                                                                                  Kulle et al. (1985)
u              "
Measurement method:   CHEM = gas phase chemilumin^n
                                                               1traviolet
                     NBKI = neutra)
^s , not significant; nd = not

^See U.S. Environmental Protection Agency (1986).
^Measured in ambient air (mobile laboratory)

 "Suggested" significance based on Bonf.™,  ,nequality
                                           vent;)ation:   IE  =
                                          "Bn.f.c.nc. based on  difference between
                                                                                                        03 and filtered air (0.0 ppm 03) exposures:
                                                       co,.tC,lon (p 
-------
                             VII-10
 the 03 concentration inhaled during exposure.  Group mean data
 pooled.from numerous controlled human exposure and field studies
 and summarized in the CD  (p. 12-80) indicate that,
 on average, pulmonary function decrements occur at:
       (a)  > 0.5 ppm 03 when at rest (Ve . 5-10 L/min; e.g.,
           sitting);
       (b)  > 0.37 ppm 03 with light exercise (Ve = 10 to 23 L/min;
           e.g., slow walking); (c) > 0.30 ppm 03 with moderate
           exercise (Ve = 23 to 43 L/min; e.g.,  brisk
           walking);
       (d)  > 0.24 ppm 03 with heavy exercise (Ve = 44  to 63  L/min;
           e.g., easy running);
       (e)  > 0.18 ppm 03 with very heavy exercise (Ve  >64 L/min;
           e.g., competitive running).
      Although the group mean changes in lung function in the
.above studies are small, considerable intersubject variability in
 the magnitude of individual pulmonary response exists,  and some
 subjects experienced responses which were quite large (See
 Section VI.c.2).   Controlled exposures to 0.12  ppm 03 during very
 heavy exercise have resulted in individual pulmonary function
 decrements up to 16% for adults (McDonnell et al.,  1983) and up
 to 22% for children (McDonnell et al.,  I985b).   Individual
 subject data  from the Avol et al.,  1984,  McDonnell  et al.,  1983,
 and Kulle et  al.,  1985  studies have been used to estimate the
 fraction of population  which experiences > 10%  and  >  20% FEV,
 decrements due to 03  exposure  (Figure VII-2).   The  exposure-
 response relationships  represented in  Figure VII-2  suggest  that
 between 2 and  20% of the heavily  exercising population  might
 exhibit > 10%  decreases in FEV;L when exposed to 03 levels of
 approximately  0.12  ppm.   It is  estimated that between 0 and 5% of
 the heavily exercising  population might  experience  FEVX
 decrements of  > 20%  when O3  exposures are  0.12  ppm.   These
 estimates must be considered approximate considering:   (i)  the
 amount of variability of response between  the three studies
 analyzed,  (2)  the small  number  of subjects tested in  each study

-------
                            VII-11
  I
  I
I
                         D(FEV1) >= 10%
                               0-2        0.3

                     OZONE CONCENTRATION (ppm)
                                                            Avol
                                                             McDonnell
                       D(FEV1) >s 20%
0

OB
         o.o
X"
4
a
[
I .•
f
• Avol
a Kuito
At A^n !•
McDonnell

                   OZONE CONCENTRATION (ppm)
                                                  0.4
                 >20* r                    POPULATIOH

-------
                             VII-12
 (i.e.,  between 20-50 subjects at any given exposure level), (3)
 the fact that the subjects in these three studies were not
 derived from population based sampling,  and (4)  the use of a
 fitted  function based on empirical, rather than biological
 grounds.   Further discussion of this analysis can be found in
 Section VII.B.  of the staff paper and in Hayes et al.  (1987).
      Recent controlled human exposure,  epidemiology, and animal
 toxicology studies have increased concern about enhanced
 respiratory effects associated with multi-hour (6-8 hour)
 exposures to  03.   Clinical  research conducted  since closure on
 the CD  has provided data which indicate  lung  function  decrements,
 symptoms,  and even inflammation during prolonged 03 exposures
 below the current 03  NAAQS  level  of 0.12  ppm.  In addition,  camp
 and field studies offer evidence which corresponds well  to
 clinical  data on  lung function decrements and  symptoms.  New
 animal  toxicology research  is  helping to  make  a  stronger case  for
 needing to consider time as an important  variable along  with
 concentration and Ve  in  assessing health  effects  of  O3.
      Folinsbee  et al.  (1989) published data indicating that ten
 healthy non-smoking males  (18-33  years) exposed  to 0.12  ppm O3
 during  intermittent moderate exercise.  Exposures lasted 6.6
 hours and  consisted of  six  50-ininute  exercise Ve  =  40  L/min)
 periods interspersed  with 10 minute  rest  breaks  and a  35 minute
 lunch break.  Subjects  experienced  progressively  decreasing FEY-,^
 over the exposure period.   At  the end of  O3 exposure,  group mean
 FEVi had decreased by 13.0%, FVC by 8.3%,  and FEV25_75  by 17.4%
when compared to  clean  air  exposures.  One individual  experienced
a 47.6% decrease  in FEV^.  There were also progressive increases
 in symptom ratings  of cough and pain on deep inspiration for
subjects exposed to 0.12 ppm 03 as well as a marked increase in
methacholine airway responsiveness.
     In a continuation of the  Folinsbee et al. (1989)  work,
Horstman et al. (1988,1989) exposed each of 22 nonsmoking,
healthy, male subjects to 0.0, 0.08, 0.10, and 0.12 ppm  O3  on
separate days.  Exposure protocol was similar to that  described

-------
                              VII-13
  above for the Folinsbee et al. (1989) study.  Substantial
  pulmonary function decrements, respiratory symptoms, and
  increases in nonspecific airway reactivity were reported for all
  three O3 exposures when compared  to filtered air exposures
  Group mean decreases in FEV,  were 7.0% at 0.08 ppm o3,  7.0% at
  0.10 ppm 03/  and 12.3% at  0.12 ppm o3,  all  of which were
  statistically significant  (p  < .01).   Average ratings Qf ^ ^
  deep inspiration were low  but increased significantly (p <  05)
  following exposure to all  three 03 levels compared to filtered
  air.   Ratios  (p  < .005)  of PD100 observed for filtered air
  compared to 03 exposure were  1.56  at  0.08 ppm O3,  1.89 at O.io
  PPin  03,  and 1.89  at 0.12 ppra o3.  These results led Horstman et
  al.  (1988, 1989)  to conclude  that  clinically  meaningful  pulmonary
  responses can be  induced by 03 exposures and  exercise levels
 which simulate a  6- to 8-hour  period  of moderate to heavy work or
 play outdoors during" the O3 season.
    .  Further support of the relationship between acute 03
 exposure and pulmonary function decrements is provided by several
 epidemiological studies of children and young adults  (Lippmann et<
 al.,  1983; Lebowitz et al., I982a, 1983; Lebowitz.  1984; Bock et
 al-,  1985; Lioy et al., 1985;  Spektor et al.,  I988a,b).  These
 studies report decreased peak  flow or increased airway resistance
 for acute exposures to ambient O3  concentrations  during the study
 period.   Lioy  et  al.  (1985)  reported that a  persistent decrement
 in  lung function  of children lasted as much  as a  week after the
 end of a smog  period of four days  during which peak l-hr O3
 levels were in the range  of 0.135  to 0.186 ppm.   The persistent
 decrements suggested by Lioy et al.  (1985)  (i.e., altered
 epithelial permeability and changes in airway  secretion)
 represent a potentially more important response than the  more
 transient effects  found in controlled  exposure studies; however
 they recognize that elevated concentrations of inhalable
particulate matter associated with a large scale photochemical
smog episode also are associated with  effects  reported.   Spektor
et al. (I988a)  have followed up the work of Lioy et al. (1985)

-------
                             VII-14
 with a summer camp study of children exposed to ambient 03 in an
 area which did not exceed 0.12 ppm 03 during the study.   Multiple
 regression analyses indicate that the most explanatory
 environmental variables for daily change in lung function were:
 1)  previous hour 03  levels,  2)  cumulative  daily O3 exposure
 between 9  a.m.  and the time of measurement of lung  function,  3)
 ambient temperature,  and 4)  humidity.   Spektor et al.  (1988a)
 concludes  that the l hour O3  levels  have the strongest influence
 on  lung function and that regression slopes for FVC,  FEV,, PEFR,
 and MMEF predict average decrements  of 4.9%,  7.7%,  17% and 11%,
 respectively at 0.12  ppm 03.   Upper  decile  decrements  are
 predicted  to be 14%,  19%,  42%,  and 33% for FVC,  FEV^  PEFR, and
 MMEF,  respectively at 0.12  ppm O3.   Lioy and  Dyba (1988) have
 proposed recently that the  PEFR decrements  reported  in Lioy et
 al.  (1985)  resulted  from total O3 dose rather than persistence of
 the effect  from one  day to  the next.
      Epidemiology studies comparing  incidence of  chronic  lung
 disease  in  communities  have thus  far been  relatively unsuccessful
 due to the  lack of differences  in pollutant  levels,  inadequate
 control  of  covariables,  and insufficient individual exposure data
 (CD,  p.  11-54).   While  it has  been concluded  in the CD  (p. 11-53)
 that  not one  of  these epidemiological  studies provides definitive
 quantitative  data by  itself due to methodological problems and
 confounding variables,  the aggregation of studies provides
 reasonably good  qualitative evidence of association between
 ambient photochemical oxidant exposure and acute pulmonary
 effects.  The association is strengthened by the consistency of
these epidemiological results with the findings of McDonnell et
al.   (1985b) and Avol et al. (I985a,b) who reported small
decrements in pulmonary function for exercising children exposed
to  0.12 ppm 03 in purified air and adolescents exposed to 0.144
ppm O3 in ambient air, respectively.
     In conclusion, the weight of currently available evidence
indicates that healthy, heavily exercising subjects can
experience pulmonary function decrements during controlled

-------
                              VII-15
  exposures of > 0.12 ppm 03 for 2 hours.  Although level of
  exercise and individual responsiveness play a major role in
  determining the extent of pulmonary function loss, staff
  concludes that 0.12 ppm 03 is the lowest observed effects level
  (LOEL)  for this effect in controlled exposure studies of 2-hour
  exposures.   Effects have been reported as low as 0.08 ppm when
  exposures lasted for 6.6 hours.   Field and epidemiology data
  provide added evidence of measurable functional decrements below
  0.12  p»,  perhaps in part due to  pollutant interactions
       The.question of what degree of pulmonary  function response
  should  be considered adverse is  addressed in Section  VII.c.l.

       2.   Symptomatic Effects
       Respiratory  symptoms  have been associated with group mean
 pulmonary function  changes in adults acutely exposed  in
 controlled exposures to 03 and in ambient air containing 03 as
 the predominant pollutant.  Despite a close association observed
 between changes in group mean FEV, and group mean respiratory
 symptoms for 03 exposures (CD, p. 12-17), Hayes et al. (i987b)
 report only a weak-to-moderate correlation between FEV, changes
 and symptoms severity when the analysis is conducted using
 individual data.  It should be noted that symptoms reported are
 inherently more subjective than FEV, decrements  measured
      in  controlled O3 exposures,  some heavily exercising (V  >  65
 L/min) adult subjects have experienced  cough, shortness of "  ""
 breath,  and  pain on deep inspiration at 0.12 ppm 03, although the
 group  mean response was statistically significant for  cough  only
 (McDonnell et  al.,  1983).   Above  0.12 ppm o3, respiratory and
 non-respiratory  symptoms which have  been reported include throat
 dryness, chest tightness,  substernal pain, cough,  wheeze, pain on
 deep inspiration,  shortness of breath, dyspnea, lassitude
malaise,  headache, and  nausea  (DeLucia and Adams,  1977; Kagawa
and Tsuru, I979a,b,c; McDonnell et al.,   1983; Adams and
Schelegle, 1983;  Avol et al., 1984; Gibbons and Adams, 1984;
Folinsbee et al., 1984;  Kulle et al., 1985).  At 0.2 ppm O3 and

-------
                                VII-16
            controlled
                       exposure
     .„  n                 n
     (McDonnell et al.,  „
         been anaiyzed
               P0pulation
                syn,ptoms
    Figures VIM and VII.4
                             24 hours
             r<*°«« .
              et al.  „                      "
                subjects
    -creases.   For exanple
                                       ., 1985;  vol ^
                                                  the            '
                                                    and
                                          Varl°
                                         t              e.response
                                 ng         " thSSe studi«, the
                                           *"** ** °'
            , between o and 15*  f
                                   or chh:;tvidly          9 «•«*
            level rises to o 2 nl ^K       disc™fort.   As the 0
                                   the ««-ted  fraction of the
                                                      cl>est
 Population that night
            increases to a
    fact that the subjects in      7 giVe" aXposu" ^vel,   (3
f - ^^cn based3sa:pl ;th:ndth;;: StU«- -re not ^   d
junction based on e*piricai,  ^ther *H   ^ US'3 °*
Section VH-B Of tnis     *'  r"her than biological
        a .ore ^11
    anaiysis.                      ot the respiratory sympto»

                                     in
exercising (v. . 57
Purified air containing : ,
« -ich contained o.15 ppffl
                                         °f
                                             fieid studies,  one
                                                     in

-------
                       = Mild Chest DiscomToT
         0.0
           0.0
        0.0
          0.0
                     01         02        0.3


                     OZONE CONCENTRATION (ppm)
 01        0.2        0.3


OZONE CONCENTRATION (ppm)
                                                   0.4
FIGURE VII-3    FRA
                                              1.0
                                                                 0.8
                                                                            >= Moder. Chest Discomfort
I "]


-------
        VII-18
       Lower Respir. Symp. (Avol)
a  >. Mild



•  >« Moderate
        OZONE CONCENTRATION (ppm)
                     POPULATI™

-------
                              VII-19
  significant differences, suggesting that increased symptoms
  associated with lung function impairment were caused by O3 alone
  (Avol et al.,  1984).  Several epidemiology studies have provided
  evidence of qualitative associations between ambient oxidant
  levels > 0.10  ppm and symptoms in children and young adults such
  as throat irritation,  chest discomfort,  cough, and headache
  (Hammer et al.,  1974;  Makino and Mizoguchi,  1975).   Thus  it can
  be concluded that most symptoms  reported in  individuals exposed
  to 03  m  purified air  are similar, but not identical  to, those
  found  for ambient air  exposures.
  .    An exception is eye irritation, a common  symptom associated
  with exposure  to  photochemical oxidants, which has  not  been
  reported  for controlled  exposures  to 03 alone.  This  appears to
  hold even at 03 concentrations much higher than would be found in
  the ambient air.   it is  widely accepted that other  oxidants such
  as aldehydes and peroxyacetyl nitrate (PAN) are primarily
 responsible for eye  irritation and are generally found  in
 atmospheres containing higher ambient O3  levels (Altshuller
 1977;  National  Research Council,  1977;  U.S. Environmental
 Protection Agency, 1978; Okawada et al.,  1979).
      Pulmonary  function decrements have been reported in studies
 which  did not report increases in symptoms.   Children (age 8-11)
 intermittently  exercising (ve = 39  L/min)  for 2.5  hours at  0 12
 ppm 03  showed small,  but  statistically significant decreases in
 FEV, but showed no changes in frequency or  severity  of cough
 compared to control (McDonnell et al., 1985a,b).   Similarly
 adolescents (age  12-15)  continuously  exercising (V   =  31-33'
 L/min)  during exposure  to 0.144 ppm mean  O3 in  ambient air  showed
 no  changes in symptoms  despite statistically  significant
 decrements in group mean  FEV,  (4%, which persisted at  least one
 hour during resting post-exposure  (Avol et al., I985a,b)
 Because symptoms can  be viewed as an early warning of  related
 lung function impairment  by 03, the lack of symptoms in children
and adolescents during exposures which induce functional
decrements may be of concern.  It is possible that children may

-------
                             VII-20
 be at higher risk since, with no warning, they may not exhibit
 avoidance behavior.
      Symptoms when combined with objective measures of lung
 function are considered useful adjuncts in assessing health
 effects caused by O3 and photochemical oxidants.   Group mean
 symptoms have been shown to be closely associated with the time-
 course and magnitude of group mean pulmonary function changes
 associated with O3 exposures;  however,  only weak-to moderate
 correlations exist when data are analyzed on an individual basis.
 To the extent symptoms associated with exposure to 03 and other
 photochemical oxidant exposures are associated with discomfort,
 interfere with normal activity,  and provide subjective evidence
 of functional impairment,  the staff in concordance with comments
 made by CASAC (CASAC,  1986)  recommends that they  should be
 considered adverse health effects and as such should be included
.in identifying a lowest observed effects level (LOEL).

      3.   Exercise Performance
      A  limited data base provides suggestive evidence of
 decrements in exercise performance  associated with O3 exposure.
 A  detailed discussion  of those data can be  found  in the CD (pp.
 10-60 to p.  10-65)  and is  summarized in Table 10-6 (CD, p.  10-
 64).
      Early epidemiological evidence on  high  school students
 suggested that the percentage  of  track  team  members failing to
 improve  performance increased  with  increasing oxidant
 concentrations the hour before a  race  (Wayne et al.  1967).   The
 authors  concluded that the effects  may  have  been  related  to
 increased Raw or to associated discomfort which may have  limited
 motivation to run at maximal levels.  Controlled  exposure studies
 of heavily exercising  competitive runners have demonstrated
 decreased max Ve  at 0.3 ppm 03 (Savin and Adams, 1979) and
 decreased FVC,  FEF, and FEF at 0.20 ppm 03 (Adams and Schelegle,
 1983).  At 0.21 ppm 03, Folinsbee et al., (1984) reported
 decreases in  FVC,  FEV,  FEF, 1C, and MW at 75% max  VO2 as well as

-------
                              VII-21
  symptoms in 7 distance cyclists exercising heavily (V  = 81
  L/min).                                               e
       With regard to exercise performance studies,  reductions in
  group mean FEV,  (-5.6%)  and  mild symptoms (cough,  chest soreness
  shortness of breath)  are reported in 17  endurance  cyclists
  exposed  at 31'c  to  0.12  ppm  03  for  l  hour during very  heavy (v  =
  88 L/min)  continuous  exercise (Gong et al.,  1986).  Although no
  impairment of maximal performance is  noted at  0.12  ppm,  when Gong
  et al. (1986) exposed the same  subjects  to 0.20 ppm 03 under the
  same  exercise conditions, he  reports  that both group mean  FEV,
  decrements  (-21.6%) and  symptoms  (prominent  cough and  nausea)  are
  intensified and result in impairment  of maximal performance
  Similarly, Schelegle  and Adams  (1986) report that 10 highly
 trained endurance athletes exposed to 0.12,  0.18, and  0.24  ppm 0,
 while performing a 1-hour competitive simulation (Ve . 86 L/min)
 on a bicycle ergometer show a significant  (p < 0.05) increase in
 the inability of subjects to complete the simulation.   Even
 though one subject was not able to complete the simulation at any
 of the 03 levels  tested,  five were unable to complete  at 0.18 ppm
 03  and seven did  not finish at 0.24  Ppm o3.   Only at 0.18 and
 0.24  ppm  03 are statistically significant changes reported  in
 PEVlf  FVC,  and subjective symptoms.   Schelegle  and Adams  (1986)
 report no significant  changes in lung  function  or symptoms  at
 0.12  ppm  03.   spektor  et  al.  (i988b) offer substantiation of the
 effects of 03  on respiratory  function  in populations engaging in
 continuous exercise  for short  periods.
     Although  subjective  statements  by individuals engaged  in
 sports indicate possible  voluntary curtailing of activities
 during high-oxidant episodes,  increased temperature  and relative
humidity may be involved  in provoking  the  symptoms and  lung
 function decrements observed in the studies above.   Controlled
studies of 03 exposure have,  however, demonstrated lung function
impairment and subjective symptoms which cause individuals to
             0^ "" Perf°— at °'2 W °3 and above  (Adams
             , 1983;  Folinsbee et al.,  1984; Avol et al.,  1984)

-------
                             VII-22
 Because O3 may be implicated at least in part in reducing
 exercise performance during periods of high oxidants and some
 individuals experience this effect at levels near the 03 NAAQS,
 the staff recommends that this effect of 03 be viewed as a matter
 of potential public health concern and be considered in
 developing a margin of safety.

      4.   Bronchial Reactivity and Inflammation
      Bronchial reactivity is reportedly, associated with exposure
 to 03 (CD,  p.  10-30).   Using airway responsiveness to the drugs
 acetylcholine (ACh),  methacholine,  or histamine as a measure of
 non-specific airway sensitivity,  researchers have observed
 increased bronchial reactivity in both healthy and asthmatic
 subjects, exposed to 03  levels  in  the  range  of  0.32  to l.o  ppm.
 It should be noted that most of these studies  were conducted on
 resting  or only moderately exercising individuals.
      In  both atopic and non-atopic  subjects exposed,  using a
 noseclip,  to 0.6 ppm  O3 during  intermittent  exercise, bronchial
 response to histamine and  methacholine was  greater  when  subjects
 were  exposed to O3 than to  filtered air  (Holtzman et  al.,  1979).
 Symptoms of bronchial irritation  in the  Holtzman  et  al.  (1979)
 study were  not  detectable  by the  next day,  which  contrasts  with
 results  of  the  Golden et al.  (1979) study of healthy  resting
 subjects  exposed to 0.6 ppm 03 reporting enhanced response  to
 histamine challenge for as  much as 1 to  3 weeks after exposure.
     Other  studies of bronchial reactivity show statistically
 significant  enhancement when O3 exposures are as low as 0.32 ppm
 but not at  0.20  ppm.  Significant increases in bronchial
 reactivity were  reported with ACh challenge in healthy adult
 subjects exposed to either  0.32 ppm or 1.0 ppm o3 (Konig et al.,
 1980).  Healthy  subjects showed no alterations of bronchomotor
response to histamine aerosol when exposed to 0.2 ppm O3 for 2-
hr, although exposure to 0.4 ppm 03 did enhance bronchial
responsiveness to inhaled histamine (Dimeo et al., 1981).  Kulle
et al. (I982b)  reported significantly enhanced bronchial

-------
                              VII-23
  reactivity to methacholine of healthy subjects exposed to  0.4  Ppm
  03 as compared to filtered air.
       McDonnell et al. (1987) report increased reactivity to
  histamine following exposure of healthy subjects to 0.18 ppm o,
  during intermittent very heavy exercise (64 L/min) for 2 hours
  increased airway reactivity to methacholine following 2 hour
  exposures of subjects to 0.4 and 0.6 ppm 03 during intermittent
  moderate exercise is reported by Seltzer et al.  (1986)  to be
  associated with neutrophil influx.   Using bronchoalveolar lavage
  (BAL),  Keren et al.  (1989a,b,c)  observe  increased inflammation
  (8.2  fold increase in polymorphonuclear  leukocytes,  PMN's)  at 18
  hours post exposure  in healthy  adult males exposed for  2  hours to
  0.4 ppm 03 during  intermittent exercise,   m addition to
  confirmation  of  the  Seltzer et al.  (1986)  findings,  Koren et  al
  (1989a,b,c) provide  evidence  of  stimulation of fibrogenic
 processes  and further  suggest that the inflammatory  process
 initiated  by 03 exposure is promptly initiated (Seltzer, et al
 1986)  and persists for at least  18 hours.  Although the time
 course of the inflammatory response has not been elucidated  the
 Koren et al.  (I989a,bfc)  research demonstrates that cells and
 enzymes capable of causing damage to pulmonary tissues are
 increased as a result of 03 exposure.  Furthermore, proteins
 which  play a role in fibrotic and fibrinolytic processes are
 elevated by exposure to o3.   Studies  by Kehrl  et  al.  (1987/  1989)
 confirm  that clearance of technetium labeled DTPA from airway and
 alveoli  into the bloodstream is  accelerated after O3  exposure  and
 along  with Koren et al.  (1989a,b,c)  suggest that  this accelerated
 change is due,  in part, to  increased  epithelial permeability in
 the lung.   These  permeability  alterations  are  likely  to  be
 associated  with acute inflammation and may  allow  inhaled antigens
 and other substances  to more easily reach the submucosa
     in an  extension  of a 2-hour  exposure study, Koren et al.
 (1989a,b,c) exposed ten healthy non-smoking males  (18-35 yrs) to
 0.1 ppm 03 for 6.6 hours while subjects were engaged in a
moderate to heavy  (ve = 40 L/aln)  interinittent exercise

-------
                             VII-24
 similar to Folinsbee et al.  (1989).   Results indicated suggestive
 evidence of an inflammatory ressponse in the lower airways as
 evidenced by a 4.8-fold increase in  percent of polymorphonuclear
 leukocytes (PMN's)  in the bronchoalveolar lavage fluid (p =
 0.034).   Importance of these preliminary results lies in the fact
 that they represent evidence of inflammation in humans and
 potential for damage in lower airways from O3  exposures which may
 occur often in ambient air.
      In  conclusion,  there is some evidence to  suggest that
 bronchial reactivity and inflammation occur at ambient
 concentrations of O3.   This  is  particularly true for  subjects
 exposed  to O3  during exercise.   Because  the data base showing
 effects  at low concentrations  of 03  is relatively new and  small,
 staff recommends that this information be used to develop  a
 margin of safety.

      5.   Aggravation of Existing Respiratory Disease
      Some epidemiological  studies suggest an association between
 photochemical  smog and aggravation of existing respiratory
 disease.   No clear evidence  is  available,  however,  from
 controlled exposure  or field studies to  suggest that  individuals
 with  asthma, chronic bronchitis,  or  emphysema  have  greater lung
 function  impairment  caused by O3 or other photochemical oxidants
 than  healthy persons.   However,  individuals  with preexisting
 respiratory disease  are  considered to be  especially "at risk"  to
 O3 exposure due to their already compromised respiratory systems
 and concern that increased symptoms or pulmonary function
 decrements may  interfere with normal function  (CD, p.  12-73).
      In controlled human exposure studies,  statistically
 significant group mean decrements in pulmonary  function were  not
 reported  for adult asthmatics exposed for 2  hours at  rest
 (Silverman, 1979)  or with intermittent light exercise  (Linn et
al.,  1978) to O3 concentrations of 0.25 ppm or lower.   Similarly,
 no statistically significant group mean changes  in pulmonary
 function  or symptoms were found  in adolescent, asthmatics exposed

-------
                              VII-25
  to 0.12 ppm 03 for 1-hour at rest (Koenig et al., 1985).
  Subjects with chronic obstructive lung disease (COLD) performing
  light to moderate intermittent exercise show no statistically
  significant group mean decrements in pulmonary function during 1-
  and 2-hour exposures to < 0.30 ppm 03 (Linn et al.,  1982,  I983a-
  Solic et al.,  1982;  Kehrl et al.,  1983,  1985)  and only small,
  statistically  significant group mean decrements in FEV,  are
  observed for 3-hour  exposures of chronic bronchitics to  0.41 ppm
  03  (Kulle  et al.,  1984).   while these controlled  exposure  results
  indicate that  individuals with pre-existing respiratory  disease
  may not  be more sensitive to 03  than  healthy subjects,
  experimental design  considerations  in these studies  suggest that
  the issues of  sensitivity and aggravation of pre-existing
  respiratory disease  remain unresolved.
      Several new studies  compare effects of 03 exposure on
 asthmatics with effects on healthy persons.  Although no
 differences in 03-induced symptoms are reported between 9 normals
 and 9 asthmatics exposed to 0.4 ppm 03/ Kreit et al.  (1988) do
 report a more negative change in FEV,, FEV./FVC,  and  FEF2,•  for
 asthmatics than for normals.   SRaw  is not significantly increased
 in normals but  is  in  asthmatics, though it should be noted that
 significant increase  in SRaw  occurs  in asthmatics  even  for  air
 exposure.  Koenig  et  al.  (1987) compare the effects of 0.12 and
 0.18  ppm 03 on  10  asthmatics  and 10  nonasthmatics  exposed for 30
 mm.  at rest  followed by 10 min. exercise (32.5 L/min) .   NO
 significant effects are found at 0.12  ppm O3 while both groups
 show  increases  in  respiratory resistance  (RT, at 0.18 ppm O3.
 Koenig  et al. (1987)  conclude that-there  are no significant
 differences in  response to 03  between  asthmatics and
 nonasthmatics.  in  a  similar  study, Koenig et al.  (1988)  find
 that 12 healthy adolescents exposed to 0.12  ppm O3, 0.30 ppm  NO2
 or a combination during intermittent moderate exercise show no
 significant response,  while 12 asthmatic adolescents showed
marginal responses  in pulmonary  function for  .12 ppm O3 and o  3
ppm N02 but not  the combination.  This leads the authors to

-------
                             VII-26
 conclude that asthmatic subjects may be slightly more responsive.
 However,  replication of these observations will be required
 before this suggestion can be substantiated.
      Epidemiological studies do not provide a clear
 concentration-response relationship between O3  and aggravation of
 disease.   One study (Whittemore and Korn,  1980)  conducted in Los
 Angeles reported  increased daily asthma attack  rates  on cool days
 and days  with high  oxidants and "particulates  when median daily
 maximum hourly oxidant levels were  < 0.15  ppm;  questionable
 exposure  assessment,  lack  of control for medication,  pollen,
 respiratory infections and other pollutants limits the use of
 this  study for developing  quantitative  dose-response
 relationship.
      Despite  several  exposure and variable  control limitations,
 the Houston Area  Oxidants  Study (Johnson et al.,  1979;  Javitz et
 al.,  1983)  concluded  that  for the study period  in which the daily
 maximum hourly 03 concentrations were below- 0.21  ppm 03 near  the
 subjects'  residence,  1) there was increased incidence  of  nasal
 and respiratory symptoms and  increased  frequency  of medication
 use for asthmatics with increasing 03 levels; 2)  FEV1 and FVC
 decreased  with increasing  03 and total oxidants;  and 3) increased
 incidence  of  chest discomfort,  eye irritation, and malaise
 occurred at high peroxyacetyl  nitrate (PAN) concentrations.   In  a
 subsequent  related study,  (Stock  et al., 1983; Holguin  et al.,
 1985; Contant  et al.,  1985) increased probability  of an asthma
 attack was  associated with the occurrence of a previous attack
 and with exposure to  increased O3 concentrations and temperature
when maximum  1-hour averages  for 03 were between 0.001 and 0.127
ppm; however,  other pollutants such as S02  and particulates may
have been  involved.
     In a series of studies conducted in a Tucson community,
adults with asthma,  allergies, or airway obstructive disease
 (AOD)  were observed during an 11 month period in which 1-hour
daily maximum 03 concentrations were < 0.12 ppm (Lebowitz et al.,
 1982,  1983; Lebowitz, 1984).  After adjusting for covariables, O3

-------
                              VII-27
 and TSP  levels were  significantly associated with decreases in
 peak expiratory  flow rate  in adults  with  ADD,  and the interaction
 between  O3 and temperature was significantly associated with
 alterations  in peak  flow and symptoms  in  asthmatics.   While these
 results  suggest  an effect  of 03 in individuals with preexisting
 respiratory  disease,  interpretation  is difficult  due  to the small
 sample size  in relation to the number of  covariates and the fact
 that individual  exposure data were not available.
      Bates and Sizto  (1987)  conducted a study in  southern  Ontario
 and reported that there is a consistent summer relationship
 between  SO4/  03  and  temperature and hospital respiratory
 admissions.  The correlation between 03 and hospital respiratory
 admissions was not affected  by substituting  the 8-hour average 03
• for mean hourly maximum 03 levels.  They further suggest the
 adverse, health effects studied may not be associated with  either
 O3 or sulfates but rather with some species not yet monitored
 which they refer to as "acid  summer haze."   Bates and Sizto
 (1988)  failed to confirm association between excess hospital
 admissions and 03 for June 1983,  the month of the highest O3 of
 any month in their analyses.   In a separate  study conducted  in
 Ontario,  Raizenne and Spengler (1987) have preliminarily reported
 that lung function decrements in children attending summer camp
 are associated with maximum daily 03 levels as well as
 temperature,  fine particles,  and sulfates.  Over the course of a
 41-day-study, Raizenne and Spengler (1988) continuously measured
 03/  N02,  S02, and acid aerosols  (H2SO4) .  In plotting individual
 6-hour estimated dose for 03  and H2S04  (separately) versus
 percent change in PEFR,  they  found a negative trend in lung
 function  as cumulative dose increases for both 03  and  H2S04
 although  the  slopes  did not differ from zero (p > .10)  when
 dosimetry was used.
      None of  these controlled human exposure or epidemiology
 studies alone demonstrates  a  clear relationship between exposure
 to 03 and aggravation of preexisting  respiratory disease.   All of
 the epidemiology  studies  cited report effects which may be

-------
                             VII-28
 related  to  inhalable particle  exposure  and  most have inadequate
 characterization  of  exposure.   However,  the group  of studies as a
 whole  supports  the contention  that  exposures to ambient levels of
 03 and other photochemical oxidants recently reported  in many
 cities  (See Appendix A) may  increase the rate of asthma attacks
 for asthmatics.   The staff believes that increased asthma  attack
 rate is  an  adverse effect but  recognizes the limited and
 uncertain data  base  relating this effect to 03  exposure.   Thus,
 in concordance  with  uncertainty expressed by CASAC (1986,  1987,
 1988)  it is recommended that data associating 03 exposure  with
 aggravation of  existing respiratory disease be considered  only  in
 the development of a margin  of safety for the primary  standard
 unless further  research provides more conclusive evidence.

     6.  Morphological Effects
     Morphological effects of 03 have been reported primarily in
 animal toxicology studies (See Table VII-2).   For  this  reason  it
 is important to consider the differences in  dosimetry  and
 sensitivity between  humans and laboratory animals,  which is
 discussed in Chapter  9 of the CD.   The following discussion
 provides an indication of some of the structural changes which
 might occur in human  lungs as a result of repeated 'and  long-term
 exposures to O3.
     Despite the differences in lung structure between  humans,
 dogs,  monkeys,  mice,  rats, and guinea pigs,  a characteristic
 lesion occurs at the junction of the conducting airways and the
gaseous exchange tissues.   In all  species examined the  effect is
typically damage to the ciliated and Type 1  cells and hyperplasia
of non-ciliated bronchiolar and Type 2 cells; an increase  in
 inflammatory cells is also observed (CD, p.  12-58).  These
effects were reported after 7 days,- 8 hr/day exposures to  0.2 ppm
03 in monkeys  (Dungworth et  al., 1975;  Castleman et al., 1977)
and in rats exposed for 7 days, 8  and 24 hrs/day (Schwartz et
al.,  1976).   Similar effects with  different  exposures were

-------
                                        VI1-29
           TABLE  vn-2.   MORPHOLOGICAL EFFECTS OF
   Effect
                                OZONE IN EXPERIMENTAL ANIMALS
      Ozone
concentration,ppn
      Exposure
      duration
                                                          Species   Reference.
   Damage to
   ciliated and  .
   centriacinar
   alveolar type 1
   cells
         J-2       8 hr/d x 7 d
         °-2       8 hr/d x 7 d
         0-25      12 hr/d x 6 wk
         0-25      12 hr/d x 6 wk
         0-5        2 hr
         °-5        2 hr
                      Rat
                      Monkey
                      Rat
                      Rat
                      Rat
                      Rat
           Schwartz.et al., 1976
           Castleman et al., 1977
           Barry et al.,  1983
           Crapo et al.,  1934
           Stephens,  et al.,  I974a
           Stephens,  et al.,
  Hyperplasia of
  nonci Hated
  bronchiolar and
  alveolar type  2
  cells
        0.2
        0.35

        0.5
       •0.5
        0.8

        0.8
  8  hr/d x  7 d
  continuous
  for 1-8 d
  8  hr/d x 90 d
24  hr/d x 180 d
  continuous
  for 7 d
24 hr/d x 10 d
Monkey
Rat

Monkey
Rat
Rat
Castleman et al., 197?
Evans et al., I976b

Eustis et al.,  1981
Moore and Schwartz,
    et al.,  1973
                                                        Mouse     Ibrahaaet al.,  1980
 Inflammation
                          0-2       8 hr/d x.7 d
                          0-2       8 hr/d x  7 d
                          0-5       2 hr
                          0.5       s hr
                          0.64       8 hr/d  x
                          0.8       4 hr
                          1 yr
                    Rat
                    Monkey
                    Rat
                    Rat
                    Monkey
                    Monkey
         Schwartz et al., 1976
         Castleman et al.,  1977
         Stephens et al., I974a
         Stephens et al., 19745
         Fujinaka et al., 1985
         Castleman et al.,  1980
Continued
inflammation with
remodeling of
centriacinar airways
and increased collagen
      0.5      24 hr/d x 14 d
      0.5       8 hr/d x 90 d
      0.5      24 hr/d x 180 d
      0-64      8 hr/d x 361 d
      1-0       8 hr/d x 18 mo
                    Rat
                    Rat
                    Rat
                    Monkey
                    Oog
         Last et al., 1979
         Boorman et al.,  1980
         Moore and Schwartz,  1881
         Last et al., I984b
         Freeman et ah,  1973

-------
                             VII-30
 reported  in rats  (0.26 ppm, 6  hr,  endotracheal  tube,  Boatman et
 al., 1974), mice  (0.5 ppm, 35  days,  Zitnik et al.,  1978),  and
 guinea pigs (0.5  ppm, 6 months, Cavender et al.,  1978).
 Inadequate data limit quantitative comparisons  of human  health
 effects with these reported for monkeys and rats, but a  rough
 equivalency of responses has been observed under similar exposure
 conditions between species.  Because all species tested  show
 similar morphological responses to O3 exposure,  there  is no
 reason to believe that humans exposed to O3 would not  respond
 similarly, although such effects may not necessarily  occur at  the
 same O3 exposure concentrations or averaging times of  exposure.
      Changes in lung structure of monkeys and rats tend  to
 decrease after extended exposure to 03/  although structural
 changes in the centriacinar region have been reported  after  long-
 term exposures of rats (Boorman et al.,  1980;  Moore and  Schwartz,
 1981; Barry et al.',  1983;  Crapo et al.,  1984),  monkeys (Eustis et
. al., 1981; Fujinaka,  1984;  Fujinaka et al.,  1985), and dogs
 (Freeman et al.,  1973).   While cell repair begins within 18 hours
 of exposure (Castleman et  al.,  1980;  Evans et  al., 1976 a,b,c;
 Lum et al.,  1978), cell  damage continues throughout long-term
 exposures but  at a slower  rate (CD,  p;  12-47) .'
      Increases of lung collagen content  in the  centriacinar
 interalveolar  septa  may  be  indicative of lung  fibrosis and, thus,
 of structural  damage.  Thus morphological  effect has been
 associated with biochemical changes in activity  of prolyl
 hydroxylase and in hydroxyproline  content,  both  related to  lung
 collagen content  and  fibrosis  (Last et al.,  1979;  Bhatnagar et
 al.,  1983).  Lung collagen  content increased after exposures to
 03  concentrations  as  low as 0.5 ppm  (Last et al.,  1979) and
 continued to increase during long- term  exposures  (Last et  al.,
 1984b).   Weanling  and adult rats exposed for 6 and 13  weeks,
 respectively,  and  young monkeys for one  year to  <  i.o  ppm 03  also
 showed  increased  collagen content  in  the lungs  (Last et al.,
 1984b).   In the Last  et al., (l.984b)  study, examination of  some
 of  the  exposed  weanlings and controls at six weeks post-exposure

-------
                              VII-31
  indicated a continued increase in lung collagen content, a result
  demonstrating that damage continued to occur during the post-
  exposure period.   Also,  there was little apparent difference in
  collagen levels of rat lungs for those animals exposed
  continuously vs.  intermittently.   This would suggest that
  intermittent periods  of  clean air were insufficient for recovery.
      The centriacinar inflammatory process  also continues during
  long-term 03  exposures and appears  to  be 'related to remodeling  of
  the centriacinar  airways  (Boorman et al., 1980;  Moore and
  Schwartz,  1981) and to increased  lung  collagen which appears
  mainly in  the centriacinar regions  (Last et  al.,  1979;  Boorman  et
  al., 1980; Moore  and  Schwartz, 1981; Last et al.,  I984b).   in
  addition there  is morphometric (Fujinaka et  al.,  1985),
 morphologic  (Freeman  et al.,  1973), and functional  (Costa et al
  1983; wegner, 1982) evidence  of distal airway  narrowing.
      Several new animal toxicology studies have  focused on effect
 of prolonged exposure and appropriate selection of an averaging
 time associated with respiratory damage from 03 exposure
 Rombout et al.  (1989)  conducted a study in which rats were
 exposed to 03 levels  of 0.13  to 2.0 ppm and  time of exposure
 ranged  from l to 54 hours.  With time course of protein and
 albumin concentration as  the endpoint,  Rombout et al. (1989)
 concluded that the response  model  indicates  a strong influence of
 time  on the response,  that increases with increasing
 concentration.   Van Bree  et  al.  (1989)  exposed  rats to 0.4 ppm
 for 12  hours  on  3  consecutive nights and  found  that acute
 pulmonary injury and  inflammation,  as measured  by lung lavage
 protein increase and neutrophil influx, were  induced by the  first
 exposure  and  showed full reversibility  in spite of continued
 exposure.   However, biochemical indices for cell  proliferation
 and lung tissue  repair continued to  show increases with
 subsequent  exposure.   A polynomial model has  been  developed  by
Costa et al.  (1989) to depict  lung injury from  the interaction of
03 concentration (C) and exposure duration (T).   The model was
based on lung fluid protein values in rats exposed to O3 levels

-------
                             Vi:C-32
 between  0.1  and  0.8  ppm for  2,  4  or  8  hours  and suggests that
 impact of  T  on 03 toxicity appears to  be C-dependent.   Further
 development  of these models  is  expected as efforts  continue to
 improve  understanding  of c x T  relationships.
     While the distal  airway narrowing and lesions  at  the
 junction of  the  conducting airways and gaseous  exchange region
 are similar  to the changes which  have  been found in lungs of
 cigarette  smokers  (Niewohner et'al., 1974; Cosio et al.,  1980;
 Hale et  al., 1980; Wright et al., 1983), there  is no evidence of
 emphysema  in the lungs  of animals exposed to only O3 based  on the
 currently  accepted definition for emphysema cited in the  CD (p.
 9-49).   Many of the  lung structural  changes that have  been
 reported for long-term  exposures to  <  1.0 ppm 03  in  several
 different  species are considered adverse and relevant  to  standard
 setting  if demonstrated  in humans at ambient exposure
 concentrations of 03.  Although the lowest O3 concentrations and
 shortest averaging times which could produce structural changes
 in human lungs is uncertain  at this time,  the staff  concludes
 that there is a need to  protect the public from  O3 exposures
 which may  induce such adverse effects.   Further  analysis  is
 needed to clarify how animal data should be used  to  estimate  lung
 structural changes in humans caused by O3  exposures.  The staff
 suggests that these data be used in developing a margin of  safety
 for the  primary standard.

     7.    Effects of Ozone on Host Defense Mechanisms  in
          Experimental Animals
     Respiratory systems of mammals are protected from bacterial
and viral infections by the closely interrelated particle removal
 (both mucociliary and phagocytic)  and immunological defense
systems.   Numerous factors such as poor nutrition, preexisting
disease,  and environmental stress may influence or alter these
host defense systems of individuals sufficiently to permit
development of respiratory infections.   Animal studies have

-------
                              VII-33
  indicated that exposure to O3 is one of those factors (See Table
  VII-3).   Both in viva (live animal)  and in vitro (isolated cell)
  studies  have demonstrated that O3 can affect the ability of the
  clearance and immune systems to defend against infection
  increased susceptibility to bacterial infection has been reported
  in mice  at 0.08  to  0.10  ppm 03  for a single  3-hour  exposure
  (Coffin  et al.,  1967;  Ehrlich.et al.,  1977;  Miller  et al.,  I978a)
  and at o.io ppm  03  for subchronic exposures  (Aranyi  et al
  1983).   several  related  alterations  of the pulmonary defenses
  caused by short-term and subchronic  exposures  to 03  include-   i)
  impaired  ability to inactivate  bacteria  in rabbits  and mice
  (Coffin et  al.,  1968; Coffin  and  Gardner, 1972;  Goldstein et  al
  1977; Ehrlich et al., 1979);  2) delayed mucociliary  clearance
  (Phalen et al., 1980; Frager  et al., 1979; Kenoyer et al.,  1981-
 Abraham et al., 1980); 3) immunosuppression  (Campbell and
 Hilsenroth, 1976; Aranyi et al.,  1983; Fujimaki et al., 1984)- 4,
 significantly reduced number of pulmonary defense cells in
 rabbits (Coffin et al., 1968; Alpert et al.,  197!);  and 5)
 impaired  macrophage  phagocytic activity, less macrophage
 mobility,  more fragility and membrane alterations, and reduced
 lysosomal enzymatic  activity (Dowell  et al.,  1970; Hurst et al
 1970;  Hurst and Coffin,  1971;  Goldstein et al., I971a,b;
 Goldstein et al.,  1977;  Hadley et al.,  1977;  McAllen et'al
 1981;  witz et al., 1983;  Amoruso et al.,  1981).  some of  these
 effects have been shown to occur in a variety of species
 including  mice, rats, rabbits,  guinea pigs, dogs,  sheep,  and
 monkeys.
     Other studies indicate  similar effects for short-term and
 sub-chronic  exposures of  mice  to 03 combined with pollutants such
 as S02, N02, H2S04 and particles (Gardner et al.,  1977;  Aranyi  et
 al.  1983; Ehrlich, 1980,  1983; Grose et al., 1980, 1982; Phalen
 et al., 1980; Goldstein et al., 1974).  similar to human
pulmonary  function response to 03, activity levels of mice
exposed to O3 has been shown to play a role in determining the

-------
                                     711-34
TABLE VII-
EFFECTS OF'OZONE ON HOST DEFENSE MECHANISMS  IN EXPERIMENTAL ANIMALS
_.. • ozone Exoosura
Effect concentration^ duration
Increased
susceptibility to
bacterial
respiratory
infection
Impaired ability
to inactivate
bacteria and
viruses
Delayed rauco-
ciliary clearance


Iiwnunosuppression




Impaired macrophage
function






0.08
0.08
0.1
0.1
0.1
0.5
0.5
0.5
0.4
0.3
1.0
1.2
0.1
0.59

0.8

0.25
0.25
0.5
0.5
0.5
0.62
0.99
1.0
3 hr
** III
3 hr
3 hr
5 hr/d x 103 d
3 hr/d x 3 mo
3 hr
** llf
3 hr
2 hr
4 hr
4 hr
2 hr
4 hr
5 hr/d x 90 d
continuous
for 36 d.
continuous
for 14 d
3 hr
3 hr
8 hr/d x 7 d
2 hr
3 hr
4 hr
17 hr
4 hr
                                                   Species
                                                   Mouse
                                                   Mouse
                                                   Mouse
                                                   Mouse
                                                  Mouse
                                                  Rabbit
                                                  Rabbit
                                                  Rat
                                                  Rat
                                                  Rat
                                                  Sheep
                                                  Rat
                                                 Mouse
                                                 Mouse

                                                 Mouse
                                                 Rabbit
                                                 Rabitt
                                                 Rabbit
                                                 Rat
                                                 Rabbit
                                                 Mouse
                                                 Mouse
                                                 Rat
                                                    Reference
                                                    Coffin et al., 1967
                                                    Miller et al., l978^
                                                    fhrlich et al., 1977
                                                    Aranyi et a?., 1983
                                                   Ehrlich et al.,
                                                   Coffin et al., 1968
                                                   Coffin and Gardner,
                                                   Goldstein et al., 1
                                                   Kenoyer et al.,
                                                   Phalen et al.,  1980
                                                   Abraham et al., 1985
                                                   Frager et al.,  197S
                                                  Aranyi et al.,  1985
                                                  Campbell and Hilse?iru
                                                  1976
                                                  Fujimaki et al., 1984
                                                  Hurst et al.,  1970
                                                  Hurst and Coffin,  l§?i
                                                  Oowell  et al.,  1970
                                                  Goldstein et al.,  13?7
                                                  Hadley  et al.,  1977
                                                  Goldstein et al.,  1S71
                                                  Goldstein et al.,  1971
                                                  McAllen  et al.,  1981

-------
                              VII-35
  lowest effective concentration which alters the immune defenses
  (Tiling et al., 1980).
       Although this large body of evidence clearly demonstrates
  that short- term and subchronic exposures to O3 can impair the
  inunune defense systems  of animals,  technical and ethical
  considerations have limited similar research on human subjects.
  Thus inferences have been drawn and models developed to assess
  the  relevance of animal data to humans.   For example,  animal
  endpoints  other than increased  mortality  should be appropriately
  compared to  increased morbidity in  humans,  such as the increased
  prevalence of respiratory illness in the  community (CD,  p.  12-
  50) .
      Based on current understanding  of physiology,  metabolism
  and  immune defenses,  it  is generally accepted that the  basic '
 mechanisms of  action  associated with defense against infectious
 agents are similar in humans and animals.  Similarities between
 human and rodent antibacterial systems have been discussed in
 detail (Green, 1984), but differences in lung structure and
 biological response will inevitably cause differences'in
 dosimetry,  sensitivity,  and endpoints of response.   in addition
 other factors such as preexisting disease, nutrition, presence  of
 other pollutants and environmental stresses (e.g. No2,  S02   PM
 high  temperature or humidity)  can influence the  effect of
 exposure  to O3 and  infectious  agents. Despite the  differences  in
 related factors which make precise estimation of human  response
 from  animal data difficult,  it is  reasonable to  hypothesize that
 humans  exposed to 03  could experience impairment of host defenses
 (CD,  p. 12-50).   The  staff recommends that these data be used in
 developing  a  margin of safety.

      8.  Extrapulmonary Effects
      Extrapulmonary effects which have been demonstrated  in
hvomans or laboratory animals following exposure  to  O3 include
alterations in red blood cell morphology and enzyme activity
cytogenetic effects in circulating lymphocytes, and subjective

-------
                             VII-36
 limitations in vigilance tasks.   Additionally,  animal toxicology
 studies provide limited evidence for cardiovascular,
 reproductive,  teratological,  endocrine system,  and liver
 metabolism effects.   This wide variety of extrapulmonary effects
 is probably caused by oxidative  reaction products of  O3.   Due to
 the high reactivity of 03 with biological tissue,  mathematical
 models  predict that only a small fraction of 03  actually  reaches
 the circulatory system (Miller et al.,  1985).
      Of the extrapulmonary effects reported,  cytogenetic  and
 mutational effects are probably  the most controversial (CD,  p.
 12-51).   Statistically significant increases in  frequency of
 sister  chromatid exchanges (chromosomal  alterations)  have been
 caused  by in vitro exposures  to  0.25  ppm for 1 hour  (Guerrero et
 al.,  1979),  suggesting a mutagenic response.  However,  in vitro
 responses are  not generally extrapolatable to in vivo responses,
 and homeostatic mechanisms are not represented.   Therefore,
 isolated  in  vitro exposure studies have  not  been used to  provide
 estimations  of risk  for exposure  to 03.   In vivo animal studies
 have  shown  significant  increases  in the  number of  chromosomal and
 chromatid aberrations  following 4  and 5  hour exposures  to 0.2 and
 0.43  ppm  03, respectively  (Zelac et al.,  I971a,b,; Tice et al.,
 1978).  However,  other  animal  studies  (Gooch et  al.,  1976) and
 controlled human  exposures to  O3  levels as high as 0.5 ppm have
 shown no  significant cytogenetic effects  attributable to  O3  (Merz
 et al.,  1975;  McKenzie  et  al., 1977; McKenzie, 1982; Guerrero et
 al.,  1979),  and epidemiology studies provide no evidence  of
 chromosomal  changes induced by ambient 03 (Scott and Burkart,
 1978; Magie  et al., 1982).  Thus, while the animal studies are
 suggestive of possible  cytogenetic  effects from 03, human studies
have not demonstrated such effects  for realistic ambient
exposures to O3.
     Limited hematological and serum chemistry effects data
 indicate that 03 or one of its reactive products can cross the
blood gas barrier and interfere with biochemical mechanisms in
human blood erythrocytes and sera  (CD, p. 1-132).  Much of the

-------
                              VII-37
  data is In Yi£ro and useful primarily in studying mechanisms of
  action.  Ozone-induced effects such as pentobarbital- induced
  sleep time alterations and hormone level changes have been
  reported (Graham et al.,  1981)  and cause-effect hypotheses made
  With regard to other extrapulmonary effects,  exact mechanisms
  remain to be elucidated and physiological significance for humans
  remains uncertain.   Nonetheless,  the body of  this evidence
  suggests that 03 can  cause  effects  distant from  the lungs  in
  animals,  and hence  possibly in  humans.   it is the staff's
  recommendation that these data  be used  in developing  a  margin of
  safety.

  B.  Pulmonary  Function  and  Symptom  Health Risk Assessment
      This section summarizes an assessment (Hayes et al.,  •
  1987a,b; Whitfield,  1988)  of risks  for two categories of'effects
  (pulmonary function and symptoms)  associated with attainment of .
 alternative 1-hr 03  NAAQS.   These  health risk estimates
 characterize acute responses based on results of l- .and 2-hour
 controlled human exposure  studies  reviewed in the CD.   While
 recent controlled human exposure research and field studies
 suggest that longer  exposures (6-8 hours) at  less strenuous
 levels of exercise may also pose a health threat, EPA  is awaiting
 completion of additional research  prior to extending the risk
 assessment to these  subchronic  exposures.  EPA also plans to
 pursue risk assessment for  chronic,  possibly  irreversible,
 effects on the lung  observed in  various  animal toxicology'studies
 upon  completion of important dosimetry and lung  injury research
 currently underway.

      1.   Overview of Lung Function and Symptom Risk Assessment
     The  objective of  the pulmonary  function/symptom risk
 assessment is to estimate the magnitude and extent of risk  to  the
most susceptible population  for these effects  (i.e., heavily
exercising individuals) while characterizing,   as explicitly as
possible, the range and implications of uncertainties in the

-------
                             VI1-3 8
 existing scientific data base.  While the risk assessment
 estimates should not be viewed as demonstrated health  impacts
 they do represent EPA's best estimate as to the possible
 magnitude and extent of risk for these effects given the
 available scientific information.  Although it does not cover all
 health effects caused by 03, the risk assessment is intended as a
 tool that may, together with other information presented in this
 staff paper and the CD and CDS aid the Administrator in judging
 which alternative 03  NAAQS  provides  an adequate margin of safety.
 Risk estimates for a number of urban areas,  the aggregated total
 for those areas,  and the methodology used to generate these
 estimates are described in  detail in Hayes et al.  (1987a,b)  and
 Whitfield (1988).
      The three major  types  of inputs to the  risk  assessment are:
      (1)  exposure-response  relationships  used to  characterize
 pulmonary function  and  symptom  effects  of O3  exposure  in  heavily
 exercising  individuals  developed  from data obtained  in three
 controlled human exposure studies:   Avol  et  al.  (1984), Kulle et
 al.  (1985),  and McDonnell et al.  (1983);
      (2)  distributions  of O3 hourly concentrations upon
 attainment of  alternative NAAQS and  under the  "as is"  situation
 obtained  from  the 03-NEM project  (Paul et al., 1986); and
      (3) distributions  of population  exposure, both  in  terms  of
 people exposed and occurrences of exposure, upon attainment of
 alternative  03 NAAQS and under the "as is- situation obtained
 from the O3-NEM analysis (Paul et al., 1986).
     Chapter V of this staff paper presents an overview of the O3
 exposure analysis methodology and summarizes exposure estimates
 for the aggregated 8-urban area population for alternative air
quality scenarios.   The pulmonary function and symptoms risk
assessment considers the same four scenarios used in the exposure
analysis:  0.08, O.io,  and 0.12 ppm,  daily maximum l-hour 03
NAAQS with an expected exceedance rate of once per year and the
"as is" situation.

-------
                              VII-39
       Risk estimates have been developed for both pulmonary
  function (as measured by FEVl decrement)  and lower respiratory
  symptoms (specifically,  cough, chest discomfort, and as a group
  of symptoms).   A health  effect,  or endpoint, can be defined in
  terms of a  measure of biological response and the amount of
  change in that measure thought to be of concern.  To permit EPA
  decision makers and the  public to examine the implications of
  alternative definitions  of  adversity,  risk estimates are
  presented for  a range of different  health endpoints:
       (1)  Pulmonary function.  TWO  different level  of 03-induced
           FEVi decrement: (a)  at  least  10 percent and;  (b) at
           least  20  percent.
       (2)  Lower  respiratory  symptoms.   Cough, chest  discomfort,
           and  lower respiratory symptoms  as  a group  for  two
           different severity levels:   (a) any symptoms  (mild,
           moderate, or severe) and  (b) moderate  or severe
           symptoms.
      Estimated exposure-response relationships derived from each
 of the three health studies  (Avol et al., 1984; Kulle et al
 1985; and McDonnell et al.,  1983)  are used separately to derive
 independent  risk estimates.   While the three studies are similar
 in enough respects (e.g., health endpoints, young heavily
 exercising healthy subjects,  similar 1-2 hour 03  exposures)  to
 make useful  comparisons,  there are enough differences in
 experimental protocol (e.g.,  i-hour  continuous exposure in Avol
 vs.  2-hour intermittent in Kulle  and McDonnell and differences in
 exact exercise  level and  exposure  concentration)  to  make
 statistical  combination of these data bases  undesirable
 Estimated  exposure-response relationships  for FEV, decrement and
 symptomatic  effects  used  in the risk assessment are  shown in
 Figures VII-2 through  VII-4 earlier  in this  staff paper.   These
 relationships were derived by fitting a  four-parameter  logistic
 function to  each  data  set and subtracting  out the effect  of
exercise alone based on the response at 0 ppm  (see Chapter 3 of
Hayes et al., i987a for details).

-------
                             VI1-40
      Uncertainty attributed to sampling error due to sample size
 considerations in the exposure-response relationships is
 reflected in the pulmonary function and symptom risk estimates.
 This uncertainty was estimated using a Bayesian approach
 described in Chapter 3 of Hayes et al. (I987a).  other sources of
 uncertainty due to differences in experimental  protocol subject
 population,  measurement error,  etc.  have not been quantitatively
 addressed.   The calculation and presentation of separate risk
 estimates for each of the three data sets provides a rough
 picture of  the degree of uncertainty due to  these other factors.
      Two distinct types of risk measures are provided by the O3
 health  risk  assessment.   The first measure,  "benchmark risk,"
 focuses on the probability or risk of unhealthful air.   The
 second  measure,  "headcount risk,"  focuses on the  number of people
 affected and number  of incidences of  a given health  effect
 considering  individuals'  personal exposures  as  they  go  about
 their daily  activities (e.g.,  from indoors to outdoors,  moving
 from place to place,  and  engaging in  different  exercise  levels).
     More specifically, benchmark risk  is the probability  that
 pollutant levels  will  be  sufficient to  trigger  a  defined health
 endpoint in  at  least  a fraction  of the  sensitive  population  if
 they were exposed.  The second measure, headcount risk, assesses
 the number of people  or the percent of  the sensitive population
 that would be adversely affected given  the normal movement and
 activity patterns of the population of  interest.  Headcount  risk
 also provides estimates of the number of incidences of adverse
 effects  there would be.  Staff believe that these risk measures
 taken together capture two important perspectives that should be
 considered in selecting an O3 primary standard that provides an
 adequate margin of safety.

     2.   Benchmark Risk Results
     For the 03 pulmonary function and symptom risk assessment
the benchmark risk is defined as the probability that, upon just
attaining a given 03 NAAQS or under "as is" conditions,  the

-------
  maximum hourly concentration will equal or exceed the level that
  would cause i, 5,  or 10% of the heavily exercising population to
  exhibit particular health endpoints 1 or more times per year
  The benchmark risk is estimated assuming the entire sensitive
  population is exposed while heavily exercising (defined as Regime
  3  in 03-NEM which  is  3  or more  10 minute  periods  in an  hour at a
  >  44 L/min).   As indicated  in the risk report:

                     ^
                      thi
      Benchmark risk is calculated. by combining exposure-response
 relationships and probability distributions of hourly O3 ambient
 concentrations,  either based on observed air quality data during
 1981-1984 (the particular year depending on area)  for the "as is"
 case or based on conditions of exact attainment of alternative
 NAAQS (0.08,  0.10,  and 0.12 ppm,  daily max hourly  average  1
 expected exceedance per year) .  The  benchmark risk model and more
 detailed discussion of the inputs to the model are contained in
 Hayes et al.  (I987a,b) .
      Benchmark risk estimates  are calculated  for the  10  urban
 areas shown in Table V-i  and are  presented  in Hayes et al
 C1987a,b).  Due  to  differences  in the  degree  of non-attainment  of
 the 03 NAAQS among the ten areas  (ranging from Los Angeles to
 Tacoma)  the benchmark risk associated with the "as is" scenario
varies significantly among the urban areas.  However, benchmark
risk estimates for a particular endpoint, data base, and choice
of benchmark case are very similar for a given NAAQS -attainment
scenario across most of the 10 urban areas.  Selected results for

-------
                             VII-42
 a single urban area, St. Louis, are presented here which are
 generally representative of benchmark risks for the other urban
 areas.
      Figure VII-5 presents benchmark risks for two different
 degrees of FE^ decrement:   at least 10% in the top chart and at
 least 20% in the bottom chart.  Risk estimates are shown as three
 groups  of bars, one group for each health data set used to
 generate exposure-response relationships.  Each group consists of
 four bars:   one bar for each of the three alternative 03 NAAQS
 analyzed (0.08, 0.10,  and 0.12 ppm, daily max hourly average,  1
 expected exceedance per year)  and one bar for the "as is"
 situation.   The height of the bar represents a 1%-benchmark case,
 that is,  "the probability that, at least  r percent or more of the
 heavy exercisers would respond with the  specified health endpoint
 than under  background  conditions, one or more times during the o'3
 season"  (Hayes et al.,  1987a,b).  The top of the  vertically-
 shaded portion corresponds  to  a 5%-benchmark case and the top  of
 the  slant-shaded portion  corresponds  to  a 10%-benchmark  case.
      As  an  example of  how to read the benchmark risk  figures,
 consider  the  > 10 percent FEVl  decrement  case  in  Figure  VII-5
 (top):
      Under  exact  attainment of  the  current 03 NAAQS  (0.12 ppm) ,
      the  probability that at least  a  10 percent FEVl decrement
      would  occur  in at  least 5  percent more heavy  exercisers than
      at background, at  least once during  the 03 season,  is nearly
       H  2   *     V01 data Set' about °'1 for ^e  Kulle  data set,
      and  about  0.5 for  the McDonnell data set  (Hayes et  al
      1987a,b, p.  4-7).                           *

     Using the  same approach,  Figure VII-6 presents benchmark
risk estimates  for any  chest discomfort  (top chart) and  for
moderate  or severe chest  discomfort (bottom chart).  It  should be
noted that risk estimates for  the Avol data bases are for an
aggregate lower respiratory symptom score that reflects  chest
discomfort,  cough, and other lower respiratory symptoms.
     Figures VII-5 and VII-6 illustrate the importance of certain
key decision parameters in interpreting risk estimates.  Both

-------
                             VII-43
        1.0
£ 0.6

jt

o



I 0.4
       0.0
                        BENCHMARK RISK

                          => 10%, Heavy Exercise —
                 .12
1
                    •
                Avol
                                                   A«-ta
                                 .tz.
                              .10
                                              .10
                                           M
1
   1

                               Kulta          McDonnell



                 Alternative Ozone Standards (ppm)
                                                           * IX Uer. R«*p«fMMng
                                                             than of Baekareund
                                                             (Total Bar
                                                           *5X Mora ft«tpondlna
                                                             (N»rt Un« Down)

                                                           *10X Uer.
                       BENCHMARK RISK

                  D(FEV1) => 20%; Heavy Exercise


0.8
jt
m
•c u.o
!
j...
0.2



0.0


48

-
-

„.

TTTT

.10







. s

^§§
I
i
1

Oo^
\s\^
SXSS
£ ^






.10

J>a ]
\ 1 1 1 1





.12
••^^H



[mil


A«-h




1 1|
Ul
^







JOB
1 — 1
1





.10








.12




(III




1
1
\ssv
NSN
ooo
\\V












n *1X Uor* Responding
than at Background
(Total Bar Height)
|Q] J5S More Responding
(Next Line Down)
J£3 *10S More Responding






               Avol
                               Kulle
                                            McDonnell
                 Alternative Ozone Standards (ppm)


                           St.Louls


FIGURE  VII-5.Benchmark r1sk in St>  Louis for  FEV1 decrements Qf

^ 10% and > 20%,  under heavy exercise, for three exposure-response
data sets (Avol,  Kulle, and McDonnell).


 Source:   Hayes et al.,  1987

-------
              1.0
             0.6
             0.2
                              BENCHMARK RISK

                   	Any Chest Discomfort. Heavy Exercise	

                       _.t2 »«-*«

             0.0"™


1

I
                                        .12
                                     .10
                                           I
                                           I
                                                         .12
                                                      .10



                                                    McDonnell


                       AlUmatfv* Ozone Standard* (ppm)
         * 1X Mar* Responding
          tho« of Boeicareund
          CTetal Bar Height)

         *9X Mere Responding
          (Nert 4lne Down)

         *10X Mere Responding
       Aval data Is for lower
       respiratory symptoms
                              BENCHMARK  RISK

           — Moderate or S«ver« Chest Discomfort, Heavy Exercise	
             1.0
            0.6
          I""
            0.2-
            0.0
                                                        .12
                                                     .10
                                                  jM
                      .12
                                       .13
                                    .10
1
%
                     Avol
                                     Ku|to
                                                    McDonnell


                      Alternative Ozone Standards (ppm)


                                  St. Louis
                                                                   i * 1X More Responding
                                                                     than at Background
                                                                     (Total Bar Height)
                                          *SX Uors Responding
                                           (Went Un« Down)

                                          klOX More Responding
      Aval data Is for lower
      respiratory  symptoms
           FIGURE vn-6.Benchmark risk  in St.  Louis for  chest  discomfort
           symptoms  (any and moderate/severe), under heavy exercise,  for
           three exposure-response data sets  (Avol, Kulle, and McDonnell)
Source:   Hayes  et al.,  1987

-------
                              VII-4 5
  selection of endpoint definition (e.g., >io% or >20% decrement in
  PEVi)  and n%-benchmark case (e.g.,  at least 1,  5,  or 10% of the
  sensitive population which must experience the health effect)
  significantly affect the risk estimates.   In addition,  comparing
  risks  across health data bases results in significant
  differences.   FEV^  risk estimates for a given benchmark
  definition,  say 1%,  are highest for the Avol data  base  and lowest
  for  the  Kulle-data  base,   in contrast,  for moderate/severe chest
  discomfort and lower respiratory symptoms,  risk estimates are
  highest  for  the McDonnell  data base and lowest  for the  Avol data
  base.

      3.  Headcount Risk Results  "
      For the 03 risk assessment headcpunt risk is characterized
 by calculating- the expected headcount, that  is, the  expected
 number of people experiencing a defined effect and the  expected
 number of incidenbes of that effect projected to occur during the
 03 season,  given that a particular  NAAQS is just attained or
 under "as is" air quality conditions.  Headcount risk estimates
 the risk posed to the sensitive population (in this case
 individuals engaged in heavy exercise) as  they, go about their
 normal  activities.
      A  major  input  to the headcount  risk model is the series of
 population exposure  distributions for "as  is" air quality and the
 three alternative NAAQS analyzed by  EPA.   Using  available
 exposure  estimates,  headcount risk estimates were calculated for
 eight of  the  ten urban areas  listed  in Table V-i.   NO headcount
 risk  estimates  are calculated for Chicago  and New York,  since EPA
 is  in the process of  revising these  estimates. Appendix  C of
 Hayes et al.  (1987a,b)  presents  headcount  risk estimates for each
 of the  8 urban  areas.
      Risk estimates presented here are an  aggregation of  the
results from the  8 urban areas.  The total population living in
the 8 urban areas is approximately 25.9 million.   Of this
population, 9.3 million people  (or 36%) are estimated to exercise

-------
                             VII-46
 sufficiently heavily to reach NEM Exercise Regime 3  (3 or more
 10-minute periods in an hour at heavy (Ve > 44 1/min) which could
 include one 10-minute period at very heavy (Ve > 64 1/min)
 exercise).
      Figure VII-7 presents risk estimated for each of the three
 health data sets for two levels of FEV1  response:   at least 10%
 decrement on top and at least 20% decrement on the bottom for the
 three alternative O3 NAAQS analyzed (0.08,  o.io,  and 0.12 ppm,
 daily max hourly average,  1 expected exceedance)  and the "as is"
 situation.   Similarly,  Figure VII-8 presents  expected headcount
 estimates for any chest discomfort (top  chart)  and moderate or
 severe chest discomfort (bottom chart).   Again,  it should be
 noted'that risk estimates  for the Avol data base  are for lower
 respiratory symptoms which include chest discomfort  as well as
 cough and other symptoms.   The  complete  headcount  risk results
 are  contained in Hayes  et  al.  (1987a,b),  including risk estimates
 for  cough and for expected number of  incidences of each of the 6
 health endpoint definitions examined, as  well as estimates for
 each of the  8  urban  areas  examined.
      In both  Figures VII-7  and VII-8  the  expected  headcount  in
 millions  of people is indicated  by the filled-circles  and
 triangles and  90% credible  intervals  are  displayed as  vertical
 bars.   The credible  interval reflects uncertainty  in the  expected
 headcount attributed to uncertainty in exposure-response
 relationships due to sample size considerations.  The  90%
 credible  interval can be interpreted as a 0.9 probability that
 the  headcount lies within the interval,  and a O.io probability
 that  it falls outside of the interval.
     As with the benchmark risk results,  for FEVX decrement the
Avol data base results in the highest headcount estimates and
Kulle shows the lowest estimate and for chest discomfort  the
McDonnell data base results in the highest headcount estimate and
Avol shows the lowest estimate.   One also observes a greater
reduction in headcount estimate going from the "as is" situation
to the 0.12 ppm NAAQS than in going from  0.12  ppm to a 0.10 or

-------
        23
    |  15.0

    O


    0

   3:  10.0

   T)
    X
   u
        S.O
        0.0
                           VII-47


                    HEADCOUNT RISK (People)

                	D(rEV1) => 10%. Heavy Exercise



Ai

A.-*
•*
T:
•
. mi
X
<
.:
«
4
0
•

<
-<•




               Avcl           »•"•        McDonn.M



                 Alternative Ozon* Standards (ppm)
      23
   §  13.0

   O


   O

  X  10.0
  O
  4>   S (
  a   5-(
  X
  LJ
      0.0
                  HEADCOUNT RISK (People)

              — 0(FIV1) => 20%, Heavy Exercise
              Avo1

A.-4,


A*
At-t.
.12
.10 r
Til r-?f
M T° f '
-ta


                                        WcDonn.ll



                Alternative Ozone Standards (ppm)



                   Eight City Aggregation
                                                             
-------
                    HEADCOUNT RISK  (People)

              	Any Chest Discomfort, Heavy Exercise	
0 2S.O
X
| 20.0
O
0
T3 1S.O
O
«
I
•O 10.0
"o
0
0- 5.0
X
Ld
0.0


A*

A*
'
•
-te
A*

.1
JO '
.u -2?
.10 f T
TTl I-
i <
2 '
i
H«
'
.
' M




0


2



-ta .





                Aval
                              Kulte
                                           UcOonn«ll



                  Alternative Ozooe Standard*  (ppm)
                                                                MX Crwdlbto
                                                                Interval
                                                             Avol data U for low«r
                                                             rwiplratofy •ymptom*
   O


   x
   T>
   »

   T)
   e>
   Q.
   X
                   HEADCOUNT RISK (People)

      — Moderaia or Savere Chest Discomfort, Heavy Exercise —
23.0
Headcouni


8
b
5
b
       0.0
                                              .12
                                           .10
                           A*-to
                   A.-*.

               Av«<
                                          McDonnell


                 Alternative Ozone Standards (ppm)


                    Elghi City Aggregation
                                                      Avol data l« for lower
                                                              •ymptoms
FIGURE vn-8.  Expected  headcount  (chest discomfort) aggregated  for
eight U.S. urban areas (number of heavily  exercising  people responding
during the ozone season).

Source:  Hayes et al.;  1987

-------
                               VII-49
  0.08 ppm NAAQS.  However,  one must note that the expected
  headcount in Figures  VII-7 and VII-8 for the "as is" case areas
  dominated by LOS Angeles•  contribution to headcount risk and
  should not be viewed  as representative for the  nation.   For
  example,  in Figure VII-7,  the expected headcount for FEV
  decrement > 10% based on the Avol et al. (1984)  study is about
  1.7  million people, of which 1.3  million people  are from Los
 Angeles.    Headcount risk  estimates can also be  expressed as a
 percentage of the heavily  exercising population.  Table VII-4
 presents  the range of mean percentages  of heavy  exercisers
 responding 1 or more times per year upon attainment  of  3
 alternative  NAAQS and for the "as  is- situation  for  each of  the 6
 health endpoint definitions examined.   The  range is  due  both to
 variation  in the shape of air quality distributions  in the 8
 urban areas  and to  differences in  exposure-response relationships
 based on the  three  health data bases  (Avol, Kulle, and
 McDonnell).  Again,  it is  important to recognize the influence  of
 the  Los Angeles  estimates  when considering the  »as-is» case
 Removing Los Angeles from  the "as-is" column  would for example
 reduce the upper end of the range  from 36.3%  to  14.0% for FEV  >
 i n&                                                             1 —
10%.
Health
Bndpoint
           TABLE VII-4  Percent of Heavy Exercisers Responding
                 Under Alternative Air Quality Scenarios

                                  Mean Percentages of Heavy Exercisers
                                  Responding Under NAAQS Attainment
                                  Across Eight U.S. Urban Areas
                                   0.12 ppm    0.10  oom    0^,08
                Responding ,»,-!•.)
        >20%
Cough    Any
        Mod/Sev
Chest    Any
Discom-  Mod/Sev
fort
                1.1 to  36.3
                0.5 to  20.7
                3.1 to  41.8
                0.2 to  27.7
                2.7 to  20.4
                0.2 to  18.7
                                  0.5 to 4.2
                                  0.2 to 2.1
                                  0.9 to 18.8
                                  0.1 to 4.8
                                  0.8 to 8.5
                                  0.1 to 7.1
0.1  to 3.0
0.0  to 1.3
0.2  to 15.2
0.0  to 2.9
0.1  to 7.0
0.0  to 5.8
0.0 to 1.9
0.0 to 0.7
0.1 to 10.1
0.0 to 1.4
0.1 to 4.8
0.0 to 3.8

-------
                             VII-50
 4.   Caveats and Limitations
      A number of assumptions and limitations should be kept  in
 mind in interpreting results of the pulmonary function and
 symptom risk assessment.  Extrapolation of results of the Avol,
 Kulle, and McDonnell data sets to the heavily exercising
 population at large is affected by a number of considerations.
 These include the following:
      (1)  Exercise group.  The CD defines heavy exercise as lung
 ventilation rates of 44-63 L/min and very heavy exercise as >64
 L/min.  The Avol study (57 L/min)  corresponds to the mid-to-upper
 portion of the heavy exercise range (54 L/min is the mid-point).
 The McDonnell study (65 L/min)  falls just in the lowest portion
 of the very heavy exercise range,  with the Kulle study (68 L/min)
 some 5 percent higher.   Exposure-response relationships obtained
 from the  three data sets have been assumed to represent exposure-
 response  in all  heavy exercisers.   Additional research is  needed
 to determine the extent to which exposure-response  relationships
 observed  at 57,  64,  and 65 L/min change at lower exercise  levels
 such as 44-56  L/min.  To  the extent that exercise rates  among
 many heavy  exercisers are lower than in the  subject  studies,  lung
 function  and symptom headcount risks may be  overstated.  The
 extent to which  this  is so is unknown  and  must be regarded as an
 additional  risk  assessment uncertainty  (Hayes et  al.,  1987a,b).
      Since  the December 1987 CASAC  meeting,  an addendum to the
 lung function and symptom risk assessment  has been prepared
 (Whitfield,  1988) which combines the estimated exposure-response
 relationships from the same three studies discussed previously
 with the NEM very heavy exercise exposure estimates to generate
 expected headcount estimates.  The resulting expected headcount
 estimates for very heavy exercisers are approximately 100-150
 times smaller in magnitude than the heavy exercise expected
 headcount estimates.  This is due principally to the fact that
 the population of very heavy exercisers is much smaller (by about
two orders of magnitude) than the population of heavy exercisers.

-------
                             VII-51
      (2)  Interaction between o^ and other pollutants.   The health
 studies used in the risk assessment involved only 03 exposure.
 It is assumed that the health effects of interest in the real
 world where other pollutants are present are due solely to O3.
 While controlled human exposure studies have not consistently
 demonstrated enhancement of respiratory effects for O3 when
 combined with SO2/  NO2,  CO,  H2SO4/ or other aerosols, there is
 some animal toxicology research suggesting additive or possibly
 synergistic effects.   (See Section  VII.C.5 for further discussion
 of interactions with other pollutants).
      (3)  Reproducibility of Q^-induced  response.   It is assumed
 that 03-induced respiratory responses are  reproducible for
 individuals.   The CD cites both Gliner  et  al.  (1983) and
 McDonnell et al.  (1985a)  in concluding  that respiratory effects
 of 03 are highly  reproducible.   Analysis of the Avol and Kulle
 data sets by Hayes  et al.  (l9-87b) also  supports the
 reproducibility of  individual  responses.
      (4)  Age.   The  risk  assessment  has  been applied to all
 heavily exercising  individuals regardless  of age.   However,
 controlled human  exposure  and  recent field epidemiology studies
 in children have  reported  pulmonary function,  but  not
 symptomatic,  effects  for O3-exposures.  Therefore,  the  headcount
 symptomatic effect  estimates which  rely on population  exposures
 that  include  children may  overstate symptom headcount  estimates.
 Pulmonary function  risk estimates are not  affected, and the lack
 of  apparent symptoms  does  not  mean that biological  processes
 associated  with O3 symptoms  in adults are not also present in
 children.
      (5)  Sex.  There  is some limited evidence that women may be
more responsive to O3 in terms of FEVj impairment than  men.  The
data sets used here to derive  exposure-response relationships
 involved mostly male  subjects.  To the extent that women are more
responsive than men, risk estimates may be understated.
      (6) Smoking status.  There is some limited evidence that
smokers may be less responsive to 03 than nonsmokers.  The risk

-------
                             VII-52
 assessment was applied to all heavy exercisers regardless of
 smoking status.   To the extent that smokers are less responsive
 than nonsmokers,  risk estimates may be overstated.
      (7)  Attenuation or enhancement-. Of response.   The risk
 assessment assumes that the 03 -induced response in  any particular
 hour is not affected by previous O3  exposure history.   The extent
 of  attenuation and/ or enhancement of O3 -induced responses due to
 previous O3  exposures cannot  be addressed quantitatively  and must
 be  regarded as an additional  uncertainty in interpreting  the risk
 estimates.
          Exposure  and  air quality estimates.  A major  input  to
the headcount risk is  the O3 exposure analysis estimates  for the
heavily exercising population.  Uncertainties about human
activity  patterns  and  the procedures used to estimate  03
concentrations upon attainment of alternative standards,  as  well
as other  uncertainties about the exposure analysis model  and
inputs to the model, must be regarded as additional uncertainties
in interpreting the headcount estimates.  As noted in  Chapter  V,
ASB plans to use a  different version of O3-NEM in the  future to
address multiple-hour  O3 exposures and to more explicitly analyze
various exposure-related uncertainties.  When new O3 NEM
estimates are available, ASB plans to revise the risk  assessment
discussed in this section, assuming contract funds are available
for that  purpose.   The benchmark risk estimates are affected by
uncertainty in projecting O3 concentrations upon attainment of
alternative NAAQS at the "critical" monitor.

C.   Related Health Effects Issues
     The purpose of this section is to provide an overview of
important issues related to the health effects of O3 and other
photochemical oxidants.  Some issues such as attenuation of
effects and the acute/ chronic relationship involve more than one
health effect,  while other issues involve multiple pollutants.
Detailed discussion of the individual studies associated with
these issues can be found in chapters 9 to 12 of the CD.

-------
                              VII-53
       1.   Adverse Respiratory Health Effects of Acute Ozone
            Exposure
       What constitutes an adverse respiratory health effect for
  acute exposure to 03 has been a matter of some controversy and
 •diversity of opinion.  Because the NAAQS for 03 is intended to
  protect the population from exposure to 03 levels  which  might
  produce adverse health effects,  it is important to identify the
  adverse effects of 03.   Although incapacitating effects  or
  irreversible damage are widely accepted as being adverse,  that
 type  of response has not been reported for acute exposures of
 humans  to  O3  levels  near the  current  standard of 0.12 ppm.
       Public  health  organizations,  such  as  the American Thoracic
 Society (ATS),  have  developed general guidelines (ATS, 1985)  as
 to what constitutes  an  adverse respiratory health  effect with
 respect  to interpretation of  epidemiological studies.  While
 recognizing that perceptions  of  "medical significance" and
 "normal activity" may differ  among physicians and experimental
 sub3ects, the ATS (1985) defines adverse respiratory health
 effects as "... medically significant physiologic or pathologic
 changes generally evidenced by one or more of the following-   m
 interference with the normal activity of the affected person or
 persons, (2)  episodic respiratory illness,  (3)  incapacitating
 illness, (4)  permanent respiratory injury,  and/or (5)  progressive
 respiratory dysfunction."  Although acute 03  exposures of human
 subjects have been associated  only with the first of the  above
 effects,  animal studies suggest that longer-term O3  exposures  may
 cause  more  serious health effects as well.
     The most commonly reported and well-established respiratory
 health effect  of acute 03  exposure,  as discussed earlier in this
 chapter,  is reduction in lung  function.   In order to address the
 adversity or clinical  significance  of this response, however,  it
 is important to  consider:  (1)  the magnitude of a functional
 change on a test-specific basis  (e.g., FEV, FVC); (2) the
duration or persistence of the effect; (3) the respiratory
symptoms which are associated with functional changes; and  (4)

-------
                            VII-54
 limitation of  activity  which might  result  from functional  losses
 and  respiratory  discomfort.
      In  conjunction with health  scientists of  the  Office of
 Research and Development, OAQPS  staff has  developed  Table  VII-5
 which shows a  gradation of individual response to  acute  03
 exposure.   While Table  VII-5 is  not intended to imply  exact
 quantitative relationships, it does categorize three variables as
 being mild, moderate, severe, or incapacitating depending  on
 severity of response.   Relationships among the response  variables
 not  only assist  in judging the severity of effects but can  be
 used as  a  general framework for  assessing  adversity  of effects
 for  acute  03 exposure.
      During the  CASAC meeting of  December  14-15, 1987  and
 December 14-15,  1988, and in written comments  received since, a
 wide  diversity of opinion has been  expressed regarding what
 should be  considered an adverse  respiratory^health effect.  Most
 comments indicated that Table'VII-5 was generally accepted  as *
 reflecting  the range of effects  associated with acute  (1-2  hr)
 exposures to O3.   At the CASAC (1987) meeting some CASAC members
 expressed the belief that either  limitation of  activity  or
 symptoms could be considered the primary determinant of  adversity
 while others believed the more objective spirometry measurements
 were more appropriate.  The point at which an effect can be
 described adverse was discussed  in some detail.  Some  CASAC
members  felt that individuals would experience adverse effects
when O3  exposure induced any  of  the responses categorized as
moderate.  Other CASAC members believed that adverse effects
would not be experienced until 03 induced severe effects.  A
point of clarification was made that Table VII-5 refers only to
healthy  individuals rather than to persons with respiratory
 impairment.  Individuals with preexisting respiratory disease may
experience adverse effects even when encountering responses
categorized as mild for healthy persons.   Also, it was pointed
out at the CASAC (1988)  meeting that because children show few,
 if any,  symptoms when exposed to ambient O3 levels,  it would be

-------
TABLE VII-5.
                                        HEALTHY I(IDIVIDUALS
 RESPONSE

CHANGE IN
SPIHOMETRY
FEVj Q. FVC
MODERATE

10-20%
                                                                SEVERE

                                                                20-40%
                                                                INCAPACITING

                                                                >40jt
     DURATION
       OF
      EFFECT
    SYMPTOMS
   LIMITATION
      OF
    ACTIVITY
   COMPLETE
   RECOVERY
  IN <30 HIN
MILD TO MODEHATE
    COUGH
     NONE
COMPLETE
RECOVERY
IN <6 HR
                                      Mlin  TO MODERATE
                                      COUGH. PAIN ON
                                      DEEP  INSPIRATION.
                                      SHORTNESS OF
                                      BREATH
                                 FEW  INDIVIDUALS
                                 CHOOSE TO
                                 DISCONTINUE
                                 ACTIVITY
                                                       COMPLETE RECOVERY
                                                         IN 24 HOURS
                                                      REPEATED COUGH.
                                                      MODERATE TO SEVFRE
                                                      PAIN ON DEEP
                                                      INSPIRATION AND
                                                      SHORTNESS OF BREATH;
                                                      BREATHING DISTRESS
                                        SOME INDIVIDUALS
                                        CHOOSE  TO
                                        DISCONTINUE
                                        ACTIVITY
                                      RECOVERY  IN
                                      >24 HOURS
                                      SEVERE COUGH.
                                      PAIN ON DEEP
                                      INSPIRATION. AND
                                      SHORTNESS OF
                                      BREATH; OBVIOUS
                                      OISTRESS
                                      MANY INDIVIDUALS
                                      CHOOSE  TO
                                      DISCONTINUE
                                      ACTIVITY
                                                                   <
                                                                   M
                                                                   M
                                                                   I

-------
                             VII-56
 inappropriate  to  recommend  that  all  categories  of  response be
 seen  in  children  before  describing the  effect as adverse.
      Based  on  these  comments of  CASAC  (1987, 1988),  staff
 recommends  that the  responses identified as  "mild"  for healthy,
 adult response to 03 not be  considered  an adverse respiratory
 health effect.  Mild individual  responses to O3 exposures  are
 viewed as involving  only 5 to 10% decrements in spirometry which
 last  30  minutes or less accompanied  by  only mild to  moderate
 cough and no limitation of activity.  Mild response  probably
 would not be considered medically significant and should not
 interfere with normal activity of most  individuals.  Staff
 recommends, however, that moderate,  severe, and incapacitating
 responses be considered adverse  respiratory health effects.
 These categories  of  response  are more likely to be considered
 medically significant and could  interfere with normal  activities.
 For example, some adults with preexisting respiratory  disease or
 heavily  exercising healthy adults who experience a moderate
 response (i.e., lung function loss of 10-20% which persists for
 up to 6 hours accompanied by multiple symptoms)  would  tend to
 curtail activity.   Due to lack of a clear consensus and the wide
 diversity of CASAC (1987,1988) opinions, staff recommends  that
 all criteria specified for moderate response be met for an effect
 to be deemed adverse in healthy adults.   The adverse nature of
 longer-term (6 to 8 hour) and chronic effects of O3 will be
 discussed following completion of future research which addresses
 these effects.

     2.   Attenuation of Acute Pulmonary Effects
     Attenuation of acute pulmonary response to 03  after repeated
daily exposures to 03 is  a well-established and  well-documented
phenomenon.   Until recently,  descriptive terms other than
attenuation have commonly been used to describe the response,
such as "adaptation" and "tolerance".  These terms  imply a
reduced impact of  repeated  exposure to 03 whereas recent evidence
suggests that lung injury continues during  the process of

-------
                              VII-57
  attenuation.  A thorough review of the large body of supporting
  evidence can be found in the CD (p.  10-47 to 10-60)  and CDS
  Functional decrements are generally greater on the second
  exposure day;  by the fourth or fifth day of exposure,  small or no
  decrements are observed (Farrell et  al.,  1979; Horvath et al
  1981;  Kulle et al.,  I982b;  Linn et al.,  1982).  Attenuation of
  functional response  to a particular  03 level  does  not  attenuate
  response to higher O3  levels,  nor is the  attenuation process
  permanent.   Subjects repeatedly exposed to  O.2 ppm 03  for three
  consecutive days  exhibited  attenuation to that level of  03  but
  showed no  attenuation  of response to higher (0.42  or 0.50 Ppm)  o
  levels (Gliner et al.,  1983).   once full  attenuation is  achieved
  it does not  continue for more  than 3 to 7 days  for most  subjects
  (Horvath et  al., 1981; Kulle et al., I982b; Linn et al., 1982)
 however,  partial attenuation has been shown to persist for  as  '
 long as 2 weeks (Horvath et al., 1981).  Attenuation of  symptoms
 correlates with magnitude of functional response and was shown  in
 one isolated statistical result to last as long as 4  weeks  (Linn
 et al., 1982).   Thus  individuals living in areas with high O3
 peaks may develop attenuated response to  repeated lower O3  levels
 and still respond to  peak exposures,  but  after exposure to O3
 ceases the attenuated response eventually will return to normal
      Evidence reviewed in the CDS suggests the possibility  that
 ambient oxidant exposure during summer  months  produced  an
 "adaptation" response which  persisted in human subjects for
 several months  (Hackney et al.,  1988).  The  authors suggested
 that  allergy or atopy may be a  risk factor for excess response
 and,  further, that nonresponders could be  at increased  risk  for
 chronic health effects  of cumulative ambient O3 exposure  since
 they would  be less likely to avoid such exposures due to  lack of
 symptomatic adaptation  or attenuation of o3-induced deficits with
 sustained but not worsening protein accumulation in lavage fluid
 (Costa et al., 1989).   However, histopathology of the animals
appeared to worsen and evolve from an acute to a more chronic
inflammatory pattern.

-------
                             VII-58
      Attenuation of functional or symptomatic response does not
 necessarily imply attenuation of morphological or biochemical
 response.   As indicated in the CD (p.  12-42),
      Responses to 03/ whether functional,  biochemical   or
      morphological,  have the potential for undergoing  changes
      during repeated or continuous exposure.   There is interplay
      between tissue inflammation,  hyperresponsiveness,  ensuing
      injury (damage), repair processes,  and changes in response.
      The  initial response followed by  its  attenuation  may be
      viewed as sequential states in a  continuing  process of luna
      injury and repair.                        .                 y


 Public health implications of attenuation,  therefore,  are that
 this  response is not a  protective  mechanism but may in  fact
 result in an increased  03  dose to  the deep  lungs and potentially
 cause greater tissue damage by permitting  individuals to exercise
 outdoors during elevated  03  episodes.

      3.  Relationship Between Acute  and  Chronic Effects
      In order to better assess attenuation  of pulmonary
 responsiveness  to 03 and other effects of O3 which may not be
 reversible,  it  is important  to understand the relationship
 between acute  and chronic  effects.   Although animal research
 provides some  evidence,  little is  known with certainty regarding
 either long-term implications of repeated acute 03 exposures or
 chronic effects  of prolonged  exposure of humans to 03.   Human
 chamber studies  involving  long-term  exposures to 03 have not been
 conducted due to concern for  serious health effects which might
develop in subjects.
     Various pulmonary effects demonstrated in animal studies
suggest that recovery from chronic exposure to O3  is not complete
even after an extended period.  Monkeys exposed to 0.64 ppm O3
for as long as a year (8 hr/day,  7 days/wk) continued to show
either statistically significant or substantial decreases in lung
function (e.g. static lung compliance)  even after  a 3 month post-
exposure period  (Wegner, 1982).  This was interpreted as
suggesting recovery was  not complete even after three months.  In

-------
                             VII-59
 several studies, increases in lung collagen content occurred
 after short- and long-term exposures to 1.0 ppm 03 (Last et al.,
 1979; Last and Greenberg, 1980; Last et al., 1981; Last et al.,
 1984b).  Adult and weanling rats exposed for 6 and 13 weeks and
 monkeys exposed for one year to 03  below 1.0 ppm showed increased
 lung collagen content,  and even six weeks post-exposure weanling
 rats continued to show increases in collagen (Last et al.,
 1984b).  This suggests that damage  continued during the post-
 exposure period.
      Even though the above and numerous other animal studies show
 structural changes caused by repeated short-term and long-term 03
 exposures,  there is no evidence of  emphysema in animals exposed
 to 03 according  to the  CD (p.  9-52).   However,  studies  of
 Bartlett et al.  (1974)  and Costa et al.  (1983)  provide  evidence
 that the recoiling force of the lungs of rodents is reduced by
 chronic exposure to O3.   Reevaluation of three  studies  (P'an  et
 al.,  1972;  Freeman et al.,  1974;  Stephens et al.,  1976)  cited in
 the 1978 criteria document (U.S.  EPA,  1978),  which reported
 emphysema in animals after prolonged  exposure to 
-------
                             VII-60
 possible that the resting, nasal-breathing rats exposed to 0.7
 ppm 03 may be receiving a lower O3  dose to the lungs than
 exercising, oral breathing humans exposed to lower O3 levels.
 Pickrell et al.  (1987)  observed that rats exposed to l.l ppm 03
 for 19 hours per day for 11 days had a 40% increase in total lung
 collagen and developed interstitial pneumonia, significant weight
 loss,  and pulmonary fibrosis based  on examination 2 months after
 exposure was initiated.   A lower level of O3  (0.57 ppm)  did not
 produce these effects,  thus indicating that these effects may be
 concentration or dose dependent.  Furthermore, chronic exposure
 of  rats and monkeys to  > 0.40 ppm O3  causes a  concentration-
 dependent peribronchiolar inflammatory response (Barr et al.,
 1988;  Moffatt et al.,  1987).
     In a chronic exposure study, Gross et al.  (1989)  showed  that
 rats exposed for 12 months to an. episodic profile of 03  (2  hours
 at  0.25 ppm 03 peak over  9  hours; 0.19  ppm  03  integrated
 concentration) exhibited (1)  functional lung changes indicative
 of  a stiffer lung;  (2) biochemical  changes  suggestive of
 increased antioxidant metabolism; and  (3)  no observable
 immunological  changes.   Other recent  studies indicate lower
 levels  of 03 may be  responsible for epithelial  injury.   Barry et
 al. (1985)  found that rats  exposed  to  both  0.12 and  0.25 ppm  03
 for 12  hours per day over  a six week period showed statistically
 significant  concentration  dependent changes in  alveolar  type  l
 epithelial cells suggesting that low concentrations  of 03 may
 cause a chronic epithelial  injury in the proximal alveolar
 region.   For chronic 03 exposures of 0.12 - 0.30 ppm, a lesion is
 evident at the junction of the conducting airways and the gas
 exchange  regions of  the lungs, characterized by cell population
 shifts along with interstitial inflammation and thickening  (Crapo
 et al., 1.984; Barry  and Crapo, 1985; Barry et al.,1985,  1988;
 Sherwin and Richters, 1985).  This occurs without increased lung
 collagen  (Filipowicz and McCauley, 1986; Wright et al., I988a)
unless exposure is  intermittent (Tyler et al.,  1988).  The
 increased lavagable  lipids found in lungs of rats (Wright et al.,

-------
  1988b)  after chronic exposures to 0.15 - 0.30 ppm 03 are
  consistent with cell population shifts and/or inflammation.
       Exposure of monkeys by Harkema et al.  (I987a,b) to 0.15 or
  0.30  ppm 03  for 6  or 90  days (8  hours/day)  resulted in
  "quantitative changes in nasal transitional and respiratory
  epithelium...,  ciliated  cell necrosis,  shortened cilia,  and
  secretory cell  hyperplasia.'•  The authors conclude that  ambient
  levels  of 03  can cause nasal  epithelial  lesions  which  may
  compromise defense mechanisms of  the upper  respiratory tract
  (Harkema  et  al., I987a,b).   This  did not occur  in  a controlled
  exposure  study  of  humans  exposed  to 0.4  ppm 03 for  4 hours
  (Carson et al., 1985) despite  increased  neutrophils (Graham et
  al., 1988).  The tracheal region  of monkeys shows  similar acute
  lesions as the nasal region but adaptation may occur (Hyde  and
 Plopper,  1988).
      Since CD closure, new information h'as become available which
 has improved the ability of EPA to extrapolate 03 toxicology data
 to human health effects levels.  An 03  dosimetry model  has been
 used to simulate local absorption of O3 in the lower respiratory
 tracts of guinea pigs and rats (Overton et  al.,  1987).   This
 model  along with information on respiratory  system uptake of 03
 in both  rats  and humans suggests that humans retain a greater
 fraction (97%)  of  inhaled 03  than  rodents (50%)  (Wiester  et  al
 1987,  1988; Gerrity and Wiester,  1987;  Gerrity,  1987; Hatch  and'
 Aissa, 1987)  and enhances EPA's ability to extrapolate  animal
 data to  human health  effects.   According to  Miller  et al.
 (1987a,b),  application  of an  03 dosimetry model to  obtain intra-
 and interspecies dose-response curves from collective assessments
 of toxicological data will improve the  overall risk assessment
 process, particularly in  its quantitative aspects,   while this
 new information  indicates great strides have been made  toward
understanding animal-to-human dosimetry relationships,  large
uncertainties remain due to species sensitivity differences
 (Hatch et al., 1986) as evidenced by significant variations  in
lung txssue concentrations of antioxidant enzymes among mammalian

-------
                             V3I-62
 species (Slade et al.,  1985;  Bryan and Jenkinson,  1987).   These
 differences will have to be dealt with in any animal to human
 extrapolation.   Furthermore,  extrapolation across  averaging times
 or  from high to low doses of  03  introduces  additional
 uncertainties into any extrapolation  of these effects.
     Until  the degree of uncertainty  associated  with using animal
 studies to  estimate human effects and effect  levels  is
 substantially reduced or methods are  devised  to  study humans
 during  long-term exposures to O3,  the  relationship between acute
 and chronic response to 03 will  remain  unclear.

     4.  Effects of Other Photochemical Oxidants
     It has been postulated that oxidants such as peroxy  acetyl
 nitrate (PAN)  and hydrogen peroxide (H202) may play a role  in
 producing health effects  associated with photochemical  oxidant
 exposures.   However,  relatively  few controlled human exposure or
 animal  toxicology studies  routinely have investigated these
 pollutants,  and  field and  epidemiology  studies evaluate mixtures
 of pollutants thus  making  it difficult  to judge which oxidant
 caused  effects.
     No  significant  effects have been reported in controlled
 exposures of  intermittently exercising  healthy young and middle-
 aged males  to PAN concentrations of 0.25 to 0.30 ppm (Drinkwater
 et al.,   1974; Raven  et al., 1974a,b, 1976; Gliner et al.,  1975).
 Only one study suggested a possible simultaneous effect of  PAN
 (.3 ppm) and O3  (.45 ppm)  (Drechsler-Parks et al.,  1984),  but as
with the other studies, this occurred at PAN concentrations much
higher than those reported for relatively high oxidant areas
 (0.047 ppm).  Except for an association of PAN with eye
 irritation  (Okawada et al., 1979; Altshuller,  1977; Javitz  et
al., 1983;  U.S. EPA, 1978; National Research Council, 1977) few
effects  of PAN have been reported. Field and epidemiological
studies  also have reported few relationships between health
effects  and PAN, while animal toxicology studies suggest that
only very high PAN concentrations produce effects in animals,

-------
                              VII-63
  such as significant alterations in host pulmonary defenses
  (Thomas et al., 1979, 1981) .   Review of ^ llterature on
  ofMI can be found in the CD (pp.  9-205 to 9-306 and 10-80 to

       Even fewer studies on the health effects of H20,  are
  available.   No significant effects  were observed in rats exposed
  for  7 days  to 0.5  ppm H202  in the presence of ammonium sulfate-
  it is generally assumed to not  penetrate into alveolar regions'
  possibly  due to the high solubility of  H202  (Last et al., 1982)
  Other studies are  of  little more than mechanistic value
       in conclusion, the  limited evidence available  suggests  that
  at levels found  in ambient  air, PAH and  H202 are not responsible
  for adverse respiratory  effects of  photochemical air pollution
 ozone „ considered to be chiefly responsible for the  adverse
 health effects of oxidants  largely due to the relative abundance
 compared to other oxidants  (CD,  pp.  „.„ to 12.6S) .      ndan°e

      5.  Interactions  with other Pollutants
      Although controlled human exposure  studies  have not
 demonstrated consistently any  enhancement of respiratory  effects
 for 03 when  combined with S02 , N02,   CO, and H2SO4 or  other
 particulate  aerosols,  animal toxicology  research suggests
 additive or  possibly synergistic effects  (CD,  p.  12-67)
 controlled exposure studies  of animals provide evidence ' that  03
 in combination with HO2 increases susceptibility to bacterial
 infection  (Ehrlich  et  al.,  1977, 1979; Ehrlich,  1980) and
morphological  lesions  (Freeman et al., 1974).  Exposure to O3 and
H2SO  has been reported to produce additive or even synergistic
effects in host defense mechanisms (Gardner et a!., 1977,  Last
       SS'
t
    .,
                M
                Mixtures of 03 and (NH4)2so4 have produced
                     " colaagen synthesis

-------
                             VII-64
 polluted air or O3  in purified air  (Avol et  al.,  1984).   Results
 of  the  study showed no  differences  thus suggesting that O3  is
 largely responsible for respiratory effects  in oxidant-polluted
 air.  However,  it should be  noted that combinations of oxidants
 with 03  may  contribute  to decreased  pulmonary  function and
 increased symptomatic effects in asthmatics  (Whittemore and Korn,
 1980; Linn et al.,  1980,  I983a; Lebowitz et  al.,-1982,  1983,
 1985; Lebowitz,  1984; Holguin et al., 1985)  and  in children'and
 adolescents  (Lippmann et al., 1983;  Lebowitz et  al.,  1982,  1983,
 1985; Bock et al.,  1985; Lioy et al., 1985).   Interactions
 between  03 and total suspended particulate matter were reported
 for decreased expiratory flow in children  (Lebowitz  et al.,  1982,
 1983, 1985;  Lebowitz, 1984) and in adults with airway  obstructive
 disease  (Lebowitz et al., 1982, 1983).  Thus it  appears
 reasonable to conclude  that 03 may cause most of the respiratory
 effects  in oxidant-polluted air, but effects may be  exacerbated
 by other pollutants.  Because efforts to examine pollutant
 interactions  have been  incomplete in clinical and epidemiological
 studies thus  far, the potential for pollutant interactions
warrants emphasis in the consideration of the margin of safety.

-------
                           VIII- 1
 VIII.   Staff Conclusions and Recommendations for Ozone Primary
        Standardise

     Drawing upon the evaluation of scientific information
 contained in the CD and the CDS,  which reviews new data made
 available since  CASAC closure on the CD,  this section provides
 preliminary staff conclusions and recommendations for
 consideration by the Administrator in selecting a pollutant
 indicator,  averaging time,  form,  and level  of the primary O3
 standard.   Staff conclusions and recommendations contained herein
 are based upon the  totality of scientific information and diverse
 effects reported in clinical,  epidemiology,  and toxicology
 research.

     A.   Pollutant  Indicator
     When the Environmental Protection Agency promulgated the
 NAAQS for photochemical  oxidants  in the Federal Register  (36 FR
 8186) on  April 30,  1971, the scientific data  base for health
 effects was  very limited.   The Schoettlin and Landau  (1961)
 study, which associated  increased  incidence of  asthma attacks
 with ambient photochemical  oxidants,  served an  important  role  in
 developing the basis  for the primary  photochemical oxidants
 standard.  Subsequent  health research,  however,  indicated  that  O3
 was the predominant oxidant of  concern for public health,  and  on
 February  8,  1979, the chemical  designation of the primary
 standard was  changed  from photochemical oxidants to O3  (44 FR
 8202).
     Since 1979,  a substantial human health effects research base
has established  03 as being  "...chiefly responsible for the
adverse effects  of photochemical air pollutants, largely  because
of its relative  abundance compared to  other photochemical
oxidants  (CD, p.  12-65)."  As discussed in section VII.B.3. of
the staff paper  and section  12.6 of the CD,  relatively few
controlled human studies have investigated the health
significance of peroxyacetyl nitrate  (PAN) or hydrogen peroxide

-------
                           VIII- 2
 (H202).   Of the controlled human exposure studies of PAN,  only
 one (Drechsler-Parks et al.f 1984) suggested a possible
 simultaneous effect of O3 and PAN.  Other controlled studies have
 reported no significant effects for PAN exposures of 0.25 to 0.30
 ppm,  much higher than PAN levels commonly reported in high
 oxidant areas (CD,  p. 12-65).  Because H2O2  is  highly soluble in
 aqueous media,  it is believed that H202 deposits  on  upper  airway
 surfaces rather than penetrating to the alveolar region (Last et
 al.,  1982).  However, investigations of H202 effects in the
 alveolar region have not yet been reported.
      Regarding  interactions with other pollutants,  the CD  (p. 12-
 67) has concluded that O3  alone  is considered responsible  for
 observed respiratory effects reported in  controlled  human
 exposures of 03  with SO2, N02, CO, and H2SO4  or  other particulate
 aerosols.   Animal toxicology studies,  however,  have  produced
 varied  results,  depending  on the pollutant combination evaluated
 and the  variable measured.   Additive and/or  possibly synergistic
 effects  have-been described from exposure to 03 and  NO2  (e.g.,
 increased susceptibility to bacterial  infection,  morphological
 lesions)  and from exposure  to O3 and H2SO4 (e.g.,  host defenses
 effects  and collagen synthesis).   Although a controlled human O3
 exposure vs.  ambient oxidant  exposure  comparison  study indicated
 that 03  was  the principle cause of respiratory effects,  several
 epidemiology studies suggest  that  combinations of oxidants may
 contribute  to such effects  as decreased lung function  and
 exacerbation of  symptoms in asthmatics and in children  and young
 adults  (CD,  p. 12-68).
     Despite the  apparent interactions of O3  with other
pollutants  as reported in toxicology and epidemiology studies,
controlled human  exposure and field studies have not consistently
demonstrated that respiratory effects observed in combined
exposure studies of 03 and other pollutants are  caused by any
pollutant(s) other than 03.  This divergence  in  the data base
does not reject the hypothesis that some portion of the human
respiratory effects associated with exposure  to photochemical

-------
                             VIII-  3
  oxidants may be attributed to pollutants  other  than o,.  However
  until further human research is  performed to support or  refute  '
  thls hypothesis, the staff  beUeve the case for oontrolling  03 as
  a surrogate for protecting  public health  fro* human  exposure to
  03 and other photochemical oxidants remains valid.
       The question of whether 03 can serve as an abatement
  surrogate for controlling other photochemical oxidants has been
  addressed in the CD (p. 12-15,  vith a quote from Altshuller
  (1983,  who concluded that "...   ambient air measurements indicate
  that 03  may serve directionally but cannot be expected to serve
  quantitatively  as a surrogate  for the other products."  This
  conclusion  appears  to  apply to the subset  of pnotochelnical
  products of concern -  03,  PAN, PPN, and ^ . iaentified .„
  CD,  even though Altshuller (1983)  examined the use of 0,  as an
  abatement surrogate for all photochemical  products.   Lack of  a
  quantitative, monotonic relationship  between 03  and other
  Photochemical oxidants  is  discussed in Chapter 5 of  the CD and
  demonstrated in Table 12-2 of the CD  in which average PAN/03
  ratios for different sites and years  vary  from 2 to 9   in
  addition, it is emphasized in the CD  (p. 12-17,   that  no single
 measurement -thodology can quantitatively and reliably measure
                                                 « « ambient air
      in spite of the above limitations,  it is generally
 recognised that control of ambient 03  levels  currently provides
 the best means of controlling photochemical oxidants of potential
 health  concern (o,,  PAN  PPN   ana  n n i   ™. •            potential
 „,„,       ,.       3      '    "'  and  H2°2>-  This recognition along
 with  a  controlled-exposure human health  data  base,  which
 implicates only 03 among  the photochemical  oxidants  at  levels
 reported  in ambient  air,  supports  the recommendation that O3 be
 retained  as the pollutant indicator for  controlling  ambient
 concentrations  of photochemical oxidants.   in addition,  using c,
 as a surrogate  for oxidant control  deemphasizes the  need for
monitoring of PAN, H202 and other  oxidants.   Unless significant
additional evidence which demonstrates human health effects from

-------
                           VIII- 4
 exposure to ambient levels of  non-03 oxidants becomes  available,
 it is  the staff's  recommendation that  03 remain as the surrogate
 for protection of  public  health from exposure to  all
 photochemical  oxidants.

 B.   Form of the Standard
     The current primary  03 NAAQS is expressed as an hourly
 average  which  is the  concentration not to be exceeded  on  more
 than 1 day per year on average.  During the  last standard  review
 the decision was made to  change from the deterministic to the
 statistical form of the standard.  The deterministic form, which
 permitted only a single hourly  exceedance of the  standard level
 in  any given year,  did not adequately  deal with the inherent
 variability in hourly 03 concentrations due to the stochastic
 nature of meteorological  factors affecting formation and
 dispersion of  03 in the atmosphere.   In addition,  EPA  further
 modified  the standard so  that one expected exceedance  would be
 given a daily  interpretation; that is,  a day with two  or  more
 hourly values  over  the standard  level  counts as one exceedance  of
 the  standard level  rather than  two.   These changes reduced the
 magnitude  and  increase stability of the "design value"  (or
 characteristic highest concentration;   see p. IV-2) used to
 evaluate  precontrol air quality  (Hayes et al.,  1984).
     If it  is desired to further increase stability of the design
 value indicator by allowing more exceedances,  the standard level
would have to be reduced to preserve equality of protection for a
multiple  exceedances  standard formulation.   For example, going to
 a 5 expected exceedances standard would require a reduction in
the standard level to the range of 0.09 to 0.11 ppm to maintain
the same  level of protection as the current 0.12 ppm 03 NAAQS.
 (A 5 exceedances standard has a design value 80% lower than a 1
exceedance standard, on average, but there is  variability in this
relationship among urban areas.)
     Since a multiple expected exceedances formulation for 0.12
ppm 03  does not provide  as much protection  against short-term

-------
                           VIII- 5
 peak concentrations as does the current O3 NAAQS standard,  it is
 recommended that the 1 expected exceedances form of the current
 O3 NAAQS  be retained for a short-term primary 03 standard.

 C.    Averaging Time(s)
      Exposure durations for studies reporting effects at or near
 ambient O3  levels  fall  into three general  categories—short-term
 (1 to 3 hours),  prolonged (6 to 8 hours),  and chronic (months to
 years).  Controlled chamber and field studies of acute pulmonary
 effects of  03  have reported statistically  significant impairment
 of group  mean lung function at  O3 concentrations <  0.20  ppm for
 exposure  durations of l to -2 hours (McDonnell et.  al., 1983,
 1985a,b,c;  Kulle et al.,  1985;  Folinsbee et.  al.,  1984;  Avol et
 al.,  1983,  1984,  I985a,b;  Linn  et al.,  1980,  1983a.,b; Adams and
 Schelegle,  1983;  Gong et al., 1986;  Linn et al., 1986; Schelegle
 and  Adams,  1986).   Prolonged and chronic exposure studies have
 reported  a  variety of pulmonary and  extrapulmonary  effects  for 03
 exposure  durations ranging from less than  a day  to  more  than a
 year,  but for  ethical reasons chronic studies have  been  performed
 using only  animal  subjects.
      As was the  case  during the 1978 review of the  ambient  O3
 standards,  acute effects  of O3  documented  in  controlled  human
 studies continue to provide the most quantitative and strongest
 support for a  short-term primary 03  standard  (Table VII-i,  staff
 paper section  VILA.).  Although 03  exposures  continue in these
 studies for 1  to 2 hours,  periods of intermittent exercise  during
 exposure  at 15 to  30  minute intervals have  caused effects to
 occur at  much  lower O3 levels than for resting subjects.  It has
 been  demonstrated  that maximum  impairment of  lung function  occurs
 during  or immediately after the  exercise period  (Folinsbee  et.
 al.,   1977a,b).  Although many individuals may  exercise and  work
 outdoors  for extended periods, most  do not  exercise heavily for
prolonged periods; therefore, the maximum impact of O3 for most
 individuals probably  occurs  after a  relatively short  time period.
While consideration may be  given  to  standards with averaging

-------
                           VIII- 6
 times of less than one hour, such a standard does not appear to
 be warranted.  As shown in Section V, all of the alternative 03
 standards investigated reduce the probability of experiencing a
 high acute O3 exposure.
      With regard to the need for a separate longer-term  (6-8 hr)
 primary O3 standard,  recent controlled human exposure,
 epidemiology and toxicology studies provide evidence of increased
 respiratory impairment caused by multi-hour 03  exposures.
 Preliminary results of controlled human exposure (6.6 hour)
 studies indicate that 03  levels  as low as 0.08  ppm produce
 statistically significant lung function decrements and
 respiratory symptoms  (Folinsbee  et al.,  1988; Horstman et al.,
 1988a,b).   In a  related  study suggestive evidence of inflammation
 was  reported in  humans exposed for 6.6 hours to 0.10 ppm (Koren
 et al.,  I988a,b,c).
      In the case of epidemiology studies interpretation  of
 exposure averaging times  associated with this effect remains
 uncertain.   Associations  have been reported  between  ambient
 photochemical  oxidant  levels  and both  asthma attack  rates
 (Whittemore  and  Korn,  1980; HoLguin et al.,  1985),  increased
 respiratory  hospital admissions  (Bates and Sizto,  1983,  1987),
 and  lung function decrements  (Lebowitz et al.,  1983;  Lebowitz^
 1984; Lippmann et al., 1983;  Bock  et al., 1985; Lioy  et  al.,
 1985).   in spite of the inherent exposure and environmental'
 uncertainties of these studies, they provide evidence of more
 serious  health effects associated with longer-term exposures than
 the apparently transitory effects reported for 1 to 2 hour
 exposures in most controlled exposure  studies.  For example, Lioy
 et al.  (1985) have reported that healthy, active children  (age 7-
 13) experience "a persistent decrement in function lasting for as
much as a week after the end of a smog period of about four
days."  Lioy and Dyba (1988) have proposed recently that the PEFR
decrements reported in Lioy et al. (1985) may result from total
03 dose rather than  persistence of effect from one day to the
next.  Finally, animal studies have provided collaborating

-------
                           VIII- 7
 evidence for time of exposure as an important factor in
 determining seriousness of effect.  Until further evidence is
 published which confirms exacerbation of FEV-L decrements from
 extended 03 exposure or more serious effects such as inflammation
 and persistence of lung function effects, the basis for a
 separate 6 to 8 hour O3 standard will remain inadequate.
      Much uncertainty continues to exist for extrapolation to
 human health effects of levels and averaging times associated
 with chronic exposure effects reported in animal toxicology
 studies.  Although the animal toxicology data base which
 documents most of the reported chronic effects provides extensive
 support for effects more serious than lung function decrements,
 uncertainties regarding dosiitfetry and species sensitivity must be
 addressed in any extrapolation to an effect level in humans.
 Animal studies investigating continuous and intermittent  exposure
 to 03  lasting  weeks  to  years  have  reported  changes  in  lung
 function (4 weeks to 1  year), morphology (i week to 18  months),
 biochemistry (l week to l  year),  host  defenses (1 week  to 3
 months)  and various  extrapulmonary effects  (up to 1 year).  with
 respect to  some of  these alterations (e.g.,  lung function,
 biochemistry)  the postexposure period  has been reported to be one
 of continued damage  rather than  complete recovery (CD,  p.  12-48).
 New research,  recent  developments  in animal  extrapolation, and a
 chronic effects risk  assessment  will provide a significantly
 improved basis  for addressing the  need  for  a separate chronic
 exposure standard in  the future.
     With respect to  protection  from longer-term and repeated
 peak exposures  afforded by a  1-hour  standard,  Appendix  A
 discusses relationships among air quality indicators in urban
 areas.   These relationships have been graphically presented in
 Appendix A by Figures A-2 through A-4.   Interpretation  of these
 figures  suggests that if the  current 0.12 ppm  1-hour daily
maximum 03 standard is met by all sites in the data set, then 10%
of all metropolitan statistical areas (MSAs) might have 17 or
more days with an 8-hour daily maximum > 0.08 ppm O3.  Although

-------
                           VIII- 8
 it is premature to suggest that 0.08 ppm O3  is  the level of
 concern for an 8-hour exposure,  this does suggest that the
 current standard does permit numerous potential 8-hour exposures
 to 0.08 ppm 03.   On  the  other hand,  if 0.10  ppm 03  is  determined
 to be the level of concern for  8-hour exposures,  the current O3
 standard provides reasonable protection from multiple 8-hour
 exposures in most areas.   The scientific basis  for judgment
 regarding the level  of concern  for health effects has not been
 determined for 8-hour or longer-term exposures.   Until adequate
 research has been completed and  accepted for incorporation into
 the CD,  determination of a level of  concern  for health effects
 arising from 8-hour  or longer-term exposures to O3  is  premature.
      In conclusion,  comments made by CASAC  (1986,  1987,  1988)
 support the  need  to  protect the  public from  health  effects caused
 by  acute exposures to O3.   The staff concurs and recommends  that
 a primary O3 standard with  a  l--hour averaging time be maintained
 to  provide protection from acute exposures.  The  CASAC (1986,
 1987,  1988)  also  discussed the need  and  basis for multi-hour O3
 standards; however,  no consensus was reached regarding  the need
 for alternative primary  standards.  New  clinical, epidemiology,
 and toxicology research has  provided additional support  for  the
 occurrence of more significant health  effects resulting  from 6  to
 8 hour,  multiple, and chronic exposures  to O3 than result  from
 single  l-hour O3 exposures.  A CDS of new published research
which addresses both acute  and longer-term O3 exposures has been
prepared  by the Environmental Criteria and Assessment Office
 (ECAO) and closure reached  on the document.  Staff recommends
that this  information be used by the Administrator in assessing
the need  for longer-term 03 primary standard(s)  while carefully
considering the conclusions reached by CASAC in the closure
letter  (McClellan, 1989).  Although the CASAC recognized the
emerging data base on acute effects resulting from multi-hour
exposures, it was the CASAC view that it would be some time
before enough such information would be published and included  in
a criteria document.  Thus, "CASAC concluded such information can

-------
                           VIII- 9
 better be considered in the next review of the ozone standards"
 (McClellan,  1989).   Also,  CASAC reached "closure on the staff
 position paper recommending a 1-hour standard ..." (CASAC,
 1988;  McClellan,  1989).  With this portion of the review complete
 and after considering CASAC's views on all issues,  the
 Administrator will  be in a position to make a regulatory decision
 on  how and when to  best act on the 1-hour standard.

 D.   Level of the  Primary Standard(s)
     In selecting the level of the primary NAAQS for O3,  the
 Administrator is  faced with consideration of a large and diverse
 health data  base.   Judgments regarding the scientific quality and
 strength of  that  data base already have been made in Chapters 9
 to  12  of the CD.  The preliminary assessment of  health effects
 attributed to 03 presented  in  Section VII  of  this staff  paper
 presents a variety  of health effects and discusses  the
 seriousness  of these  effects.   In addition to establishing a
 lowest observed effects  level  from these health  data,
 consideration must  be given to the uncertain evidence which bears
 on  a margin  of safety needed to protect  public health.   In
 addition,  the Administrator must  judge which  of  the  health
 effects  attributed  to O3 should be considered  adverse before  a
 final  judgment on the level  of  the O3 primary  standard can be
 made.  Although scientific  literature supports the conclusion
 that particular 03 concentrations and exposure patterns may pose
 risks  to human health, scientific  data can only  identify  the
 limits of  a  range within which  a  standard  should be  set.
 Specific numeric standard levels,  frequency of allowable
 exceedances,   and averaging times are largely a public policy
 judgment.
     This  section draws conclusions regarding  the scientific
 literature used by staff to  identify a lowest  observed effects
 level for  short-term health effects attributed to 03.  In
addition, the more uncertain or less quantifiable evidence,  which
forms the basis for judgements about which standard provides  an

-------
                           VIII- 10
 adequate a margin of safety,  is considered for use in
 recommending short-term O3  standard  options.   Although concern
 has been expressed for the more serious  effects reported in
 chronic animal exposure studies,  the major uncertainties
 associated with dosimetry and species sensitivity differences
 limit  direct quantitative extrapolation  of effects in animals to
 human  health effects at this  time.
     The strongest evidence of human health effects from acute
 exposure to 03  comes  from controlled  human  exposure  studies
 because 03  exposures  are  known  accurately,  other pollutants are
 not present,  temperature  and  humidity are  monitored,  and human
 subjects are used.   Field studies also provide strong evidence
 for similar reasons  with  the  exception that other  ambient
 pollutants  are  present  during exposure.  This  exception,  however,
 can provide support  for effects occurring under real  ambient
 conditions.   The  ethical  limitations  of controlled human exposure
 and field studies  (e.g. restriction to short-term exposures,  non-
 invasive techniques,  and  limited use  of infectious agents) have
 prevented investigation in humans of  the more  serious  chronic
 exposure effects which  have been reported in animal toxicology
 studies.  Field and epidemiology studies have  the advantage of
 providing associations  between  health effects  and "real world"
 ambient  exposures to  photochemical oxidant pollution;  however,
 available epidemiology  studies  provide less certain exposure-
 response evidence for O3 than the controlled-  exposure and field
 studies.  As a result, epidemiology data tend to be relied on
 less quantitatively in establishing a lowest observed effects
 level for 03.  Epidemiology data must be given consideration in
 developing the level of the NAAQS because they do provide
 evidence of effects experienced in more realistic environments
 than the artificial chamber exposure studies.
     Staff observations, conclusions, and recommendations
 regarding health effects of O3 are presented below.  They are
 based upon the scientific review in chapters 9 to 12 of the CD as
well as in the CDS and the preliminary assessment and analyses

-------
                            VIII- 11
  discussed in Section VII of this  staff paper.
  are as follows:
These conclusions
                                                         Lults to n.
  .   m .,        	    -- ~. .,. , .,—U.J.Q ppm.  AS summarized
  in Table VII-!,  statistically significant FEV, decrements have
  been reported in controlled exposure studies of intermittently
  heavily  exercising,  healthy children and adults exposed for two
  hours to 0.12 ppm O3 and of  continuously heavily exercising
  healthy  adults exposed  for  one hour to 0.16  (McDonnell et al
  1983; I985b;  Avol  et al., 1984,. Field studies have demonstrated
  that  statistically significant group mean FEV,  decrements  have
  been  induced  in continuously heavily exercising,  healthy
  adolescents at mean ambient 03 levels of  0.144  ppm and in
  continuously heavily exercising, healthy  adults at mean ambient
 03  levels of 0.153 and 0.165 ppm (Avol et al.,  I985a,b; Avol et
 al., 1984; Avol et al.,  1983; Linn et al., I983a)
 experiencing > i n s-
             	*-
 r.anqe from .1-0 i o Ifigur^ VTT-?).  These estimates are based
 upon  an  analysis  (Hayes et al.,  1987b)  of data obtained from
 individual subjects in three controlled exposure human studies
 (Avol  et al.,  1984;  Kulle  et al.,  1985;  McDonnell et al.,  1983)
 Because a  potentially  large group  of  the population  may
 experience a substantial reduction (>lo*,  of FEV, when  exposed to
 0-12 ppm 03, this 03 level should be considered as the lowest
observed effects level  (LOEL, but  not necessarily the lowest
observed adverse effects level (LOAEL) based on the  staffs
recommended definition of adverse effects in section VII c 1

-------
                           VIII- 12
      3•   Despite a close association reported (cpf p

 between  changes in group mean FE^ and occurred of CTroup
 respiratory symptoms — for acute o^ exposures of anmt-o i~
 chambers,  Haves et al.  ri987t^ report  only a
 correlation between FE^ changes and  symptoms  severity when +-h*
 analysis is conducted using individual data,   one reason that
 group mean analysis outcomes are different from results of
 individual data analysis may be that  symptoms reported are
 inherently more subjective than FE^  decrements measured;  also
 some  individuals may experience symptomatic effects without
 notable  changes in  FEVX  or vice versa due  to  differences in
 mechanisms causing  FEV^^  decrements versus  symptomatic  effects.

      4 '   Estimates  of the f-rac-hion of heavily exercising adul-hs
 experiencing — moderate or severe chest discnmfnr-t-  ranged from n
 to 15% and those experiencing  moderate or severe  nnucrh ranged
'from  0-13% when exposed  to 0.12 ppm P.,   (Figure  VU-3K  These
 estimates  are based upon an analysis  (Hayes et al.,  I987b  of data
 obtained  from individual subjects in  two controlled human  studies
 (Kulle et  al.,  1985;  McDonnell  et al.,  1983).

      5-  The pulmonary function  and symptom risk assessment for
 alternative o  hourly  NAAOS  rHayas ot al..  I987b?
1988) is one of the factors that should be considered  in
selecting an hourly primary NAaps that provides  an adequate
margin of safety.  From an air quality perspective without regard
to personal exposure patterns, the benchmark risk estimates show
probabilities from 0.5 to 1.0 that 03 concentrations will exceed
levels sufficient to cause > 10% decrements in FEVj^ or > mild
symptoms (cough and chest discomfort) in 1,5, or 10% of heavily
exercising individuals upon attainment of a 0.08, o.io, or 0.12
ppm O3 NAAQS.   Benchmark risk estimates are < 0.2 for two of the
three health data bases used when the endpoint is >20% FEV,
decrements or moderate/ severe respiratory symptoms.  When
personal exposure patterns are considered,  mean expected

-------
                           VIII- 13
 headcount estimates range from 0.5 to 18.8% of the heavily
 exercising population for > 10% FEVX  decrements or >'mild
 respiratory symptoms depending upon health data base used and
 urban  area examined upon attainment of a 0.12  ppm 03 NAAQS.   Mean
 expected  headcount estimates drop to  0.1 to 7.1% of the heavily
 exercising population for > 20% FEVj^  decrements or
 moderate/severe  respiratory symptoms.   The reader is referred to
 Table  VII-4 for  a  more detailed summary of expected headcount
 estimates and  to the discussion on pp.  VII-41-43 of caveats  and
 limitations that should be considered in interpreting the risk
 estimates.

     6.   Preliminary results indicate that prolonged (6 to 8-
 hour)  exposures  to  O^ mav cause  enhanced  lung  function
 decrements  and respiratory symptoms as  well as more persistent
 respiratory effects  than the luna function decrements produced
                                                          •
 bv single one- and two-hour exposures  to O3.  Preliminary
 evidence  of statistically significant  lung function decrements
 and respiratory  symptoms has been reported for  exercising healthy
 persons exposed  to O3  levels  of 0.08-0.12  ppm for  6.6 hours
 (Folinsbee  et  al.,  1988;  Horstmann et  al.,  1988a,b).   Similar
 exposure  protocols produced suggestive  evidence of inflammation
 at 0.10 ppm 03 (Koren  et  al., 1989a,b,c).  Persistence  of  lung
 function  response  in  children,  as  measured by peak expiratory
 flow rate (PEFR),  has  been reported to  last for as long as a  week
 following a  smog period of about  four days  duration in  which  the
 03 concentration peaks were  in the range of 0.120  to  0.185 ppm
 (Lioy et  al.,  1985).   While  the respiratory responses reported
 above appear to be primarily  caused by O3/ the authors suggest
 that other pollutants  such as acid  sulfates may  have  contributed
 to the persistent  lung  function effects.  Also,  Lioy  and  Dyba
 (1988)  have proposed recently that the PEFR decrements reported
 in Lioy et al.  (1985) resulted from total 03 dose rather than
persistence of the effect  from one day to the next.  Although  it
 appears to be premature to draw firm conclusions regarding

-------
                           VIII- 14
 persistent or enhanced respiratory effects associated with 6 to 8
 hour or longer term exposures to ambient 03," the staff concludes
 that there is reason for concern about enhanced respiratory
 function changes beyond the transient pulmonary function
 decrements thus far reported in most controlled human exposure
 and field studies.   The emerging data base of clinical and
 epidemiology research suggests that standards with longer-term
 averaging times may need to be considered as alternatives or
 additions to the current 1-hour averaging time primary standard
 after adeguate information is available to make an informed
 judgment regarding  need for such standards.
      — - Work Performance eiirrPirMy appears  tr» h* limited by
 exposure to O3; however,  too small  a data base ja available  -ho
 quantify the magnitude  of this  impairment
 this  effect  should  be  used  only  to develop  an  aH^i^te margin nf
 safetv-   Results  of exposure  to  03 during high exercise levels
 indicate  that discomfort may  be  an important factor  in limiting
 performance  (Adams  and Schelegle, 1983; Folinsbee et al.,  1984;
 Gong  et al., 1986;  Schelegle  and Adams, 1986).

          Subjective statements  by individuals engaged in  various
      sport activities  indicate that these individuals  may
      voluntarily  limit  strenuous exercise during high-oxidant
      concentrations.  However, increased ambient temperature  and
      relative humidity  are also  associated with episodes of hiqh-
      oxidant concentrations,  and these environmental conditions
      may also enhance subjective symptoms and physiological
      impairment during  O3 exposure (CD, p.  10-65.)

While it may be difficult to differentiate performance  effects
caused by 03 from those caused by other environmental conditions,
exacerbation of effects caused by 03,  beyond those caused by
other conditions,  may prevent or curtail normal activities and
should be viewed as adversely affecting individual performance;
however,  limitations of the data base suggest using this
information only in margin of safety considerations.

-------
                           VIII- 15
      8-   Based on the CDS review of new clinical evidence, both
 allergic and asthmatic subjects have a greater increase in airway
 resistance after 0^ exposure than do healthy subjects (CDS,  p.
 MLs	Epidemioloaical data provides additional qualitative
 evidence of exacerbation of asthma in adults at ambient Q3
 concentrations below those generally associated with symptoms or
 functional changes in most healthy adults.   Given the nature of
 these data,  they should be used in developing an adequate margin
 of safety.

      Whittemore and Korn (1980)  and Holguin et al.  (1985)  found
      small increases in the probability of  asthma attacks
      associated with previous attacks,  decreased temperature,  and
      incremental increases in oxidant and 03  concentrations,
      respectively.   Lebowitz et al.  (1982,  1983)  and Lebowitz
      (1984)  also showed effects in  asthmatics,  such  as  decreased
      peak  expiratory flow and increased respiratory  symptoms,
      that  were related to the interaction of  O,  and  temperature.
    .  (CD,  p.  12-55.)                           .

 Uncertainties  concerning individual  exposure  and confounding
 environmental  variables limit developing exposure-response
 relationships  with  these epidemiological data at this time and
 staff recommends  that  these  data  should be  used  to develop an
 adequate margin  of  safety.   However,  if it  is demonstrated that
 an  increased incidence of  asthma  attacks occurs  as a function of
 03 exposure, this would be an effect of  great medical
 significance and would require careful  consideration.

     9j	Luna  structure damage induced  bv long-term exposures to
QT_ has been demonstrated in several animal species.  These data
should be used in establishing an adequate margin of safety.
Significant morphological alterations in the  centriacinar  region
have been reported after long-term O3 exposures of rats (Barry et
al., 1983; Crapo et al., 1984), monkeys  (Eustis et al., 1981;
Fujinaka et al., 1985) and dogs (Freeman et al., 1973).  "There
is morphometric  (Fujinaka et al., 1985), morphologic (Freeman et

-------
al
a1'' 19?3),
and
               function -,
    *    SU
it is
   al
                                       s
                                         possible that

-------
                            VIII- 17
  humans exposed to 03 could experience decrements in host
  defenses; but at'the present time, the exact concentration at
  which effects might occur in man cannot be predicted, nor can the
  severity of the effect" (CD, p.  12-50).  The staff recommends
  that these data be used in selecting a standard which provides an
  adequate margin of safety.
       11.
                                 c, of o,^ reported  in
 addition to  linjtftd -vMenea  for-  ri«^iovaBeil1jtT.
 teratoloqical, mutaqpnin, endocrine  svstem.  *nr. Uv
 effects-  Until further analysis  better establishes the
 physiological significance of these  effects  on  human health   the
 staff recommends that, these effects  data be  used in selecting a
 standard which provides an adequate  margin of safety.
 susceptibilitY to o,. exposure are activity ieve]
 environmental  stress (..g.   humidit
                                                           Those
 factors which either have not been adequately tested or remain
 uncertain include age,  sex,  nutrition,  and smoking status.
 The Administrator should consider taking them into account in
 establishing an adequate margin of safety.
ion response
      13 '  Attenuatior> of  acute pulmo
by repeated daily  exposes  to o,  is a
phenomenon, which  should  hi  viaw^
increasing dosP of o^ to  the lungs.  Functional decrements are
greater on the second exposure day,  with smaller decrements  on
each successive day until the  fourth or fifth day when  small or
no decrements are  observed (Farrell  et al., 1979; Horvath et al
1981; Kulle et al., I982b; Linn et al., 1982).  Attenuation  of
functional response,  however, does not necessarily imply
attenuation of morphological or biochemical response to 0

-------
                           VIII- 18
      There is an interplay between tissue inflammation
      hyperresponsiveness, ensuing injury (damage), repair
      ?o??«SSr^ *?? Chan9es in response.  The initial response
      followed by its attenuation may be viewed either as
      sequential states in a continuing process of lung inlurv and
                                                               and
 Attenuation of pulmonary function or symptomatic: effects,
 therefore,  should be seen as a human response of concern 'because
 it may result in an increase in the total O3 dose reaching the
 lungs by permitting individuals to exercise more heavily outdoors
 when ambient O3  levels are high,  thus resulting in continued,  and
 potentially greater,  lung injury.  However,  the possibility
 exists that this response may indicate the body's ability to
 adapt to increased O3  levels.   Attenuation should,  therefore, .be
 considered  in developing an adequate margin of safety.
     MJ - Exacerbation  of  rpgpi T^.torv effects  by interaction of
other pollutants  with o, has not been  den.on.cH-r*^  in  controlled
human exposure  or field studio; however,  epidemiology studies
suggest that respiratory effects reported  near the  n   standard
may be caused by  0± in  combination with other  pn
Furthermore, animal studies -sugges* that combin
other pollutants may act additivelv or syn^rgi st Lcallv .  depend i
on the pollutants and endooints chosen for study.  While the  lack
of individual exposure analysis may limit development of
quantitative exposure-response relationships in epidemiology
studies, the large body of evidence cited in the CD  (p.  12-68)
supports the conclusion that ambient pollutants other than O3 may
interact with 03 to contribute to respiratory effects observed in
those studies and should be considered in developing an  adequate
margin of safety.
     •^ - Two groups have been iH^n^jfjed as being "potentially
at-risk" from exposure to O,;   m  that subgroup of the general
population characterized as having preexisting r-c.gpiratorv

-------
                           VIII- 19
 disease,  and (2)  those individuals whose activities outdoors
 result in increases in minute ventilation which includes
 responsive individuals who experience significantly greater
 decrements in luncr function from exposure to p., than the average
 response  of groups studied. rCD. pp.   12-88 to 12-89K
      In conclusion,  short-term controlled exposure studies
 (McDonnell et al.,  1983)  suggest that a group mean lowest
 observed  effects  level (LOEL)  for pulmonary function and symptoms
 for healthy,  exercising subjects is 0.12 ppm O3 for an averaging
 time of one to two hours.   Effects somewhat below 0.12 ppm 03
 have been reported in recent camp studies (Spektor et al.,
 1988a,b),  but it  has been  suggested that these effects may be  due
 in  part to interactions with other pollutants such as acid
 aerosols.   Analysis  of clinical  studies indicates that a small
 fraction  of the population may experience measurable lung
 function  and  symptomatic effects for  O3  exposures  even below 0.12
 ppm (Hayes et al.,  I987a,b,) while  some individuals show no
 effects at 0.12 ppm  (Linn  et al.,  1988).   Although significant
 lung function decrements and symptoms have been demonstrated in a
 single  study  (Horstman et  al., 1988,  1989)  to occur in healthy
 human subjects  exposed for 6.6 hours  to  03  levels  as  low  as 0.08
 ppm,  staff believes  a  larger data base is  required to establish a
 LOEL for prolonged exposures.  Uncertainty associated with
 assessment of epidemiology and animal data prevents identifying a
 lowest  observed effects level  for chronic  exposures at this time.
 Factors which staff  believe  should be considered  in developing a
 1-hour  standard which  provides an adequate  margin  of  safety are:
      (1) the  possibility that  individuals  with  pre-existing
respiratory disease may experience effects  at levels  below those
producing  effects in healthy subjects;
      (2) possible exacerbation of respiratory effects  by  other
pollutants in combination with O3 during ambient exposures;
      (3) the possibility that attenuation of pulmonary  function
decrements may  increase 03  dose inhaled and result in more
serious effects;

-------
                           VIII- 20
      (4)  possible limitation of work performance by exposure to
 °3;
      (5)  evidence of  exacerbation  of asthma  at  ambient O3
 concentrations;
      (6)  evidence of  bronchial  reactivity  in subjects  exposed to
 >  0.3 ppm 03;
      (7)  evidence indicating that  repeated acute and chronic
 exposures to O3 may cause  lung  structure damage;
      (8)  evidence of  increased  susceptibility to infection;
      (9)  extrapulmonary effects of O3'which  are  of uncertain
 biological importance;
      (10) potentially "at  risk" groups that  have not been
 adequately tested; and
      (11) factors  which may  affect susceptibility to O3.

 E.  Summary of Staff Recommendations
     Based upon information  and discussions  contained  in the CD
 the CDS,  and the  staff conclusions drawn in  Sections VIIIA-D,  the
 following staff recommendations; regarding the primary  O3 standard
 are as follows:
     1.    In consideration  of the large base  of health  information
 attributing effects to 03 exposure and the lack of evidence which
 demonstrates human health  effects  from exposure  to ambient levels
 of photochemical oxidants  other than 03,  staff recommends that O3
 remain as the surrogate for controlling ambient concentrations of
photochemical oxidants.

     2.   The current primary 03  NAAQS is  attained when the
expected number of days per calendar year with maximum hourly
average concentrations above the level of the standard is equal
to or less than one (44 FR 8202).   Staff  recommends that these
attributes of the O3 standard be retained.

     3.   Because research has demonstrated that respiratory
effects  of public health concern are associated with acute (1 and

-------
                           VIII- 21
 2  hour)  exposures it is recommended that the 1-hr averaging time
 be retained for the 03  primary NAAQS.   Although there are studies
 which have identified health effects associated with O3 exposures
 having longer than one-hour averaging times,  there remains great
 uncertainty about which exposure  characteristics (e.g., level,
 time,  repeated peaks)  are most important.   Recent clinical
 research provides evidence of enhanced respiratory effects of 03
 levels below the current NAAQS level when subjects are exposed
 for a 6.6 hour period (Folinsbee  et al.,  1988;  Horstman et al.,
 1988).   Epidemiology research suggests persistent changes in lung
 function,  aggravation of respiratory disease, and possibly
 increased hospital admissions,  all  of which are associated with
 03  levels  near  the current  NAAQS.   In  addition,  data  reported in
 animal toxicology studies provide evidence of lung structure
 damage,  increased susceptibility  to respiratory infection,  and a
 variety  of extrapulmonary effects following O3  exposures  lasting
 from-a few hours to more than a year.   The seriousness of these
 effects  indicates a need to protect public health from ambient 03
 exposures  which reasonably  can  be expected to induce  such effects
 in  humans.   This protection could be provided by setting  separate
 6 to 8-hour,  monthly, seasonal, or  annual  standards,  or by
 setting  a  1-hour standard which controls  longer-term  exposures of
 concern.
     Two types  of data  are  necessary for considering  the  possible
 need for a primary 03 standard with  an averaging time  of  greater
 than 1 hour:  (l)  health effects data and  (2) aerometric data,
 both empirical  and statistical.   Rombout et al.  (1986,  1989)
 published  air monitoring  data from  the Netherlands and New Jersey
 on relationships  between  the  1-hour  and 8-hour  averaging  times.
 The  results of  their particular data sets  indicate that 1-hour
maximum 03 levels do not always predict 8-hour average O3
 concentrations  above 0.10 ppm and, therefore, that the  daily
maximum hourly  03 standard currently in effect "... does not
adequately indicate the intensity of long duration and  high
concentration exposures to ozone" (Rombout et al., 1986).  They

-------
                           VIII- 22
 concluded from their data that a longer-term averaging time,
 between 7 and 10 hours, would be necessary to provide adequate
 protection of public health and, further, that a standard should
 be established for the highest daily 8-hour period based upon
 running 8-hour averages.  Consideration of the Rombout
 et al. (1986, 1989) analyses and conclusions must be tempered by
 recognition that a very limited data base was used and that
 extrapolating the New Jersey data to all areas of the U.S. is
 subject to large uncertainty.
      The New Jersey data used by Rombout et al.  (1986)  were for
 specific days,  at the sites chosen,  on which hourly O3
 concentrations  did not exceed 0.12  ppm,  the level of the current
 primary NAAQS.   It should be noted,  however,  that those New
 Jersey sites are  located in a  nonattainment area  and that diurnal
 03 curves at  those  sites,  therefore, are  not necessarily
 representative  of diurnal  curves  (i.e.,  the distribution  of
 hourly 03 concentrations over a day) observed in  areas  in which
 the standard  is actually attained.
      Thus, it seems prudent  to the staff  that the relationships
 among 1-hour  O3 concentrations and other potentially appropriate
 exposure statistics in  attainment areas or  under  attainment
 scenarios must be fully  explored before adopting  a  longer-term
 standard.  Whether  the current 1-hour primary standard can
 effectively serve as a surrogate for one or more longer-term
 averaging times depends upon concentrations of potential concern
 associated with longer-term exposures.  Those concentrations of
 potential concern must first be identified before the question of
 surrogacy can be resolved.
     OAQPS has analyzed whether it is possible to set the short-
 term standard at a  level which reduces the probability of
 experiencing an unacceptable long-term exposure.   As discussed in
Appendix A,  there is a marginally statistically significant
association between short- and longer-term O3 indices.   (For
example, if the current 0.12 Ppm l-hour daily maximum O3 standard

-------
                           VIII- 23
 is met by all sites in the data set, then 10% of all metropolitan
 statistical areas might have 17 or more days with an 8-hour
 daily maximum >0.08 ppm O3.)
      Although the CASAC recognized the emerging data base on
 acute effects resulting from multi-hour exposures, it was the
 CASAC view that it would be some time before enough such
 information would be published and included in a criteria
 document.   Thus,  "CASAC concluded such information can better be
 considered in the next review of the ozone standards" (McClellan,
 1989).   Also,  CASAC reached "closure on the staff position paper
 recommending a 1-hour standard ..." (CASAC,  1988;  McClellan,
 1989).   With this portion of  the review complete and after
 considering CASAC?s views on  all issues,  the Administrator will
 be in  a  position  to make a regulatory decision on how and when  to
 best act on the 1-hour standard.

     4.  Based on the CD the  CDS,  the staff assessment of the
 acute  (1-2  hour)  controlled exposure,  field,  epidemiology,  and
 animal toxicology data and recommendations  of  the Clean Air
 Scientific  Advisory Committee (CASAC,  1986,  1987,  1988),  staff
 recommends  that the range of  1-hour  03  levels  of  concern  for
 standard-setting  purposes should remain 0.08 ppm  to  0.12  ppm.
 The upper end  of  this  range (0.12 ppm)  represents lowest  03
 exposure concentrations  in controlled  exposure studies  in which
 healthy, heavily  exercising children,  adolescents, and  young
 adults have  experienced  statistically  significant, though small
 group mean decrements  in  lung function.  Analyses of  individual
 data from these studies  indicates that  larger  individual
 decrements have been observed in some  subjects with cumulative
 frequency distributions suggesting that some fraction of  subjects
 experience FEVj, decrements greater than 10% at 0.12 ppm O3.   Mild
 cough also has been reported for individuals in these studies at
the upper end of the range.  On the other hand, recently
published acute exposure research suggests that 0.12 ppm  and even
 0.14 ppm 03 may be a no effects level in some individuals for

-------
                          VIII- 24
     and symptoms.  As discussed in Section VII.c.l., staff has
recommended that mild individual responses  (i.e.., 5-10%  FEV,
decrements, 30 minute recovery, mild cough, no  limitation  of
activity) should not be considered adverse health effects,  if
this response is deemed to be not adverse or if 0.12 ppm is not
identified as a lowest observed effects level for O3, then it can
be argued that a standard level of 0.12 ppm does provide a margin
of safety.  Some field and epidemiology studies provide
suggestive evidence of acute lung function impairment at similar
or lower exposure levels, though these studies are limited by
uncertainties concerning individual exposure levels and  presence
of other pollutants contributing to or causing health effects
measured.  Other epidemiology studies have provided evidence of
aggravation of asthma and airway obstructive disease, again
limited by uncertainties about individual exposure and other
pollutants.  Given the difficulty of extrapolating these findings
to the U.S. population and the fact that the effects of  O3 on
individuals with preexisting respiratory disease have not yet
been fully characterized, staff believes that the upper  end of
the above range provides a relatively small if any margin of
safety.  These uncertainties and the nature of potential effects
are important considerations in developing an adequate margin of
safety.
     With regard to the lower end of the range,  neither the
evidence provided by animal toxicology studies (e.g.  increased
susceptibility to lung infection,  lung structure and biochemical
changes, and extrapulmonary effects)  nor other epidemiological
evidence suggesting respiratory effects in children and
susceptible groups provides scientific support for health effects
below 0.08 ppm for short-term (1-2 hour)  exposures.   These data
as well as estimates provided by exposure and risk analyses will
be considered in evaluating the. margin of safety provided by
alternative 1-hour standards in the range of 0.08 to 0.12 ppm.
     Health effects information discussed in Chapter VII of the
staff paper and the CDS provides reason for concern about multi-

-------
                           VIII- 25
 hour exposure (6-8 hours) of humans to O3.   Recently published
 clinical and epidemiology research suggests that extended
 exposure of exercising individuals to 0.08  to 0.12 ppm O3 may
 produce biologically important, as well as  statistically
 significant,  lung function decrements, respiratory symptoms and
 inflammation.   Epidemiology evidence further indicates a possible
 association of ambient 03 and other pollutants with increased
 respiratory hospital admissions over an extended summertime
 exposure.   Animal toxicology data show reversible epithelial
 injury in  the  lungs following sub-chronic exposure of animals to
 03  levels  as low  as 0.12  ppm;  structural  alterations in  the lung
 and irreversible  deposition of collagen in  the lungs previously
 have^been  reported in animals chronically exposed to much higher
 levels (0.61 ppm).
      Based on  the health  data base  discussed briefly above,  EPA
 has considered alternatives in addition to  the existing  1-hour
 primary O3  standard.   One obvious alternative  is  to provide
 public health  protection  from prolonged 03  exposures by  setting a
 separate longer-term primary st'andard  or  standards.   Great
 uncertainty exists  at this time, however, regarding either
 appropriate averaging times or levels  of  concern.   Another
 alternative under -consideration is  to  set a 1-hour  03 standard
 which  provides adequate protection  from longer-term exposures.
 The current 1-hour  primary standard of  0.12  ppm O3  is
 approximately  equivalent  to a  0.1 ppm,  8-hour  standard.   If  it
 should be determined  that protection of public health requires
 limiting 8-hour exposures to  lower  levels,  then it  would  probably
 be  necessary to tighten the 1-hour  standard, to be  useful  as  a
 surrogate for protection  from  longer exposures.  Again, however,
 this decision will  continue to  be premature  until new research
 has been fully evaluated  and formally incorporated during  the
 next review cycle in  a criteria document because such a change
would be based upon the need for protection  against prolonged
 exposure effects.

-------
                          VIII- 26
     Because there is a good health effects data base available
on 1-2 hour exposures, the staff concurs with- CASAC that review
of this scientific information, including the staff paper and its
assessment of this information be closed out.  With this portion
of the review complete, and after considering CASAC's views on
all issues, the Administrator will be in a position to make a
regulatory decision on how and when to best act on the 1-hour
standard.

-------
                               IX-  1
       This section reviews and assesses research on welfare
  effects attributed to 03 and other photochemical oxidants as
  summarized in Air Ouaii+y cri^-ri*  for Qg;nno ar^ ^.^
  ghotochemic*] Orld.n^ (CD) .  where possible, judgments are
  provided which identify adverse effects and levels of these
  effects for secondary-standard setting.
       Important categories with respect to welfare include
  agricultural vegetation effects (e.g.,  yield loss),  natural
  ecosystem effects,  personal  comfort and well-being effects  and
  materials damage.   Because the possible indirect  contributions of
  the photochemical oxidants to  visibility  degradation,  climatic
  changes,  and  acidic deposition cannot  at  present  be  quantified
  these atmospheric effects  and  phenomena are  not addressed  in  this
  document  (CD, p. i-1N  Tney have been  addressed/ haw      ^
  other recent  air quality documents  (U.S.  EPA, 1982a,b)   The
  approach taken in this staff paper  is: 1) to describe each type
 of 03  effect, 2)« to discuss what is known regarding exposure-
 response relationships of 03  exposure,  and 3) to evaluate the
 factors which should be considered in selecting the level'
 averaging time and form of the secondary NAAQS for O
 Discussions of the status of  the rural 03  exposure analysis,  the
 03  economic analyses,  and the welfare risk assessment are
 included.
A.
         Mechanisms  of  Action  for  Vegetation
                                                          as the
            of a sequence of physical, biochemical  and
Physiological events.  For ambient 03 to exert a phytotoxic
effect, lt must first diffuse into the plant,  such an effect
will occur only if . sufficient amount of 03 reaches the
,v.^.                    Space throu9h the stomata, which can
exert some control on O3 uptake to the active sites within the

-------
                             IX- 2
 leaf.  Once within the leaf, O3 quickly dissolves in the aqueous
 layer on the cells lining the air spaces.   Ozone or its
 decomposition products then diffuse through the cell wall and
 membrane into the cell,  where it may affect cellular or
 organellar processes.   Ozone injury will not occur if i) the rate
 of 03 uptake is  sufficiently small  so that  the  plant is able to
 detoxify O3  or its metabolites  or 2)  the plant  is able to repair
 or compensate for the  03  impacts  (Tingey and Taylor,  1982).   The
 uptake and movement of 03  to the  sensitive  cellular  sites are
 subject to various physiological  and biochemical controls (CD,  p.
 6-22).
      At any point along  this pathway,  03 or  its  decomposition
 products may react with  cellular  components.  Altered cell
 structure and function may result in changes in  membrane
 permeability,  carbon dioxide fixation,  and many  secondary
 metabolic processes (Tingey and Taylor,  1982).   The  magnitude of
 the 03-induced effects will  depend upon the physical  environment
 and the  chemical  environment of the  plant (including  other
 gaseous  air  pollutants and  a wide variety of  chemicals),  and
 biological factors  (including genetic potential,  developmental
 age of the plant  and interaction with plant pests).   Cellular
 injury manifests  itself in  a  number  of ways,  including  foliar
 injury,  premature senescence, reduced yield or growth or  both,
 reduced  plant vigor, and sometimes death.  Depending upon  the
 intended  use of a plant species (for food, forage, fiber,
 shelter,  or amenity), any of these effects discussed above could
 constitute a "welfare" effect, with potentially adverse effects
 on society.

     1.  Biochemical Response
     When 03 passes into the liquid  phase,  it undergoes
transformations that yield a variety of free radicals (e.g.,
 superoxide and hydroxyl radicals).  Whether these species result
 from decomposition of 03 or reactions between 03  and  biochemicals
 in the extracellular fluid has not been determined.  Ozone or its

-------
                             IX- 3
 decomposition products, or both, will then react with cellular
 components, resulting in structural and/or functional effects.
      The potential for 03,  directly or indirectly,  to oxidize
 biochemicals in vitro has been demonstrated.  Ozone can oxidize
 several classes of biochemicals including nucleotides, proteins,
 some amino acids and various lipids.  Data acquired from in vitro
 studies are best utilized to demonstrate that many cellular
 constituents are susceptible to oxidation by O3.   New approaches
 are needed to assess the full range of in vivo  biochemical
 changes caused by 03.   (See CD,  Section 6.3.1.3  for further
 details.)

      2.   Physiological Responses
      Physiological responses have been more useful  than
 biochemical changes in characterizing cell responses to oxidants
 (CD,  p.  6-26).   Many ,consider membranes to be the primary site of
 action  of  O3  (Heath, 1980;  Tingey  and  Taylor,  1982).   Whether  the
 plasma  membrane or some  organelle  membrane is the primary site of
 03 action  is open  to speculation  (Tingey  and  Taylor,  1982).  The-
 alteration in  plasma membrane function,  however,  is clearly an
 early step in  a series of 03-induced events that  eventually leads
 to  leaf  injury  and subsequent  yield  loss.
      The effect of 03  on key steps in  photosynthesis has been
 measured for several plant  species as  shown  in Table  IX-1  (CD,  6-
 28).  Reductions in photosynthesis may reflect the  direct
 impairment of chloroplast function or  reduced CO2 uptake
 resulting  from  O3-induced stomatal closure, or both. Regardless
 of the mechanism,  a sustained  reduction in photosynthesis will
 ultimately affect  growth, yield, and vigor of  plants.  Some
 examples of 03-induced reduction in apparent photosynthesis at
 concentrations  above and below 0.25 ppm are presented  in Table
 IX-1.  These data highlight the potential of O3 to reduce primary
productivity at various air quality levels, some of which are
close to ambient.

-------
                                         IX-4

                         TABLE IX-1.   EFFECT OF OZONE ON PHOTOSYNTHESIS
Species
Loblolly pine

Slash pine

Bean
Alfalfa

Ponderosa pine



Eastern white pine
Eastern white pine
Sensitive

Intermediate


Bean
Black oak
Sugar maple
White pine
Sensitive
Tolerant
Poplar hybrid
Ponderosa pine


03
concentration,
ppm
0.05

0.05

0.072
0.10
0.20
0.15

0.30

0.15
0.10
0.20
0.30
0.10
0.20
0.30
0.30
0.50
0.50

0.7 or 0.9
0.70 to 0.95
0.90
450, 700
800 ppm-hr

Exposure duration
18 wk
continuously
18 wk
continuously
4 hr/day for 18 days
1 hr
1 hr
9 hr daily/
60 days
9 hr daily/
30 days
19 days
4 hr/day for 50 days
4 hr/day for 50 days
4 hr/day for 50 days
4 hr/day for 50 days
4 hr daily/50 days
4 hr daily/50 days
3 hr
4 hr daily/2 days
4 hr daily/2 days

3.0 or 10
10/30 days
1.5 hr
Cumulative
dose over
1,2,3 yr
X
inhibition
15b

9b

18b
4b
10b
•25C

67C

10C
24j
42?
sr
Not sig.
different
14b
20b
22C
30 1 10d
21 ± 10d
h
100b
ob
60e
90b


Reference
Barnes (1972a)

Barnes (1972a)

Coyne and Bingham (1978)
Bennett and Hilt (1974)

Miller et al. (1969)



Barnes (1972a)
Yang et al. (1983)





Pell and Brennan (1973)
Carlson (1979)
Carlson (1979)

Botkin et al. (1972)

Furukawa and Kadota (1C!/
Coyne and Bingham (1981)


 1 ppra = 1960 ug/m.
bP < 0.05.
CP < 0.01.
 Standard deviation.
eNo statistical information.
Source:   03  Criteria Document,  U.S.  EPA, 1986

-------
                             IX- 5
      For example, Miller et al. (1969) (Table IX-1) found that 3-
 year-old ponderosa pine seedlings sustained a 25-percent
 reduction in apparent photosynthesis after a 60-day exposure to
 an O3  concentration of 0.15 ppm for 9 hours a day.   Yang et al.
 (1983) exposed three clones of white pine to 03 concentrations of
 0.10,  0.20 or 0.30 ppm for 4 hours per day for 50 days in CSTR
 chambers.   Net photosynthesis was reduced in the foliage of
 sensitive and intermediate clones by 14 to 51 percent in direct
 relation to O3 dose and clonal sensitivity.   Coyne  and Bingham
 (1978) exposed field-grown snap beans to an 03  concentration of
 0.072  ppm for 4 hours a day for 18 days.   Apparent  photosynthesis
 was reduced 18 percent in plants treated with O3.
     In addition to depressing photosynthesis in the foliage of
 many plant species,  03  inhibits  the  allocation  and  translocation
 of photosynthate (e.g.,  sucrose)  from the shoots to other organs
 (Tingey,  1974; Jacobson,  1982),  often referred  to as
 partitioning.   Tingey et al.  (1971)  found that,  when radish
 plants were exposed to  03  (0.05  ppm  for 8  hrs a  day,  5  days  per
 week for  5 weeks),  hypocotyl  (stem)  growth was  inhibited 50
 percent while  foliage growth  was inhibited only 10  percent.
 Ponderosa  pine exposed  to  0.10 ppm O3  for  6 hours per day  for 20
 weeks  stored significantly  less  sugar and  starch in their  roots
 than did control  plants  (Tingey,  1976). Several  other studies
 have measured  this  partitioning  effect of  O3 on  photosynthate in
 carrot, parsley,  sweet corn,  cotton  and pepper  (Oshima,  1973;
 Bennett and Oshima,  1976; Oshima et  al. 1978; Oshima et al.,
 1979;  Bennett  et  al., 1979).   Plants  were  exposed to 03
 concentrations of 0.12 to 0.25 ppm for 3 to 6 hours  for 0.2
percent to 7 percent  of the total growth period  of  all  the
plants.  In all species but pepper, root dry weight was depressed
much more than leaf dry weight.  The  reduction in photosynthate
translocation to roots and the resulting decrease in root size
indicate that the plant had fewer stored reserves, possibly
rendering it more sensitive to injury  from cold, heat,  or water
stress.  When less carbohydrate is present in roots, less energy

-------
                             IX- 6
 will- be available for root-related functions.  An 03-induced
 suppression of nitrogen fixation by root nodules could affect
 total biomass and agricultural yield, especially in areas where
 soil nitrogen is low.
      Reproductive capacity (flowering and seed set)  is reduced by
 03 in ornamental plants,  soybean,  corn,  wheat and some other
 plants (Adedipe et al.,  1972;  Feder and Campbell, 1968; Heagle et
 al.,  1972,  1974; Shannon and Mulchi,  1974).   These data suggest
 that 03  impairs the fertilization  process  in  plants.   This
 suggestion  has been confirmed  in tobacco and  corn studies using
 low concentrations (0.05  to 0.06 ppm)  of O3 (Feder,  1968,
 Mumford,  1972).   In addition to these physiological  effects,
 which are directly related to  productivity, there are  many
 secondary metabolic responses  in a plant exposed to  03,  such  as
 increase  in stress ethylene, which may contribute to the
 manifestation  of foliar  injury (CD, p.  6-31).

      B.   Factors Affecting Plant Response
      The  magnitude of plant response  to  O3 exposure can  be
 affected  by biological, physical,  and  chemical variables.   Each
 plant  exposed  to a given concentration of O3 will respond to  a
 different extent,  depending on  such biological factors as
 genetics, stage  of development,  and presence of  pests or disease,
 as well as  such  physical factors as temperature,  humidity,  light'
 intensity,  and soil moisture.   Chemical  factors  in the
 environment of plants can  also  modify plant response by either
magnifying  effects, as with multiple pollutants,  or by reducing
effects, as with antioxidant sprays.  All of these determinants
of response can overlap,  thus potentially creating a complex
myriad of causes for the effects observed.

     1.  Biological Factors
     a.  Plant Genetics
    'Differential susceptibility of individual plants to 03
exposure is determined by genetic composition.  Genetic variance

-------
                              IX- 7
  in response to O3  is  seen  among species, cultivars, and
  individuals within a  population.  Although foliar injury has been
  used as the most common measure of comparative 03 sensitivity
  other measures,  such  as yield and physiological effects
  substantiate differential  results based on genetic differences
  In fact,  the relative O3 sensitivity of cultivars within a
  species can vary with dose and  nature of response measured
  (Tingey et  al.,  1972; Heagle, 1979b).  Differences in relative
  sensitivity of cultivars have been reported between controlled 0
  exposures and ambient air field studies (Taylor,  1974;  Meiners *
  and Heggested, 1979; Hucl and Beyersdorf,  1982;  DeVos et al
  1983).  The  nature of inherited O3 sensitivity may help  to  "
  explain this disparity,   changes in gene expression  during plant
  development  and due to environmental  variations may  explain in
  part the potential  variability  in plant  response.
      b.  Developmental Factors
      The stage of plant  development plays a role in determining
 sensitivity to o3.   Just prior to, or at maximum, leaf expansion,
 Plant foliage appears  to be most sensitive because stomata are
 functional,  intercellular spaces are expanded, and barriers  to
 gas exchange are  minimal (CD, p.  6-34).  stages of plant
 development  also  can be  affected by C, exposure.   Premature  aging
 and leaf drop have  been  demonstrated in numerous  field and
 controlled studies  (Menser and Street, 1962; Heggestad,  1973;
 Hofstra et al., 1978;  Pell et al., 1980;  Reich, 1983;  Mooi
 1980).   The  premature  leaf drop and senescence decrease  the
 amount  of photosynthate that a leaf can contribute  to  plant
 !enetc*  " ^ ** C°nClUded that the  effects  of  °3 on the
 senescence process,  whenever initiated, may  be responsible for
 many of the documented reductions in yield  (CD, p.  6-34).
     c.  Pathogen and Pest Interactions with Ozone
     Disease has been defined  as  the result of a complex
 interaction between  host  plant, environment, and pathogen;

~s(:r:d r3rh rhogens and insects  have — *«-
        (CD,  p.  6-35).  ozone  can affect the development of

-------
                             IX- 8
 disease in plant populations.   Laboratory evidence suggests that
 03  (at  ambient concentrations  or  greater  for. 4  or more hours)
 inhibits infection by pathogens and subsequent  disease
 development (Laurence,  1981; Heagle and Letchworth,  1982).
 Increases,  however,  in diseases from "stress pathogens" have been
 noted.   For example,  plants exposed to  03 were  more  readily
 injured by Botrytis  than plants not exposed to  O3 (Manning  et
 al.,  1970a,b;  Wukasch and Hofstra,  1977a,b;  Bisessar,  1982).
 Both  field and laboratory studies have  confirmed  that  the roots
 and cut stumps of  03-injured ponderosa  and Jeffrey pines  are more
 readily colonized  by  a  root rot (Heterobasidion annosus).   The
 degree  of  infection was correlated  with the  foliar injury (James
 et al.,  1980;  Miller  et al., 1982).   Studies  in the  San
 Bernardino  National Forest showed that  03-injured trees were
 predisposed to attack by bark  beetles and that  fewer bark beetles
 were  required  to kill an 03-injured tree  (Miller  et  al.,  1982).
 At any  stage of the disease cycle 03 may alter the success  of  an
 invading organism  by  direct effects on  them or  by modifying the
 ability of  the host-plants to  defend against attack.

      2.  Physical  Factors
      Environmental conditions  before, and during,  plant exposure
 are influential in determining plant.response.  The  influence  of
 environmental  factors has been studied primarily  under controlled
 conditions, but field observations have substantiated the
 results.  Factors which  can potentially influence response  of
plants to 03 include light intensity, temperature, relative
humidity, soil moisture, and soil fertility.  Variations in  one
or more of these parameters can cause plants to be more or  less
sensitive to O3 exposures.  The CD has identified  the following
generalizations concerning the influence of these factors on
plant response (CD, p. 6-244) :
     a.    Light conditions conducive to stomatal opening appear
          to enhance 03 injury  due to increased  O3 absorption.

-------
                              IX- 9
       b.   No consistent relationship between temperature  and
            response to 03 has been reported; however, plants do
            not appear to be as sensitive at extremely high or low
            temperatures .
       c.   Plant injury tends to increase with increasing relative
            humidity as a result of the effect of humidity on
            stomatal opening.   McLaughlin and Taylor (1980)
            demonstrated that  plants absorb significantly more O3
            at high humidity than at low humidity.
       d.    Decreasing  soil  moisture increases  plant water stress
            causing a reduction in plant  sensitivity to  O3.  The
            reduced 03 sensitivity is apparently related  to
            stomatal closure, which reduces  03 uptake  (U.S.  EPA
            19.78; Olszyk and Tibbitts, 1981; Tingey  et al . ,  I9B'2}
           Water stress, however,  does not provide  a permanent
           tolerance to 03 (Tingey et al., 1982) .

      3.   Chemical Factors
      The chemical environment of plants can include air
 pollutants, herbicides, fungicides, insecticides,  nematocides
 InflT11^' T ^^  ^^^   Th- "ctor..  which may
 influence plant response to 03,  can be  grouped into two areas-
 multiple pollutants and chemical sprays.
      a.   Multiple  Pollutants
      Studies  indicate  that  the joint action of 03 and sulfur
      b     *                                          et  al.,
     ,b, Beckerson and Hofstra, 1979; olszyk and Tibbitts   1981,
     synergisn (injury enhancement, is most common at lot'
concentrations of each gas and when foliar injury induced by each
gas, md^duany, is small.  At higher concentrations or when
                                     the influence of S02 on
      response to 03  at ambient and higher concentrations for

-------
                             IX- 10
 several plant species - soybean (Heagle, et al., i983c; Reich and
 Amundson, 1984), beans (Oshima, 1978; Heggestad and Bennett,
 1981),  and potatoes (Foster et al., 1983).  Ozone altered plant
 yield,  but S02 had no significant  effect and did not interact
 with 03 to reduce plant yield unless the S02 exposures were much
 greater than typically found in the ambient air in the U.S.
      The applicability of the yield results from pollutant
 combination studies to ambient conditions is not known.  An
 analysis of ambient air monitoring data (SAROAD, 1981; EPRI-SURE)
 indicate that most sites wheres the two pollutants were co-
 monitored had ten or fewer periods of co-occurrence occurred
 during  the growing season (Lefohn  and Tingey,  1984).   Co-
 occurrence was defined as the simultaneous occurrence of hourly
 averaged concentrations of 0.05 ppm or greater for both
 pollutants.   At this time,  it appears that most of the studies of
 the  effects on pollutant  combinations (03  and  SO2) on  plant  yield
 have used a longer exposure duration and a higher frequency of
 pollutant co-occurrence than occur  in the  ambient air.
 Preliminary studies  using three pollutant  mixtures  (03,  S02, N02)
 have shown that  the  additions  of SO2  and NO2 (at low
 concentrations)  caused  a  greater growth  reduction  than O3 alone
 (CD,  p.  1-79).
      Although  Lefohn and  Tingey (1984) did  not  observe
 significant  co-occurrence  of 03 and N02 or SO2 in their analysis
 of ambient  air monitoring  data,  there  is some evidence to suggest
 that  03 climatology in natural ecosystems may be correlated with
 that  of other anthropogenic pollutants  (Taylor  and Norby, 1985) .
 Consequently, many forested regions of North America,  where
monitors are scarce, may experience elevated levels of  03 in
combination with other gases  (e.g., nitrogen oxides, sulfur
dioxide, nitric acid vapor, and organics) as well as various
pollutants deposited in rain or cloud water  (e.g., trace
elements, hydrogen ion, nitrate and sulfate) (Taylor and Norby,
1985).  Given the potential for these interactive effects to
occur (McLaughlin, 1985), a number  of new studies examining the

-------
                             IX- 11
  effects of 03 and other acidifying substances on forest trees are

  H^L9 rrte? a-s part °f the Forsst Rera ^- »*«
  NAPAP (National Acid Precipitation Assessment Program, .  New data
  will be available on this issue over the next two to five years
       b.  Chemical Sprays
       Chemical sprays have long been used to protect agricultural
  crops from pests and diseases.  Fungicides, herbicides,
  insecticides, and nematocides control damage caused by fungus
  weeds, insects,  and  ne.atodes, respectively.  They  a!so have been

  on :fait:r rsitivity °f piants *» •* >°"— • ^^
  of the effects of pesticides  on 03 sensitivity have shown
  dUfenng results, with some  chemicals (e.g., nematoxides,
  Phenamlphos)  increasing sensitivity and others (e.g., benomyl
  =arboX1n) reducing sensitivity of plants to 03  Miner et aT
  1976; sung and Moore, 1979).                               '
      Antioxidants, which are commonly used  to reduce rubber
 cracKang and  food sppiiage, have been reported to reduce
 vegetation injury caused by 03 (Kendrick et al.,  „„,.- Tne
 addition  of antioxidants to insecticides, herbicides and
        t                                      vegetation
       et al.,  1980; Koiwai et al.,  1977; Gilbert
 Ethylenediurea  ,„,. . widely                ~
 visible in3ury in bean plants exposed for „„ minutes to 0.8
                                    can apparentiy protect

pract ca                              ^ tO be ~«ici«*
practical to be used solely for this purpose (CD,  pp.  6-68 to
     c.  Heavy Metals

-------
                             IX- 12
 fertilized with sludge,  may have  important effects on vegetation
 and  ecosystems.   Heavy metals  may penetrate the cuticle and cause
 direct  toxic  effects  on  plants or penetrate the soil  and affect
 plant roots,  which  in turn  may contaminate the  food chain.   The
 effects of 03 in  combination with heavy metals  have been studied
 in several plant  species..   Zinc (Zn) and cadmium (Cd)  reacted
 synergistically with  O3  (0.30  ppm for 6 hours)  in  producing
 visible injury and  loss  of  chlorophyll in  garden cress and
 lettuce (Czuba and  Ormrod,  1974).  Exposure to  the combination of
 Cd and  O3  induced earlier development of necrosis  and  chlorosis
 and the injury was  observed at lower O3 plus Cd  levels  than for
 the individual treatments (Czuba  and Ormrod, 1981).
     Low concentrations  of  Cd  and nickel (Ni) have been shown to
 enhance O3 phytotoxicity on the growth peas  (Ormrod, 1977).   The
 interaction of Cd and  03 was influenced by both concentration and
 environmental conditions.   Tomato plants grown  at-  0.25  and  0.75
mg Cd/ml developed  only  slight  foliar injury when  exposed to 03
 (0.20 ppm  for 3 hours) under cloudy skies;  whereas  the  Cd
treatment  alone had no significant effect  (Harkov  et al., 1979).
 In full  sun there was  extensive 03 injury and the joint response
was synergistic.  Quaking aspen treated with 10 jug  Cd/ml for 30
days displayed significantly more  foliar injury when exposed to
ambient  air in New Jersey or exposed to 0.20 ppm O3 for 2.5  hours
 (Clark and Brennan,  1980).   When plants were exposed to  0.03  ppm
03,  the Cd enhancement of injury was not  apparent.   Although  it
is not possible at the present time to assess the risk  from  the
joint action of gaseous and heavy metal pollutants to vegetation,
the limited data available indicates that heavy metals  can
increase the phytotoxic reactions of 03  (CD, p.  6-67).

-------
                             X-  1
X.   Assessment of Welfare Effects and Related Welfare  Issues
     Considered in Selecting Secondary Standardrs^  for  Ozone

     Of the phytotoxic compounds commonly found  in  the  ambient
air, 03 is the most prevalent, impairing crop production and
injuring native vegetation and ecosystems more than any other air
pollutant  (Heck et al., 1980).  Some of the effects of  03
reported in the literature occur at 03 levels at or below natural
background concentrations in many areas of the country  (see
Section IV. for further discussion of background values).  Ozone
has also been shown to damage elastomers, textile fibers and dyes
and certain types of paints.  Other photochemical oxidants of
importance to effects on vegetation, ecosystems and materials are
nitrogen dioxide  (NO2) and peroxyacetyl nitrates.   Air Quality
Criteria for Oxides of Nitrogen  (U.S. EPA, 1982) and Review  of
the NAAOS for NO^;  Assessment of Scientific and  Technical
Information (U.S. EPA, 1984) previously assessed the phytoxicity
of N02, and thus NO2  will  not be discussed in this staff paper.
In addition, while at a given dose the peroxyacetyl nitrates are
more phytotoxic than O3 (p.  X-22),  they generally occur at
significantly lower ambient concentrations.  Because phytotoxic
concentrations of peroxyacetyl nitrates are less widely
distributed than those of 03 (CD, p.  6-1),  the focus of this
staff paper will be on the effects of O3.
     The objective of this section of the staff paper is to
assess the current basis for the 03 secondary NAAQS as contained
in Chapters 6, 7 and 8 of the CD.  In addition, the section will
summarize new analyses that address key issues of concern for the
secondary standard:  relationships of various air quality
indicators, crop loss estimates, averaging times and forest
response to 03.   Key  new studies that relate to the issue of
averaging time(s)  will also be discussed to determine whether new
effects information suggests any change in existing secondary
NAAQS for 03.

-------
                              X- 2
      A.   Vegetation Effects
           1.   Types of  Exposure Effects
      Plant response to  O3 exposure is quite varied and may  be
 expressed as  biochemical, physiological,  visible  injury,  growth,
 yield, reproductive and ecosystem effects.  When  reviewing  the
 current data  base  for vegetation effects,  it  is apparent  that  the
 bulk  of the scientific  research on crops  that has been completed
 since the last standard review  focuses on reduction  in growth  and
 yield from various  long-term  (days, months) exposures.  In
 contrast,  the majority  of short-term  (few hours)  exposure studies
 focus on  foliar injury  or physiological changes as the response
 measure.   While biochemical and physiological alterations are  the
 basis of  all  subsequent effects as described  in section IX  A.l.
 of this paper, visible  foliar injury provides one of the  earliest
 manifestations of short-term effects of O3.   However, these
 effects are not always  well correlated with reduction in  growth
 and yield  (CD, p. 6-141).
      This  section will  attempt  to assess  the  data with regard  to
 foliar injury effects and reductions in growth and yield.   For
 purposes of this staff  paper, greater emphasis will be placed  on
 damage or  yield loss than on injury.   Injury  inc:ludes all plant
 reactions, such as  reversible changes in plant metabolism (e.g.,
 altered photosynthesis), leaf necrosis,  altered plant quality, or
 reduced growth, which do not impair yield or  intended use of a
plant (Guderian,  1977).   Damage or yield  loss, on the other hand,
 involves any effect which reduces the quantity,  use,  or value
 (e.g., aesthetic)  of a plant or any impairment in the intended
use of a plant.  Although foliar injury may have some limitations
 in evaluating plant response, its presence is usually an
indicator of elevated O3 concentrations.   Growth and  yield losses
provide an important measure of the effects of 03  because  such
 losses impair the intended use of the plant and generally
constitute damage,  whereas foliar injury may or may not be
considered damage.   These growth and  yield loss effects have
become the focus  of most of the exposure response models and

-------
                               X- 3
  assessments to be discussed later,  thus providing the strongest
  and  best  documented evidence of vegetation effects of o,
  exposure.
                 a.   Visible  Foliar Injury Effects
      The  first documented observation  of O3  injury to vegetation
  in the field was  by Richards  et al.  (1958),  who described O3
  stipple on  grape  vine leaves,   other studies soon  after confirmed
  that foliar injury  was being  caused by  O3 from nearby  cities
  (Heggestad  and Middleton, 1959;  Daines  et al., i960).  Numerous
  subsequent  studies  have reported vegetation  injury at  rural sites
 caused by O3 transported long distances from urban centers (Heck
 et al.,  1969; Kelleher and Feder, 1978; Skelly, 1980;  Edinger et
 al.,  1972).
      Of  the many approaches taken to estimate the 03 levels and
 exposure times required to induce these foliar injury effects
 most  have involved short-term exposures (less than one day) and
 have  measured visible injury as 'the  response variable.  Various
 Plant species were exposed to a range of 03  levels  and exposure
 durations  by Heck  and Tingey (197!),  who evaluated  the resulting
 data  by regression analysis.   Data for  several species are
 summarized in  Table  X-l  (CD,  p.  6-224)  to illustrate the range  of
 03 levels  and exposure times required to induce 5 and  20%  foliar
 injury on  sensitive,  intermediate, and  less  sensitive  species
      Limiting-value  analysis  is  an alternative approach to
 estimating 03 levels and exposure durations which induce foliar
 inDury.  such an analysis, which  was performed on more  than 100
 studies of agricultural crops and 18 studies  of tree species
yielded the  following range of concentration  and exposure   '
durations that were  likely to induce visible  injury  (Jacobson,
-i» y * i j •
     1.    Agricultural crops:
          0.20 to 0.41 ppm for 0.5 hr
          0.10 to 0.25 ppm for l.o hr
          0.04  to 0.09 ppm for 4.0 hr

-------
                                   X- 4
 Table X-l.  Ozone Concentrations for Short-Term Exposure that  Produce 5 or
     20 Percent Injury to Vegetation Growth Under Sensitive  Conditions*
               Ozone Concentrations  that  mav Produce 5% (20%)  In-Juryi
Exposure                                                    Less
time, hr  Sensitive plants     Intermediate  plants      sensitive plants
8.0       0.02 to 0.04        0.07 to 0.12
°-5       0.35 to 0.50         0.55 to  0.70              > 0.70 (0.85)
          (0.45 to 0.60)       (0.65  to 0.85)

1-°       0.15 to 0.25         0.25 to  0.40              > 0.40 (0.55)
          (0.20 to 0.35)       (0.35  to 0.55)

2-°       0-09 to 0.15         0.15 to'0.25              > 0.30 (0.40)
          (0.12 to 0.25)       (0.25  to 0.35)

4.0       0.04 to 0.09       ^  0.10 to  0.15              > 0.25 (0.35)
          (0.10 to 0.15)       (0.15  to 0.30)
> 0.20 (0.30)
aThe concentrations in parenthesis are for the 20% injury level.  Table  is
from U.S. Environmental Protection Agency  (1986, p. 6-224).
Source:  O3 Criteria Document, U.S. EPA, 1986

-------
                              X- 5
 2.    Trees and shrubs:
      0.20 to 0.51 ppm for 1.0 hr
      0.10 to 0.25 ppm for 2.0 hr
      0.06 to 0.17 ppm for 4.0 hr

      Foliar injury in sensitive plants provides a basis for a
 good  understanding of the occurrence of elevated 03
 concentrations in a given area.   Among the.plants that have been
 identified as indicators  of pollutants are milkweed  (Duchelle and
 Skelly,  1981),  eastern  white pine (Benoit  et al.,  1982),
 ponderosa and Jeffrey pine (Miller,  1973)  Bel W-3  tobacco
 (Posthumus,  1976)  and lichens (Sigal and Nash,  1983).   Although
 the presence of visible foliar symptoms on vegetation  cannot be
 directly related to effects on growth or yield,  they do indicate
 that  elevated levels of O3  have  occurred.  The detection  of
 visible  symptoms is an  indication that additional  studies should
 be undertaken to determine if effects on growth  and yield are
 occurring (CD,  pp.  6-84).
      There are  a few studies in  which short-term exposures have
 resulted in growth  and  yield reduction:  exposure of cherry belle
 radish to 0.25  ppm  O3 for three  hours  resulted in  a 38  percent
 reduction in  root dry weight (Adedipe and  Ormrod,  1974) ;  exposure
 to 0.10  ppm O3  for  two hours  resulted  in a 9 percent reduction  in
 the average of  three growth responses of capri petunia  (Adedipe,
 1972) ; a 16 percent reduction in leaflet area growth rate was
 reported in pinto bean  exposed to  0.05  ppm O3 for  12 hours
 (Evans,  1974).   In  addition,  some  cultivars  of crops such as
 spinach  and tobacco  experience yield  losses  due  to extensive
 foliar injury at 0.10 ppm for 2-hours  (Menser and Hodges,  1972).
Thus, although  a few studies  relate short-term exposures  and
yield loss, there is very little yield  loss  data based  on
exposures that  are  easy to  relate to  the l-hr average of  the
current secondary standard.
     Despite the importance of visible symptoms, it must  be
recognized that  long-term exposure to low pollutant

-------
                             X-  6
concentrations may  adversely affect plant health without
producing visible symptoms.  Chronic  injury  from this  type of
exposure may be represented by reductions in growth, and/or
yield, or premature senescence resulting from changes  in
photosynthesis, respiration, chlorophyll content or other
processes (Dochinger et al., 1970; Feder, .1978; Heck,  1966;
Posthumus, 1976) .
     Because plant  growth and production depend on
photosynthetically  functional leaves, various studies  have been
conducted to assess the relationship  between foliar injury and
yield for species in which the foliage is not part of  the  yield.
Some research has demonstrated significant yield loss  with little
or no foliar injury (Tingey et al., 1971; Tingey and Reinert,
1975; Kress and Skelly, I985a,b,c; Adedipe et al., 1972),  while
others have reported significant foliar injury without yield  loss
(Heagle et al., 1974; Oshima et al.,  1975).  Relative
sensitivities of two potato, cultivars were reversed when judged
by foliar injury versus yield reductions (Pell et al., 1980).
Foliar injury in field corn was reported at  lower O3
concentrations than yield loss.,  but"  at higher 03 levels, yield
loss was increased to a greater extent than  foliar injury;
similarly, it was reported that foliar injury was not  a good
predictor of yield loss for wheat (Heagle et al., 1979a,b).
Therefore, while the presence of foliar injury is significant in
and of itself,  no precise relationship exists between  foliar
injury and yield loss for species of plants  for which  foliage is
not part of the yield (CD, p.  6-141).

          b.   Growth and Yield Effects
     As stated previously, the bulk of the evidence since the
last standard review has focused on reduction in growth and yield
from various long-term exposures.  The data base has been
summarized previously (Chapter 6, CD)  and consists of open top
chamber studies,  greenhouse and other controlled experiments, and
various ambient air exposures.   This section will assess the

-------
                               X- 7
  strengths and weaknesses of these various approaches.  Most  of
  these studies have characterized the exposure-response
  relationship in terms of the seasonal daily daylight mean 0,
  concentration, although various averaging times have been used
       For purposes of assessing exposure-response relationships
  derived from the existing studies,  yield loss is defined as an
  impairment of, or decrease in,  the value of the intended use of
  the plant.   This concept includes reduction in aesthetic values
  changes in crop quality,  and occurrence of foliar injury when  '
  foliage is the marketable part  of the  plant.   The actual amount
  of  yield loss  due to  decreased  aesthetic value or appearance  may
  be  more difficult to  quantify than yield loss  in weight  or  bulk
  but is  extremely  important  for  crops such as tobacco,  spinach
  and ornamentals.   Such effects  occur at  concentrations as low'as
  0.041 ppm for  several weeks or  o.io ppm  for 2  hours, and can
  constitute a yield loss when marketability of  the plant is
  decreased  (Menser and Hodges, 1972).
      Most of the recent studies of 03-induced yield loss have
 measured effects on the weight gf the marketable plant organ.
 These effects will be the primary focus of this section.   studies
 conducted to estimate the impact of 03  on the yield Of  various
 crop species have been grouped into two types,  depending on the
 experimental design and statistical methods used to analyze the
 data:  (i)  studies that developed predictive equations  relating
 03 exposure  to  plant response and  (2) studies comparing discrete
 treatment levels to a  control using analyses  of variance.  The
 advantage of  the regression  approach  is  that  it permits the
 estimation  of the  O3 impact on plant yield over the range of
 concentrations,  not just at the  treatment means as  is the case
 with the  analysis  of variance methods.
           (1)  Open-Top Chamber  Studies
      Data fro,, a series of studies conducted by  the National Crop
Loss Assessment Network (NCLAN)  have been  analyzed to develop
predictive equations relating 7-hour seasonal nean 03 exposures
to crop y.eld loss. Exampies of the relationship between 03

-------
                              X- 8
 concentration and plant  yield are  shown  in  Figure X-l (CD,  p.  6-
 229)  and  X-2  (CD,  p.  6-230).   These  cultivars/species were
 selected  because  they illustrate the kinds  of exposure response
 that  occur  and the type  of  year-to-year  variation in  plant
 response  to 03  that may  occur.   The  derived regression equations
 can be used to determine the  concentrations that  would be
 predicted to  cause a  specific yield  loss or to  estimate the
 predicted yield loss  that would result from a specific O3
 concentration.  Both  of  the approaches cited above have been used
 to summarize  the  data on crop response to 03 by use of  the
 Weibull (Rawlings  and Cure, 1985)  function.
      As an  example of response,  the  03 concentrations  that  would
 be predicted  to cause a  10- to  30-percent yield loss  have been
 estimated (Table  X-2  (CD, p.  6-232).  These cutpoints  were
 selected  to  illustrate  an effect  and do not imply an  effects
 threshold.  Other  cutpoints could  have been selected.   A brief
 review of these data  in  the table  suggests that:   (1)  a  X0% mean
 yield loss  is predicted  for several  species when  the  7-hour
 seasonal  mean concentration of  03 exceeds 0.04-0.05 ppm; (2)
 grain crops were generally less  sensitive to 03  than other  crops;
 (3) sensitivity differences within a species may  be as  large as
 differences between species.  In addition to differences in
 sensitivity among  species and cultivars,  the data in Figure X-l
 and X-2 illustrate year to year  variations in plant response to

     Although linear regression  equations have been used to
 estimate yield  loss, there appear to be systematic deviations
 from the data for some species and cultivars even though the
equations had moderate to high coefficients of determination
 (R2).   The use of  plateau-linear, polynomial equations and the
recently developed Weibull model (Heck et al.,  1983)  appeared to
fit the data better.  On the basis of available  data,  it is
recommended that curvilinear exposure response functions be used
to describe and analyze plant response to 03 (CD,  p.  7-106).

-------
                                                    X-9
     6000
     5000
  i  4000
 Jt
 a
 ut

 2  3000
 (0
    2000
    1000
            (A)
                      SOYBEAN (OAVIS)
                      RALEIGH. 1981 AND 1982
                       1982 Uk£
                       Y * «831-
-------
S
    6000

    5600

    SOOO

    4500

    4000
M   3500
Q
Z
    3000
    2500

    2000

    1500
                     COTTON (SJ-2)
                     SHAFTER. CA. 1981 AND 1982
                            1982(A)
                             .6872-KV0.088)2-1
         0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
              O3 CONCENTRATION, ppm
                                                    a
34
33
32
31
30
29
                                                    M
                                                    S!   27
                                                    u.
26
25
24
23
,(8)
TOMATO (MURIETTA)
TRACY. CA. 1981 AND 1982
           1982(A|
           V02.3-«V°-082)3-06
                   ,\
                                                           0  O.O2  0.04  0.06 0.08  0.1  0.12  0.14 0.16
                                                                   O3 CONCENTRATION, ppm
                         _
                          a.
                          a>
                         X
                         (0
                         O
                         O
                         ac
                              16
                              15
                              14
                              13
                              12
                              11
                              10
                               9
                               8
                               7
                               6
                               5
                               4
                               3
                               2
                                                 TURNIP (TOKYO CROSS)
                                                 RALEIGH. 1979 AND 1980
                                 0  0.02  0.04 0.06  0.08  0.1  0.12  0.14 0.16
                                       O, CONCENTRATION, ppm
      FIGURE  X-2.   Examples of the effects of ozone on the yield of cotton, tomato, and
            turnip. The 03 concentrations are expressed as 7-hr seasonal mean concentrations.
            The species were selected as examples of O3 effects and of year-to-year variations in
            plant response to O3.
            Source: Cotton and tomato data from Heck et al. (19846); turnip data from Heagle et
            al. (1985).

-------
      ,«  nm^       X~2' SUMMARY  OF  OZONE  CONCENTRATIONS  PREDICTED TO CAUSE
      10  PERCENT AND  30 PERCENT YIELD  LOSSES AND SUMMARY  OF YIELD LOSSES PREDICT
      TO  OCCUR AT  7-hr  SEASONABLE MEAN OZONE CONCENTRATIONS OF 0 04 and 0 06
Species
Legume crops
Soybean, Corsoy
Soybean, Davis (81)
Soybean, Davis (CA-82)
Soybean, Davis (PA-82)
Soybean, Essex
Soybean, Forrest
Soybean, Williams
Soybean, Hodgson
Bean, Kidney
Peanut, NC-6
Grain crops
Wheat, Abe
Wheat, Arthur 71
Wheat, Roland
Wheat, Vona
Wheat, Bluefaoy II
Wheat, Coker 47-27
Wheat, Holly
Wheat, Oasis
Corn, PAG 397
Corn, Pioneer 3780
Corn, Coker 16
Sorghum, DeKalb-28
Barley, Poco
Fiber crops
Cotton, Acala SJ-2 (81)
Cotton, Acala SJ-2 (82)
Cotton, Stoneville
Horticultural crops
Tomato, Murrieta (81)
Tomato, Murrieta (82)
Lettuce, Empire
Spinach, America
Spinach, Hybrid
Spinach, Viroflay
Spinach, Winter Bloom
Turnip, Just Right
Turnip, Pur Top W. G.
Turnip, Shogoin
Turnip, Tokyo Cross
==========—=—-__
03 concentrations, ppm,
predicted to cause
yield losses of:
lux

0.048
0.038
0.048
0.059
0.048
0.076
0.039 .
0.032
0.033
0.046

0.059
0.056
0.039
0.028
0.088
0.064
0.099
0.093
0.095
0.075
0.133
0.108
0.121

0.044
0.032
. 0.047

0.079
0.040
0.053
0.046
0.043
0.048
0.049
0.043
' 0.040
0.036
0.053
30%

0.082
0.071
0.081
0.081
0.099
0.118
0.093
0.066
0.063
0.073

0.095
0.094
0.067
0.041
0.127
0.107
0.127
0.135
0.126
0.111
0.175
0.186
0.161

0.096
0.055
0.075

0.108
0.059
0.075
0.082
0.082
0.080
0.080
0.064
0.064
0.060
0.072
Percent yield losses predicted
to occur at 7-hr seasonal
mean 03 concentration of:
0.04 ppm

6.4
11.5
6.4
2.0
7.2
1 7
10^4
15.4
14.9
6.4

3.3
4.1
10.3
28.8
n R
\j . j
2.2
0.0
0.4
0.3
1.4
0.0
0.0 '
0.0

8.3
16.1
4.6

0.8
10.3
0.0
6.8
2.6
6.0
5.8
7.7
10.1
13.0
3.3
0.06 ppm

16.6
24.1
16.5
10.4
14.3
5«*
. 3
18.1
18.4
28
19.4

10.4
11.7
24.5
51.2
2O
.8
8.4
0.9
2.4
1.5
5.1
0.3
2.7
0.5

16.2
35.1
16.2

3.7
31.2
16.8
17.2
9.2
16.7
16.5
24.9
26.5
29.7
15.6
                       ... — from Weibull
        in charcoal-filtered air.

Source:   Derived from Heck  et al.  (1984b).

-------
                             X- 12
     Although NCLAN provides valuable dose response  information
on a variety of crops, the program has limitations that must  also
be considered.  The adequacy of the 7-hr seasonal mean has been
questioned on several grounds.  Recent evidence has  suggested
that the diurnal cycle of O3 in natural ecosystems may differ
markedly from that of urban airsheds where the highest O3
concentrations are in the mid-afternoon.  In remote  locations it
is likely that maximum O3 concentrations shift into the late
afternoon and evening hours (Taylor and Norby, 1985; McCurdy,
1987).  In addition, the use of the seasonal mean as well as
other mean statistics in characterizing exposure implies that all
exposures over the course of the daylight period are equally
effective in eliciting a plant response and minimizes the
contribution of peak concentrations.  Thus, while the 7-hr
seasonal mean may contain all hourly concentrations for the 7-hr
period, it treats all concentrations the same (CD, p. 6-10).   An
infinite number of hourly distributions can be used to generate
the same 7-hr seasonal mean; some containing many peaks and
others containing none.  In fact, Larsen and Heck (1984)  have
emphasized that it is possible for two air sampling sites with
the same daytime arithmetic mean 03  concentration to have
different estimated crop reductions.
     Several authors have reported the greater importance of
concentration compared to exposure duration in causing injury
(e.g.,  Heck et al., 1966; Bennett, 1979,  Heck and Tingey, 1971,
Reinert and Nelson, 1979, Larsen and Heck,  1984).  More recent
findings indicate that constant concentrations have less effect
on plant growth responses than variable or episodic exposures at
equivalent cumulative doses (Musselman et al., 1983; Hogsett et
al.,  1985).   Hogsett found that over the period of 3 cuttings
(133  days),  alfalfa growth was reduced more when exposed to an
episodic 03  profile than to a  regime of daily 03  peaks, both  with
equivalent long-term means.  Generally,  assessment of human
physiological responses as well as vegetation responses to 03
have stressed the importance of episodic exposures (high

-------
                               X- 13
  concentrations over short periods of tine) in eliciting a
  biological response (Rogers, 1985).
       There are several additional concerns regarding the NCLAN
  data.  First,  there is the lack of validation of the model.  That
  is,  however,  a common deficiency among all models.  A second
  concern that  has been identified is that the use of the Weibull
  model might bias the estimated yield losses in the low dose range
  of the curve,   finally,  the most serious limitation of the NCLAN
  data stems from the inadequate sampling of environmental
  conditions and in inadequate number of  test sites.   The problem
  of inadequate  sampling is  an easy criticism to level at almost
  any  agricultural  study that  has  the  objective  of making an
  inference  over time  and  space.   The  limited time and resources
  meant that  subjective  judgments  had  to  be  made regarding which
  species were to be studied.  The  primary criteria  for these
  decisions were the relative  economic importance of the  species
  and their sensitivity to 03.  While the limitations mentioned
 above create uncertainties that should be considered in any
 application of trie results, the staff concludes that with
 appropriate caveats, the NCLAN data, as discussed,  provide useful
 information on crop loss due to O3 exposure.  A more detailed
 analysis of the NCLAN crop loss data which addresses some of the
 key uncertainties, particularly regarding exposure  dynamics,  is
 discussed on p. x-46.
           (2)   Greenhouse and Controlled Environment Studies
      The effects of O3  on plant yield may be affected by a  host
 of genetic  and  environmental  factors,  m addition  to the use of
 regression  approaches in  the  studies  previously discussed
 various  other approaches  have been used  to-  investigate the
 effects  of  03 on crop yield under more controlled (to various
 degrees) conditions,  as shown in  Tables  6-22 and 6-23 of the  CD
 (CD  P.  6-129 to -6-137,.  These studies were designed to test
whether specific O3 treatments were different from the control
using analysis of variance.  To summarize the data,  the  lowest'O
concentration that significantly reduced yield was taken fLT  '

-------
                              X- 14
 each study (Table X-3; CD,  p.  6-235); this concentration was
 frequently the lowest concentration used in the study.  Although
 it is difficult to estimate a  "no-effect" exposure concentration,
 the data generally seem to  indicate that 03 concentrations of
 0.10 ppm (frequently the lowest concentration used in the study)
 for a few hours a day for several days to several weeks induced
 yield losses  of 10-55 percent.
      One weakness of studies conducted under more controlled
 conditions (greenhouse,  growth  chamber)  is that it is difficult
 to extrapolate data from the chamber to field conditions.   The
 more controlled chamber  data, however,  do serve to strengthen the
 demonstration of 03  effects  in  the  field.   Concentrations  of 0.10
 ppm and  above appear to  cause yield reductions consistently,
 although exceptions  can  be  found.   In studies which used
 concentrations below 0.10 ppm,  the  response varied among species
 (Table 6-22;  CD,  p.  6-130).  Concentrations of 0..05 ppm in
 extended or repeated exposures  have been -shown to  cause yield
 reductions in some  species or cultivars,  no effects in others,
 and  increased yield  in others (Table  6-23;  CD,  p.,  6-134).
 Although these studies seem  to  suggest that a  higher  03
 concentration was required to cause an effect  than  was  estimated
 from the regression  studies, it should be noted that  the
 concentrations  derived from  the regression  studies  were  based on
 a  10% yield loss, while the  studies using analysis  of  variance
 indicate that 0.10 ppm frequently induced substantial  losses (10
 to 55 percent).
           (3)   Ambient Air Exposure Studies
     Ambient  air exposure studies conducted to date demonstrate
that 03 in many areas of the country can reduce plant yield.
Although the most severe effects appear to occur in areas with
the highest 03 concentrations such as the South Coast Air Basin
and the  San Bernadino Mountains in California, other agricultural
areas in the country can be  impacted as well.  Recently, open-top
chamber studies evaluating the yield of plants grown in the

-------
 Plant  species
                        TABLE  X-3. OZONE CONCENTRATIONS AT WHICH SIGNIFICANT YIELD  LOSSES HAVE BEEN NOTED FOR
                               A  VARIETY OF PLANT SPECIES EXPOSED UNDER VARIOUS EXPERIMENTAL CONDITIONS
Exposure duration
Yield reduction,
  % of control
03 concentration,
      ppm
                                                                                                          Reference
Alfalfa
Alfalfa
Pasture grass
Lad i no clover
Soybean
Sweet corn
Sweet corn
Wheat
Radish
Beet
Potato
Pepper
Cotton
Carnation
Coleus
Begonia
Ponderosa pine
Western white
pine
Loblolly pine
Pitch pine
Poplar
Hybrid poplar
Hybrid poplar
Red maple
American
sycamore
Sweetgum
White ash
Green ash
Wi How oak
Sugar maple
7 hr/day, 70 days
2 hr/day, 21 day
4 hr/day, 5 days/wk, 5 wk
6 hr/day, 5 days
6 hr/day, 133 days
6 hr/day, 64 days
3 hr/day, 3 days/wk, 8 wk
4 hr/day, 7 day
3 hr
2 hr/day, 38 days
3 hr/day, every 2 wk,
120 days
3 hr/day, 3 days/wk, 11 wk
6 hr/day, 2 days/wk, 13 wk
24 hr/day, 12 days
2 hr
4 hr/day, once every 6 days
for a total of 4 times
6 hr/day, 126 days
6 hr/days, 126 days
6 hr/day, 28 days
6 hr/day, 28 days
12 hr/day, 5 mo
12 hr/day, 102 days
8 hr/day, 5 day/wk, 6 wk
8 hr/day, 6 wk
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
51, top dry wt
16, top dry wt
20, top dry wt
20, shoot dry wt
55, seed wt/plant
45, seed wt/plant
13, ear fresh wt
30, seed yield
33, root dry wt
40, storage root dry wt
25, tuber wt
19, fruit dry wt
62, fiber dry wt
74, no. of flower buds
21), flower no.
55, flower wt
21, stem dry wt
9, stem dry wt
• 18, height growth
13, height growth
+1333, leaf abscission
58, height growth
50, shoot dry wt
37, height growth
9, height growth
29, height growth
17, total dry wt
24, height growth
19, height growth
12, heigjit growth
	 —
0.10
0.10
0.09
0.10
0.10
0.10
0.20
0.20
0.25
0.20
0.20
0.12
0.25
0.05-0.09
0.20
0.25
0.10
0.10
0.05
0.10
0.041
0.15
0.15
0.25
0.05
0.10
0.15
0.10
0.15
0.15
Neely et al. (1977)
Hoffman et al. (1975)
Horsman et al. (1980)
Blum et al. (1982)
Heagle et al. (1974)
Heagle et al. (1972)
Oshima (1973)
Shannon and Mulchi (1974)
Adedipe and Ormrod (1974)
Ogata and Haas (1973)
Pell et al. (1980)
Bennett et al. (1979)
Oshima et al. (1979)
Feder and Campbell (1968)
Adedipe et al. (1972)
Reinert and Nelson (1979)
Wilhour and Neely (1977)
Wilhour and Neely (1977)
•Wilhour and Neely (1977)
Wilhour and Neely (1977)
Wilhour and Neely (1977)
Patton (1981)
Patton (1981)
Dochinger and Townsend (1979)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
                                                                                                                                     X
                                                                                                                                      I
Source:   0   Criteria Document, U.S.  EPA,  1986

-------
                              X- 16
 presence of 03  (ambient air)  versus charcoal-filtered air have
 demonstrated losses in tomato (33  percent),  bean (26 percent),
 soybean (20 percent),  snapbean (0  to 22  percent),  sweet corn (9
 percent),  several tree species (12 to 67 percent),  and forbes,
 grasses and sedges (9  to 33  percent)  (Table  X-4; CD,  pp.  6-145
 and 6-146).   Field studies  in the  San Bernadino Forest during the
 last 30 years indicate that  ambient 03 reduced  the  height growth
 of  ponderosa pine by 25 percent, annual  radial  growth by  37
 percent and the  total  volume of wood produced by 84 percent
 (Miller et  al.,  1982).   Oxidant-induced  changes in  forest
 ecosystems  will  be discussed in more detail  in  Section B.
      Antioxidant chemical protectants appear to provide another
 objective method of estimating the impact of ambient  O3 on crop
 production  (Toivonen et al.,  1982).   Some limitations,  however,
 must  be kept in  mind.   The chemical  itself may  alter  plant growth
 and therefore results  must be  interpreted carefully.   Also,  since
 the chemical may not be effective  against all concentrations of
 all pollutants,  an  underestimation of yield  loss may  result
 (Manning et  al.,  1974).
      Ethylenediurea  (EDU), developed  as a chemical  protectant to
prevent oxidative effects of 03/  has been used extensively  to
reduce visible 03 injury in greenhouse and field studies and  to
estimate O3-induced yield loss,.  Several studies have reported
estimates of the  impact of 03 on yield by comparing yield data
from plots with  and without EDU treatment (Table X-5; CD,   p.  6-
237).  For a seven-week study  in which ambient 03 exceeded 0.15
and 0.08 ppm on  five separate days at each concentration,   EDU
treatment reduced foliar injury of onions and increased yield by
37.8% (Wakasch and Hofstra,  1977).   During a June-to-August  study
in which ambient 03 exceeded 0.08  ppm on 15  days with a maximum
of 0.14 ppm, EDU treatment increased yield of the Tiny Tim tomato
by 30% but had no effect on the New Yorker cultivar (Legassicke
and Ormrod,  1981).  Other studies  with beans (Toivonen et al.,
1982), tobacco (Bissessar and Palmer, 1984),  and potatoes
 (Bissessar,  1982) provided evidence of EDU increasing yield by

-------
TABLE X-4
	 	 	 ' — •""•" «•« «.n«notK». OH CiKtENHOUSES ON GROWTH AND VlFin (IF 0.05



0.052
0.051
0.035
>0.05

>0.05
>0.05

— ...
Exposure duration
	 __ 	
99 day average (0600-2100)

43 day average (0600-2100)
3 no average (0900-2000)

31X of hr (8:00 a.m. to
10:00 p.m.) from late
June to mid-September
over three summers; SX
of the time the concen-
tration was above 0.08 ppm
1979, 8 hr/day average
1000-1800), April-
September
1980. 8 hr/day average
(1000-1800), April -
September
1981, 8 hr/day average
(1000-1800). April-
Sepbember
Average 170 hr over 60
days exposure (1972-1974)
(6 crops)
Average 170 hr over 60
days exposure (1972-1974)
(6 crops)
Average 170 hr over 60
days exposure (1972-1974)
(6 crops)
— ~ " 	 • 	 	 	
Percent
reduction
fron control
	 j — - — • 	 —
33 , fruit fresh
wt

26d, pod fresh wt;
24 , number of pods
1, pod wt

20d, seed wt; 10d.
wt/100 +2. X pro-.
tein content, 4X
oil content

,
32. total above
ground biomass
20, total above
ground biomass
21, total above
ground biomass
+5, pod fresh
wt

14d, pod fresh
wt
3, pod fresh
wt

	 „ „___. „. Mfekk** i bu vnwrj
Location
of study
"
New York


Maryland

Maryland



Virginia
Virginia

Maryland

Maryland
Maryland

Monitoring Calibration
nethod" method0
Mast NBKI

Mast NBKI
Not given Not given

Mast NBKI. known



Chem Known 03
source,
UV
Chen

Mast IX NBKI.
Chem
Mast IX NBKI ,
Chem
Mast IX NBKI.
Chem
Fumigation '
facility0 Reference
OT Maclean and
Schneider
(1976)
OT
OT Heggestad
and Bennett
(1981)
OT Howe)) et
al. (1979)
Howell and Rose
(1980)

OT Duchelle et
al. (1983)

OT. Heggestad et
al. (1980)
OT Heggestad et
al. (1980)
OT Heggestad et
al. (1980)








X
1
i — >
-vl







-------
                                                                    TABLE X-4  (cont.)

                               EFFECTS OF  AMBIENT AIR  IN OPEN-TOP CHAMBERS, OUTDOOR CSTR CHAMBERS,  OR GREENHOUSES ON GROWTH AND YIELD OF SELECTED  CROPS
03 concn. ,
Plant species pp«
(Astro) >0.05
Snap bean >O.OS
(Gal latin 50)
(BBL 290) >O.OS
(BBL 274) >0.05
Sweet corn >0.08
(Bonanza)
(Monarch Advance) 0.08
Exposure duration
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
58% of hr (0600-2100)
between 1 July and
6 September

Percent
reduction Location Monitoring . Calibration Fumigation
from control of study Method method facility Reference
6, pod dry wt Maryland Mast
j
+1, pod dry wt Maryland Mast
10, pod dry wt Maryland Mast
22d, pod dry wt Maryland Mast
9 , ear fresh wt; California Hast
10 , no. seeds/ear
28^, ear fresh wt;
42 iK, no- seeds/ear
IX NBKI , OT
Chent
IX NBKI OT
Chem
IX NBKI OT
Chem
IX NBKI OT
Chen
UV OT

Heggestad et
al. (1980)
Heggestad et
al. (1980)
Heggestad et
al. (1980)
Heggestad et
al. (1980)
Thompson et
al.. 1976a

 Chem = chemiluminescence;  Mast = Mast oxidant meter (coulombmetrk); UV = ultraviolet spectrometry.
 NBKI = neutral buffered potassium iodide; UV = ultraviolet spectrometry.
 COT = open-top chamber;  CSTR = continuous  stirred tank reactor.
 Significant  at p = 0.05.
 eTotal above  ground biomass, 3 yr average; NF and open plot versus  CF a significant at p = 0.05

Source:   0   Criteria  Document,  U.S.  EPA,  1986

-------
                                          X-19
                          TABLE X-5.  EFFECTS OF OZONE ON CROP YIELD
                     AS DETERMINED BY THE USE OF CHEMICAL PROTECTANTS
   Species
Yield reduction,
 % of control
03 exposure,
    ppm
Reference
Beans (green)          41
Onion                  38
Tomato                 30
Bean (dry)             24
Tobacco                18
Potato                 36
Potato                 25
                  >0.08 for total
                  of 27 hr over
                  3.5 months

                  >0.08 on 5 days out
                  of 48

                  >0.08 on 15 days
                  over 3 months

                  >0.08 on 11 days
                  (total of 34 hr)
                  over 3 months

                  >0.08 on 14 days
                  during the summer

                  >0.08 ppm on 18 days
                  (total of 68 hr)
                  over'3 months
                     Manning et al. (1974)
                     Wukasch and Hofstra (1977b)
                                                           Legassicke and Ormrod (1981)
                     Temple and Bisessar (1979)
                     Bisessar and Palmer (1984)
                                                           Bisessar (1982)
                                                           Clarke et al.  (1983)
 All  the species were treated with" the antioxidant,  EDU,  except the bean study by
 Manning et al.  (1974) which used the systemic fungicide, benomyl.

 Yield reduction was  determined by comparing the yields of plants treated with
 chemical  protectants (control) to those that were not treated.
 This study was  run over 2  years when the 03 doses were 65 and 110  ppm-hr
 respectively, but the yield loss was similar both years.
 Source:   C>3  Criteria Document, U.S. EPA,  1986

-------
                                 X-20
                    .                     « — studies
       and to enable o3-induced   antl°Xidants to  improve
 -eral crop speoies/ ~ «°' »•• to  be estimate.
 concentrations occurring 
-------
                              X- 21
 threshold,  are the plateau-linear model and Weibull model.  The
 plateau-linear model incorporates a threshold value but does not
 allow curvature of any increase in yield followed by a decrease.
 The Weibull model, however,  can take on a plateau shape followed
 by curved gradual decreases.
      National Crop Loss Assessment Network (NCLAN)  yield losses
 have been modeled by Heck et al.  (1983,  1984),  who used a three
 parameter Weibull function.   The  Weibull function chosen has a
 flexible form,  which covers  the range of observed biological
 responses,  is biologically realistic,  has parameters with clear
 interpretations,  and offers  a method of  summarizing species
 responses by developing a common  proportional model (Rawlings and
 Cure,  1985).   The Weibull model subsequently has  been used with
 NCLAN data  previously modeled with  other functional forms'(CD,  p.
 6-165).
      b.   Statistics  Used to  Characterize Ozone  Exposures
      Important  factors needed to  characterize 03  exposures
 adequately  in vegetation exposure studies include time,
 concentration,  and dynamic nature (i.e.,  constant or variable).
 Which  of  the  components of exposure  actually causes plant
 response, however, remains unclear.  Although many  studies  tend
 to  report exposures  as mean O3  concentrations, the  exposure
 statistic can be  peak hourly,  daily, weekly,  monthly,  or seasonal
 means; number of  hours above  selected  concentration/  or  number  of
 hours  above selected  concentration intervals.  None of these
 statistics has  adequately  characterized  the  relationship among
 concentration,  exposure duration, interval between  exposures  and
 plant  responses.   The  lack of  correlation between exposure
 statistics and  ambient  air measurements  has  posed a major problem
 for those trying to assess the  effects of O3 exposure.
     The implication  inherent  in the use of mean 03
 concentrations  is  that  all 03 concentrations are equally
 effective in  causing plant responses.  Use of the mean statistic
minimizes the importance of peak concentrations by treating low-
 level  long-term exposures the same as high concentration short-

-------
                              X- 22
 term exposures.   Studies with beans and tobacco,  however,  have
 shown that a given dose (concentration x time)  distributed over a
 short period induced more injury  than did the same dose
 distributed over a longer period  (Heck et al.,  1966).
 Concentration was shown to be approximately  twice as  effective as
 time of  exposure at causing foliar  injury in tobacco  plants
 (Tonneijck,  1984).   The results of  several other  studies also
 have supported the greater importance of concentration in
 comparison to time in determining plant response  (Bennett,  1979:
 Heck and Tingey,  1971;  Henderson  and  Reinert,  1979; Reinert and
 Nelson,  1979;  Amiro et  al.,  1934).  Thus,  a  judgment  must  be made
 as to whether greater protection  of plants from O3 exposure is
 provided by  limiting short-term peak  exposures, long-term  average
 exposures,  or both.
      Not only are  concentration and time important but the
 dynamic  nature of  the 03 exposure is also  important; i.e.,
 whether  the  exposure is  at  a  constant  or variable concentration.
 Musselman  et  al.  (1983)  recently  demonstrated that although
 constant concentrations  cause the same  type  of plant response as
 variable concentrations  at  equivalent doses,  the constant
 concentrations had  less  effect  on plant  growth responses.   This
 finding  has been confirmed  by other studies.   Ashmore  found
 significant yield reductions  in radishes  exposed to ambient O3
 when  the maximum O3 concentration exceeded 0.06 ppm at least 10%
 of the days when the  crop was growing.   Initial studies have
 compared the response of alfalfa to daily peak and episodic O3
 exposure profiles which had the equivalent total O3 dose over the
growing  season (Hogsett et al., 1985) alfalfa yield was reduced
to a greater extent  in the episodic that the daily peak exposure.
The plants that displayed the greater growth reduction  (in  the
episodic exposure) have a significantly  lower 7-hour seasonal
mean concentration, thus raising the possibility that the 7-hour
seasonal mean may not properly consider peak  concentrations.
Studies with SO2  also showed that  plants exposed to variable
concentrations exhibited a greater plant response than those

-------
                               X- 23
  exposed to a constant concentration  (McLaughlin et al.  1979)
  in addition, the mean does not specifically include an exposure
  duration component.  Thus, it cannot distinguish between two
  exposures to the same concentration, but of different durations.
  c.   Exposure and Response to Peroxyacetyl Nitrate
       Response to .peroxyacetyl nitrate (PAN)  exposure has been
  assessed by use of the limiting-value method to estimate the
  lowest PAN concentration and exposure duration required to
  produce visible injury in plants  (Jacobson,  1977) .   The range of
  PAN  levels and  exposure  durations  which  this, analysis  suggests
  will  produce  foliar injury are:  1)  0.20 ppm for  0.5 hr;  2)  O.lo
  ppm  for l.o hr;  and 3) 0.035  ppm for 4.0 hr.  (CD, p. 6-215)    TO
  reduce  the likelihood  of  foliar injury to some  plants,  more
  recent  studies have suggested that PAN exposures may need to  be
  30 to 40%  lower  than the  values cited above  (Tonneijck, 1984)
      Few studies are available which report the effects of PAN on
 growth and yield.  A study of the effects on non-vascular plants
 indicates that growth, photosynthesis, and respiration of algae
 can be adversely affected by exposure to PAN (Gross and Dugger
 1969).  Fumigation of lichens with 0.05 ppm and O.lo PPm PAN for
 several days inhibited photosynthesis (sigal and Taylor  1979)
 Vascular plants which have been studied to determine the effects
 of PAN on growth and yield include  radishes,  oats,  tomatoes,
 Pinto beans,  beets,  and barley.  These plants were exposed in
 greenhouse studies to PAN concentrations  up  to 0.04  ppm for  4
 hours per day, twice per  week,  from germination to maturity,  with
 no significant effects  (Taylor et al.,  1983, .   In the same study
 similar  exposures of lettuce  and Swiss  chard  produced yield
.losses of  13%  and 23%,  respectively,  without  visible foliar
 «3«ry.  Although it is possible for  severe PAN  exposure to make
 some crops unmarketable, it is unlikely that PAN concentrations
th  n                                                      cur in
a ^th  7ter.6XCePt ^^^ ln — «"• of California and
a few other localized areas.

-------
                              X- 24
 d.   Economic Assessments of Agriculture
      Evidence from the plant science literature clearly
 demonstrates that 03  at ambient levels  will  reduce  the yields of
 some crops (CD,  p.  6-246).   The fact that such reductions in U.S.
 agriculture could adversely affect  human welfare has resulted in
 numerous  attempts to  assess,  in monetary terms,  the losses from
 ambient O3  or the benefits  of O3 control to agriculture.   Most  of
 the  past  1986 economic assessments  focus on  specific regions of
 the  country,  primarily California and the  Corn Belt (Illinois,
 Indiana,  Iowa, Ohio and Missouri).   This regional emphasis may be
 attributed  to the relative  availability of data  on  crop response
 and  air quality  for selected regions, as well  as the national
 importance  of these agricultural regions.  Of  the regional
 assessments evaluated in the CD (1986,  p.  6-247), Howitt et al.,
 1984a,b;  Rowe et al.,  1984; Adams and McCarl,  1985;  and Mjelde  et
 al.,  1984 are judged  to be  adequate  on  the basis of  the adequacy
 of the plant  science,  aerometric and economic  data  and  the
 assumptions used in each assessment.  Most regional  studies
 however abstract from the interdependencies that exist  between
 regions, which limits  their utility  in  evaluating secondary NAAQs'
 (CD, p. 6-252).
     National-level studies can overcome this  limitation  of
 regional analyses by  accounting for  economic linkages between
 groups and  regions.  This requires additional data and more
 complex models and  frequently poses more difficult analytical
problems.   Most  of the national assessments conducted since  1978
 suffer from either plant science and aerometric data problems,
 incomplete economic models,  or both.  As a result of the
limitations, caution should be applied in using these estimates
to evaluate the efficiency of alternative secondary NAAQS.  Two
of the studies, both based on NCLAN data, will be briefly
summarized here.   Both are judged  to be adequate in terms of the
three critical areas of data inputs  (CD, p. 6-252) and together
they provide reasonably comprehensive estimates of the economic
consequences of changes in ambient levels of  O3 on agriculture.

-------
                              X- 25
      In the first of these studies, Kopp et al. (1984) measured
 the national economic effects of changes in ambient 03 level on
 the production of corn, cotton, soybean, wheat and peanuts.  In
 addition to accounting for price effects on producers and
 consumers,  the assessment methodology places emphasis on
 developing  producer level responses to O3 induced  yield changes
 (from NCLAN data)  in 200 production regions.  The  results of the
 Kopp et al.  (1984)  study indicated that a reduction in 03 from
 1978 regional ambient levels to a seasonal 7-hr average of
 approximately 0.04  ppm would result in a $1.2  billion net benefit
 in 1978 dollars.  On the other hand,  an increase in 03 to an
 assumed ambient concentration of 0.08 ppm (seasonal 7-hr average)
 across all  regions  produced a net loss of approximately $3.0
 billion (CD,.p.  1-87).
      The second study,  by Adams etal.  (1984b)'isa component of
 the  NCLAN program.   The results were  from an economic model of
 the  U.S.  agriculture sector individual farms models for 55
 production regions.   Using NCLAN data,  the analysis examined
 yield changes  for six major crops (corn soybeans,  wheat,  cotton,
 grain,  sorghum and  barley)  that together account for over 75
 percent of U.S.  crop acreage.   The  estimated annual benefit (in
 1980  dollars)  from  03 adjustments are  substantial,   but make  up  a
 relatively small percentage of  total  agricultural  output  (about 4
 percent).  In  this  analysis a  25  percent reduction in  O3  from
 1980  ambient levels, resulted  in benefits of  $1.7 billion.   A 25
 percent  increase in  O3 resulted in an annual loss  (negative
 benefit)  of 2.363 billion.  When  adjusted  for differences in
 years and crop coverages,  these estimates  are quite close to the
Kopp et al.  (1984) benefit  estimates  (CD, p. 1-88).
     While the estimates from both of Kopp et al.  (1984)  and
Adams et al. (I984b) were derived from conceptually sound
economic models and the most defensible  plant science and
aerometric data available at the time of CD closure, there are
several sources of uncertainty  including the following:

-------
                              X- 26
      •    the issue of exposure dynamics and whether the 7-hr
 seasonal mean is an appropriate exposure statistic,
      •    the lack of environmental interactions in  the
 experiments,  particularly water stress,
      •    the lack of rural  monitoring sites in the  SAROAD system
 of  EPA,  which is necessary for extensive validation  of Kriging
 data,
      •    the economic models  themselves contain many potential
 sources  of uncertainty including  the  effects of benefits
 estimates of  market-distorting factors such  as  the Federal Farm
 Programs.   In addition to the  revised NCLAN  economic assessment,
 which  will be published in 1983,  there have  been other analyses
 that explicitly  consider the Federal  Farm Programs (see SP,  p.
 XI-10).
     Despite  these  uncertainties  these two studies,  in
 combination with the  underlying NCLAN data on yield  effects,
 provide  the most comprehensive economic  information  to date  on
 which  to base decisions  regarding the economic  efficiency  of
 alternative seqondary  standards (CD,  p.  1-88).

 B.   Natural  Ecosystem  Effects
     The previous section  discussed the  responses of  individual
 species  of  agricultural  plants, trees  and other  native  vegetation
 to 03.  The responses, which  are well documented, include:   (1)
 injury to  foliage,  (2) reductions in  growth,   (3) losses  in yield,
 (4)  alterations  in reproductive capacity, and (5) increased
 susceptibility to pests  and pathogens.  This section discusses
the effects of 03 stress on simple and complex plant  communities
to illustrate that such  effects, because of the  interconnections
and relationships among  ecosystem components, can produce
perturbations in ecosystems.   Stresses placed on plant
communities and the ecosystems of which they are a part can
produce changes that are long lasting and that may be
 irreversible.   No attempts have been made to examine  the effects

-------
                              X- 27
 of PAN on ecosystems since there are little data, and trees and
 other woody plants appear to be resistant to PAN (CD, p. 7-1).
      Evidence indicates that any impact of O3 on ecosystems will
 depend on the responses to O3 of the "producer" community.
 Producer species (trees and other green plants) are of particular
 importance in maintaining the integrity of an ecosystem since
 producers provide,  through the process of photosynthesis, all of
 the new organic matter (food, energy)  added to an ecosystem.
 Ozone-induced changes in photosynthesis influence energy
 flow and mineral nutrient cycling.   Alteration of these processes
 in ecosystems can set the stage for changes in community
 structure by influencing the nature and direction of successional
 changes (Woodwell,  1970;  Bormann,  1985),  with possibly
 irreversible consequences (see e.g.,  Odum,  1985,  Bormann,  1985).
 The sequence of responses outlined  by Bormann is  given in Table
 X-6 to assist in understanding ecosystem  effects  (CD,  p.  7-4).

      1.   Forest Ecosystems
      Anthropogenic  stress on forested  ecosystems  may result in
 reductions  in regional  tree  growth  and  a  decline  of  stands  in
 more  susceptible forest types (Johnson  and  Siccama,  1983).
 Atmospheric  pollutants, in particular oxidants  and acid
 deposition  are  important  regional factors that  are thought  to
 play  a  major role in  the  array  of stresses  affecting forests
 (Smith  1981;  Bormann  1982, 1985).   Observations of naturally
 occurring symptoms  or reduced growth as a result  of  acidic
 deposition are  limited  at the present time.  New  research results
 on  the  effects  of acidic deposition on forest are expected  emerge
 over the next two to five years.
     While the  nature and magnitude of the effects of atmospheric
pollution on North American  forest are still relatively unknown,
there is evidence that  some  forest types are negatively affected
by ambient levels of O3.  Among the more susceptible forested
areas are the mixed conifer forests of the San Gabriel and San
Bernardino mountain ranges east of Los Angeles, which have been

-------
                                   X-28
           TABLE X-6.   CONTINUUM OF CHARACTERISTIC ECOSYSTEM
                     RESPONSES TO POLLUTANT STRESS
Pnase                                   Response characteristics
 0                       No response occurs.  Manmade pollutants are absent'or
                         constitute insignificant stress.  Plant growth occurs
                         under natural conditions.

 I                       Ecosystems serve as sinks for pollutants.  Species
                         and/or ecosystem functions are relatively unaffected.
                         Self-repair occurs.

 11                      Sensitive species or individuals are subtly and
                         adversely affected.  A reduction in photosynthesis,
                         a change in reproductive capacity, or a change in
                         predisposition to insect or fungus attack may occur.

111                      Decline occurs in the populations with sensitive
                         species; some individuals will be lost.   Their ef-
                         fectiveness as functional members of the ecosystem
                         diminishes.   Ultimately, species could be lost from
                         the system.

 IV                      Large plants, trees, and shrubs of all species die.
                         The basic structure of the forest ecosystem is changed.
                         Biotic regulation is affected as forest layers are
                         peeled off:   first trees, tall shrubs, and,  under
                         the most severe conditions,  short shrubs and herbs.
                         The ecosystem is dominated by weedy species  not
                         previously present and by small scattered shrubs
                         and herbs.

  v                      The ecosystem collapses.   The loss of species  and
                         changes in ecosystem structure, nutrients,  and soil
                         so damage the system that self-repair is impossible.
Source:   Adapted from Bormann (1985).

-------
                              X- 29
 exposed to oxidant pollution since the early 1950's (Miller,
 1973).   Ozone was first identified as the agent responsible for
 the slow decline and death of ponderosa pine trees in these
 forests in 1962  (Miller et al.,  1963).   Later,  Jeffrey pine was
 also found to be injured by O3 exposure.   Oxidant injury of
 eastern white pine has been observed for many years in the
 eastern United States (although  some genotypes  of white pine
 appear  to be more tolerant).   It was first reported as needle
 blight  in the early 1900s but in 1963 was shown to be the result
 of  acute and chronic ozone exposure (Berry and  Ripperton,  1963).
      More recently,  oxidant injury of eastern white pine in the
 Blue Ridge Mountains of Virginia has been reported by Hayes and
 Skelly  (1977), Skelly 1980,  and  Benoit  et al.  (1982)  and on the
 Cumberland Plateau of east Tennessee by Mann et al.  (1980)  and
 McLaughlin et al.  (1982).   Ozone injury in natural  plant
 communities has  been reported by Treshow and Stewart (1973)  and
 by  Duchelle et al.  (1983).   In addition,  there  is evidence  from
 laboratory studies of visible injury, negative  effects  on
 physiological function and reduced productivity as  a result of
 oxidant  pollution  (Guderian,  1977;  Smith,  1981).  However,
 because  of the disparity  between effects  observed in the
 laboratory and those observed in the  field,  results  from
 controlled laboratory and  greenhouse  studies  are  not easily
 extrapolated  to  field conditions and  there  is very  little
 experimental  evidence directly linking  ambient  ozone
 concentrations with  decline of tree productivity  in  the  field.
 This  section  will  discuss  the effects of  03 on plant processes
 and growth  and the  limited data  available on  ecosystem response.
      a.   Effects on  Plant Processes
     A discussion of  individual  tree response is  the first  step
 in explaining ecosystem response.  In forest ecosystems, trees
play a critical role.  As producers, trees  influence the
structure, energy flow and nutrient cycling of forest ecosystems
 (Ehrlich and Mooney,  1983).  According to the CD, while some of
the same plant processes are affected in trees and agricultural

-------
                              X- 30
 crop species,  perennial plants,  because they live longer,  must
 cope with both short- and long-term stresses,  the effects  of
 which can be cumulative,  lasting over the  years,  or can be
 delayed,  not becoming apparent  for  many years.   Likewise,  effects
 can  possibly be mitigated through short-or long-term recovery or
 replacements (CD, p.  7-23).  Therefore,  the permanent vegetation
 in natural ecosystems receives much greater chronic exposure than
 the  short-lived vegetation that  makes up agroecosystems.   The
 single agroecosystem  has  little  resilience to pollutant stress;
 the  natural ecosystem is  initially  more  resistent to pollutant
 stress because of species diversity,  but the longer chronic
 exposures  can  disrupt the system.   These differences between
 natural ecosystems and agroecosystems raise  a key issue in the
 debate on  the  03 secondary standard regarding the adequacy of  the
 current 1-hr standard and other  exposure indicators  under
 consideration  to protect  trees as well as  crops.
      In regard to the effects of 03 on individual tree  response,
 inhibition  or  reduction in the rate of photosynthesis  is possibly
 the most significant  effect of 03 entry  into -the leaves of
 sensitive plants, although other mechanisms  such  as  increased
 foliar leaching have  been suggested  (Taylor  and Norby,  1985).
 Ozone  inhibits photosynthesis,  decreases formation of organic •
 compounds needed for  plant growth, and can alter the transport
 and allocation of the  decreased products of photosynthesis so
 that sugar  storage and root growth are reduced.  Trees  in which
O3 has been shown to reduce photosynthesis  are loblolly pine and
 slash pine  (Barnes,  1972), ponderosa pine  (Miller et al.,  1969;
Coyne and Bingham, 1981), eastern white pine, (Barnes,  1972, Yang
et al., 1983; Botkin et al.,  1972) black oak, sugar maple
 (Carlson,  1979), and a poplar hybrid  (Furukawa and Kadota,  1975)
 (Table IX-1, CD, p.  6-28 or SP,  pg.  IX-4).
     Miller et al. (1969), Coyne and Bingham (1981), and Yang et
al.  (1983) relate visible injury symptoms and reduced tree growth
to the effect of 03  on photosynthesis.  Miller et al. (1969)
found that exposure  of 3-year old ponderosa pine seedlings under

-------
                              X- 31
 controlled conditions to concentrations of 0.15 and 0.30 ppm 9
 hr/day for 30 days reduced photosynthesis by 10 and 70 percent,
 respectively,  in addition,  it was noted that a reduction in
 photosynthesis was accompanied by a decrease in the sugar content
 of injured needles.   Tingey et al.  (1976)  observed that the
 amounts of soluble sugars,  starch,  and-phenols tended to increase
 in the tops and decrease in the plant roots of ponderosa pine
 seedlings exposed to 0.10 ppm 03 for  6  hours per day for 20
 weeks.   The sugars and starches stored in the tree roots were
 significantly less when compared with the controls (CD,  p.  7-14).
      Coyne and Bingham (1981)  measured photosynthesis and
 stomatal conductance of attached ponderosa pine needles  in
 relation to cumulative 03 dose.   The  decline in photosynthesis
 and stomatal function normally associated  with aging was
 accelerated as 03  injury  symptoms increased.   Premature
 senescence and abscission of needles  occurred soon after
 photosynthesis reached its  lowest level.   In a study of  white
 pine, Yang et al.  (1983)  also  observed  that decrease in  rates of
 photosynthesis due to O3 exposure was closely  associated with
 visible  needle injury,  premature senescence and reduction in
 biomass  (CD,  p.  7-13).  Thus,  the impact of O3  on  important plant
 processes  such as  photosynthesis does seem to  be reflected  in the
 occurrence of other  symptoms of  03 injury.
           b.   Effects on Growth
     The limited observations  of tree response  to  03 in the field
 and the chamber data  that is currently available (Table X-7)
 provide the  strongest evidence to date of  tree  response to 03.
 Studies made  along the Blue Ridge Parkway  support  the view that
 exposure to 03 reduces growth in sensitive trees  (Benoit et al.,
 1982).  Eastern white pine located in experimental plots situated
 along the Blue Ridge  Parkway were studied  to determine the radial
growth increment during the 1955 to 1978 period.  Growth of trees
classified as sensitive was 25 percent less than in tolerant
trees; growth was 15 percent less in trees classified as
 intermediate  in sensitivity.  Mean radial  increments for all

-------
                              X- 32
trees during  the  last  10 years of the  study were  smaller than for
the previous  24 years.  During the period of the  study,
concentrations of 0.05 to  0.07 ppm of  03 were recorded on a
recurring basis,  with  episodic peaks of 0.12 ppm  or higher
occurring  (Benoit et al.,  1982).
     Hayes and Skelly  (1977)  monitored total oxidants and
recorded oxidant  related injury on eastern white  pine in three
rural Virginia sites between  April 1975 and March 1976.  Injury
was associated with total  oxidant peaks of 0.08 ppm for  1-hr  or
more.  Peak 1-hr  ozone concentrations  of 0.17 have been  measured
in the Blue Ridge Mountains (Skelly, 1980).  Monthly 8-hr average
O3 concentrations ranged from 0.035 to 0.065 ppm during  the
oxidant season (April - October)  and peak hourly  concentrations
ranged from 0.08  to 0.13 ppm  (Skelly et al., 1984).  Increased
injury symptoms were observed on pine  trees previously classified
as sensitive or intermediately sensitive after an O3 episode.
Hayes and Skelly  (1977) suggested that continued  exposure of
white pine to acute and chronic oxidant concentrations could
influence their vigor and reproductive ability, ultimately
resulting in replacement by tolerant species.
     A steady decline in annual ring increments was also noted on
the Cumberland Plateau of east Tennessee during the years 1962 to
1979 (McLaughlin  et al.,  1982).   A reduction of 70 percent in
average annual growth and 90 percent in average bole growth was
observed in sensitive white pine when compared to both trees
classified as tolerant to O3 and  trees  of  intermediate
sensitivity.   Decline was attributed primarily to chronic
exposure to 03,  which frequently  occurred  at phytotoxic
concentrations in the area.  For the years 1975-1979 the
incidence rates for hourly concentrations  >0.08 ppm ranged from
129-339 hours above 0.08  ppm.   Maximum 1-hr concentrations range
from 0.12 to 0.2 ppm during this  time period.   The reduction in
growth on the Blue Ridge  Parkway and on the Cumberland Plateau,
as in the case of the San Bernardino Mountains, was correlated
with the predisposing symptoms of chronic  decline, which includes

-------
                              X- 33
 the following sequence of events and conditions:  (1) premature
 senescence and loss of older needles; (2)  reduced storage
 capacity of carbohydrates in the fall and resupply capacity in
 the spring to support new needle growth; (3)  shorter new needles,
 resulting in lower gross photosynthetic productivity; (4) reduced
 availability of photosynthate for external usage (including
 repair of chronically stressed tissues of  older needles); and (5)
 premature casting of older needles (McLaughlin et al.,  1982).
 Degeneration of feeder roots and mycorrhizae  usually precedes the
 onset  of above ground symptoms (Manion,  1981).   Decreases in
 nutrient and water uptake may also occur.   These changes produce
 weakened trees.   Weakened trees are,  in turn,  predisposed to
 attack by root rot fungi,  to defoliation by insects,  and to
 attack by the pine beetle.   Growth reductions  in trees  (Table  X-
 7)  grown under controlled conditions  have  also  been  observed
 (Mooi,  1980;  Kress et al.,  I982a,b,c;  Kress and Skelly,  1982;
 Jensen,  1979;  McClenahen,  1979;  Jensen and Dochinger, 1974;
 Jensen and Masters,  1975)  (CD,  p.  6-134).
     Injury by 03  to  native herbaceous vegetation growing  in the
 Virginia  mountains was  also  observed  (Duchelle  et al., 1983).
 Ambient  O'3  concentrations were shown to reduce  growth and
 productivity  of graminoid and  forb vegetation in the  Shenendoah
 National  Park.  For each year  of the study, biomass production
 was greatest  for vegetation grown  in filtered-air chambers.  The
 total  3-year  cumulative dry weight of plants in filtered chambers
 was significantly  greater than that in non-filtered and open-air
 plots.   Common milkweed and common blackberry were the only two
 native species to  develop foliar injury.  Ozone episodes occurred
 several times during the period of the study.    Peak hourly
 concentrations ranged from 0.08 to 0.12 ppm; however, monthly
hourly average concentrations ranged from 0.03  to 0.06 ppm.

-------
 Poplar
 (Oorskamp)
 (Zeeland)
(16-SVC-23)
  (16-SYC-23)

  Sweetgum



 African Sycamore



 White ash



Green ash

     Exposure duration

     12 hr/day , 5 mo
                            0.041
                           0.05
                           0.05
                          0.05
                          0.05

                          0.05
                          0.10
                         0.15

                         0.05
                         0.10
                         0.15

                        0.05
                        0.10
                        0.15

                        0.05
                        0.10
                        0.15
   6 hr/day. 28 days


   6 hr/day,  28 day


   6 hr/day, 28 days


  6 hr/day, 28 day

  6 hr/day,  28  days
 9  • height growth


 2.  height growth


 U. height growth

 a
9 .  height growth
 6 hr/day, 28 days      +*  , .
                y       27*    9ht 9rowth
 6  hr/day, 28 days



6 hr/day,  28 days
                                                                                     Mooi (1980)


Chem

Pu__
WIICHI
Chem

Chem
Chem


1* NBKI


Constant
source,
NBKI. UV
Constant
source.
NBKI-, UV
Constant
source,
NBKI. UV
Constant
source,
NBKI, UV
LH Kress et
«)• (19826)
•
CSTR Kress et
a>- (19826)

CSTR Kress and
Skelly
(1982)
CSTR

CSTR Kress and
Skelly
(1982)
CST* Kress and
Skelly
(1982)

-------
TABLE X-7.  (cont.)   tFFECTS OF OZONE ADDED TO FILTERED AIR ON YIELD OF  SELECTED TREE CROPS
Plant species
Willow oak
Sugar maple
Yellow poplar
Yellow poplar
Cottonwood
White ash
White ash
Black cherry
Hybrid poplar
(NS 207 + NE 211)
03
concn. ,
ppm
0.05
0.10
0.05
0.10
0.15
0.05
0.10
0.15
0.10
0.10
0.10
0.10
0.20
0.30
0.40
0.10
0.20
0.30
0.40
0.15
Exposure duration
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
12 hr/day, 48 days


4 hr/day, 1 day/wk,
9 wk

8 hr/day, 5 days/wk,
6 wk
Yield, •% Monitoring
from control method
1, height growth; 2, total dry wt Chem
4, height growth; 11,, total dry wt
5, height growth; 2, total dry wt Chem
+8* height growth; 7, total dry wt
12 , height growth; 41 , total dry wt
+60e, height growth; +41, total dry wt Chem
+8, height growth; +5, total dry wt
12, height growth; +18, total dry wt
19ng, relative growth rate Chem
59n9, relative growth rate
no significant effects
+13, total height; +7, shoot dry wt Not given
0, total height; +5, shoot dry wt
0, total height; 11, shoot dry wt
0, total height; 14, shoot dry wt
+16, total height; +15, shoot dry wt Not given
+5, total height; 4, shoot dry wt
+3, total height; 4, shoot dry wt
28 , total height; 15. shoot dry wt
p
50 , dry wt new shoots from terminal Not given
cuttings
62 , dry wt new shoots from basal
Calibration Fumigation
methodc facility0
Constant CSTR
source,
NBKI, UV
CSTR
CSTR
Not given CSTR


Not given Not given
Not given Not given
Not given GH-CH
Reference
Kress and
Skelly
(1982)
Kress and
Skelly
(1982)

Jensen
(1981)


McClenahen,
(1979)
McClenahen,
(1979)
Jensen and
Dochinger
(1974)
                               cuttings

-------
TABLE X-7.  (cont.)
EFFECTS OF OZONE ADDED TO FILTERED AIR ON YIELD OF SELECTED CROPS
Plant species
Hybrid poplar
(207)
Yellow birch
White birch
Bigtooth aspen
Eastern cottonwood
Red maple (163 ME)
(167 NB)
(128 OH)
Loblolly pine
(4-5 x 523)
(14-5 x 517)
Loblolly pine
Pitch pine
03
concn. ,
ppra
0.20
0.20
0.25
0.25
0.25
0.25
0.25


0.05
0.05
0.05
0.10
0.15
0.05
0.10
0.15
Yield, % Monitoring Calibration Fumigatiog
Exposure duration from control3 method method facility
7.5 hr/day, 5 day/wk, 5, height Not given Not given CH
6 wk 8. height
8 hr/day, 5 day/wk, 9, height MAST NBKI GH-CH
15 wk
34, height
+7, height

8 hr/day, 6 wk 18, height MAST « NBKI CH
32, height
37e, height
6 hr/day, 28 days 6, height growth Chem 1* NBKI CH

6 hr/day, 28 days 18e, height growth; 14 total dry wt Chem Constant CSTR
27e, height growth; 22 , total dry wt source,
41e, height growth; 28e, total dry wt NBKI, UV
6 hr/day, 28 days 4 height growth; 8, total dry wt
13 , height growth; 19. total dry wt
26e, height growth; 24 , total dry wt
Reference
Jensen
(1979)
Jensen and
Masters
(1975)



Dochinger
and Town-
send (1979)


Kress et
al. (1982a)

Kress and
Skelly
(1982)


-------
    Plant species
   •
 Virginia pine
 White spruce



 Japanese larch
                            TABLE  X-7.  (cont.)
                                                          EFFECTS OF. OZONE ADDED TO FILTERED AIR ON YIELD OF SELECTED CROPS
0.05
0.10
0.15

0.25
 Exposure duration
•	
 6 hr/day, 28 days
            8 hr/day.  5 day/wk,
             15 wk
                                                 Yield. %
                                                from control3
 5. height growth; +2, total  dry wt
11. height growth; 3. total dry wt
14, height growth; 13. total  dry wt
                                        Mm*Jh°$n9    Ca>ibratcon    legation
                                         method        methodC       facility" Reference
                                                                                                    Chem
                                                                          Mast
                                                                             Constant
                                                                             source,
                                                                             NBKI, UV

                                                                             NBKI
                                                                                                                                 CSTR
                                                                                                      GH-CH
                                                                                                                                GH-CH
Source:    0;|  Criteria Document, U.S.  EPA,  1986
Kress and
Skelly (1982)


Jensen and
Masters
(1975)

-------
                              X- 38
      Treshow and Stewart (1973)  conducted one of the few studies
 investigating the impact of air  pollution on natural plant
 communities.   Grassland,  oak,  aspen,  and  conifer communities in
 Salt Lake Valley,  Utah were studied.   Two dominant species,
 cheatgrass and aspen,  considered key  to community integrity  were
 found to  be sensitive.   In  both  cases single 2-hr exposures  to
 0.15 ppm  03 caused severe injury.  Removal of the dominant
 species from  plant communities could  result  in  a shift  to another
 species.   In  a companion  study conducted  in  portable plastic
 chambers,  03  exposures  of 0.15 to 0.3 ppm for 3-hr per  day 5
 days/week throughout the  growing season reduced root and top
 growth'and fewer seeds  were produced  (Harward and Treshow, 1973).
           c.   Ecosystem Responses: The San Bernardino Study
      The  interdisciplinary  study of the decline of the  pine  and
 mixed conifer forest in the San  Bernardino Mountains is the  most
 comprehensive and best  documented report available on the effects
 of oxidants on an ecosystem (Miller et al.,  1982).   While San
 Bernardino may be regarded  as  a  worst case scenario  due to the
 high'levels of 03 found in  Southern California, it still  provides
 valuable  information on the potential consequences of ecosystem
 exposure  to O3.
      Many  of  the major  ecosystem processes were shown by  the
 study to  be affected directly or  indirectly.  Ozone  associated
 stress on  trees decreased photosynthesis,  affected directly or
 indirectly translocation of carbon (energy), mineral  nutrients,
 and water, and reduced  trunk diameter, tree height and  seed
 production in  ponderosa and Jeffrey pine (Miller et  al.,  1982).
 Foliar injury  of O3 sensitive ponderosa and Jeffrey pines was
 observed when  24-hr 03  concentrations  ranged from 0.05 to 0.06 '
 ppm.  During the period of the study,  average 24-hr O3
 concentrations during the months of May through September ranged
 from  a background of 0.03 ppm to a maximum of 0.10 to 0.12 ppm.
A comparison of radial  growth of ponderosa pine from 1910 to
 1940, a period of low pollution  (<0.03 ppm),  with the years 1941
 to 1971,  a period of high pollution (0.03  to 0.12 ppm),  indicated

-------
                              X- 39
 that 03  exposure reduced the average annual radial growth by
 approximately 40 percent, height by 25 percent,  and wood volume
 by 84 percent in trees less than 40 years of age.   The marketable
 volume of trees 30 years of age was reduced by 83  percent in the
 areas with the highest 03 concentrations  (Miller and Elderman,
 1977).   In addition,  stressed pines also  became  more susceptible
 to root  rot and pine  beetle as a  result of weakening by
 photochemical oxidants (Stark and Cobb, 1969).
      Ozone-induced stress on sensitive ponderosa and Jeffrey pine
 and to a lesser extent on sensitive white fir, black oak,
 increase cedar and sugar pine,  accompanied by  fire,  brought about
 the removal of the pine forest overstory.   A shift in dominance
 to self-perpetuating  fire adapted,  O3  tolerant shrub and  oak
 species  resulted.   These mixtures provide fewer  commodity and
 amenity  values than the former pine forest (Miller et al.,  1982).
      The case of San  Bernardino suggests  that  a  potential
 consequence of 03  stress  is  a  change in'the  composition and
 successional  patterns  of  some  plant communities  (Woodwell,  1974).
 With  regard to forests in eastern North America, both the  extent
 of  the decline and the causal  mechanisms  remain  controversial
 (Taylor  and Norby,  1985).   Changes  in  the  growth patterns  of
 eastern  white pine have  been attributed to  stress  resulting  from
 03  exposure that began 15 to 20 years earlier  (Miller and
 Elderman,  1977;  Miller et al.,  1982; McLaughlin  et al., 1982).
 More recently,  dendroecological studies of  the decline of  red
 spruce in  the northeast  (Johnson  and Siccama, 1983)  and of
 reduced  growth  rates of red  spruce, balsam  fir and frasier fir  in
 central West  Virginia  and western Virginia also  provide further
 evidence that  the  reductions in growth and mortality measurable
 today probably began at least 20 years ago.  In  addition,
 reductions  in growth rates of loblolly and short leaf pine
have been reported  in the piedmont regions of the southeastern
U.S. (McLaughlin,  1985).
     In regard to these most recent declines, there is currently
no agreement as to the trigger factor that precipitated the

-------
                              X- 40
 dieback,  mortality and decreased growth.   A number of stresses
 have been identified including natural processes and air
 pollution (Johnson and Siccama, 1983).  Given the regional
 distribution of 03 in North America,  the  frequent occurrence of
 elevated  O3  concentrations,  and the recognized effects of 03 on
 agricultural productivity in the  region,  the potential influence
 of 03  on  forest ecosystems  should be  given  high priority (Taylor
 and Norby,  1985).

      2.   Interrelated Ecosystems
          a.  Aquatic Ecosystems
      It is extremely important  to consider  that an adverse impact
 on a forest  or  agricultural  ecosystem may in turn adversely
 affect adjacent aquatic systems.  A variety of linkages  for
 energy and nutrient  exchange  exist.   Disruptions  induced by air
 pollution stress on  terrestrial ecosystems  often  trigger
 dysfunctions  in neighboring aquatic ecosystems, such as  streams,
 lakes, and reservoirs.  Sediments resulting  from  erosion can
 change the physical  character of  stream channels,  causing changes
 in bottom deposits,  erosion of channel banks,  obstruction of
 flow,  and increased  flooding.  They can fill  in natural  ponds and
 reservoirs.   Finer sediments can reduce water  quality, affecting
 public and industrial water supplies and recreational areas.
 Turbidity caused by  increased erosion can also reduce the
 penetration of  light  into natural waters.   This,  in turn, can
 reduce plant photosynthesis and lower supplies of dissolved
 oxygen, leading to changes in the natural flora and fauna
 (Bormann  and Smith,  1980).  Significant forest alterations,
therefore, may have a regional impact on nutrient cycling, soil
stabilization, sedimentation, and eutrophication of adjacent or
nearby aquatic systems. Interfacing areas,  such as wetlands and
bogs, may be especially vulnerable to impact  (CD,  p. 7-47).
          b.   Agricultural Ecosystems
     Natural and agricultural ecosystems possess the same basic
functional components, require energy flow and mineral cycling

-------
                               X-41
   for maintenance, and
claand                                 »»« influences of
simnT    ? SUbStrate-  »«t»»l «osyste»s vary l» diversity fro,
         SS
  simn
  spec eT    W     SW SPe°iSS t0 °°"'PleX ^sterns «*" -ny
  spec.es   Their populations also vary in genetic composition
  peLt  JPe°leS diVerSity>  *"•* a" -"-"^atin* and se f-
  perpetuat^g.  Agroecosystems ,  on the other hand,  are usually
  highly manipulated monocultures of similar genetic and age
  composition and are unable" to maintain themselves  without the

                 If any of the species, varieties, or cultivars is
  very sensitive to 03, its market value can be destroyed.  When
  this occurs, efforts are made to find a resistant cultivar, as
  With tobacco, or to grow a crop less sensitive to 03 stress
 C.  Materials. Damage
 th t oSSrrCh °VSr a Peri°d °f I°°re than tW°  decades  ^s  shown
 that 03 has the  ability  to  react with both manmade and natural
 totaT:^  rne some research has been d°ne °- ^  •««*•
 total oxidants on materials, the only components of total
 sTite^ tT "^ ^  "Udled ^^^ •» 03 and Ho
 " ro^o \     reSear°h  f°CUS °n  °3'  h°WeVer'  the «"
 from 03  to  actual  in-house materials remains poorly
 characterized.
      The materials known to be most susceptible to 03  attack are
            ™
                           and
                    and amount of 03
                                    mage functions
                        *«WJ.wlia.   ine economic impact  of  o
related damage could then be estimated by using accelerated

-------
                              X-42
 repair and replacement costs.  Because little recent work has
 been reported on the effects on nonbiological materials, however,
 it is necessary to rely on older studies which, in many cases, do
 not provide sufficient information to assess the amount and costs
 of oxidant related damage.

      1.   Elastomers
      The effects of 03 on  elastomers  are the best  documented.
 Natural  rubber and synthetic polymers/copolymers of butadiene,
 isoprene,  and styrene account for the bulk of elastomer
 production for products such as automobile tires and protective
 electrical coverings.   The mechanism  of  03  damage  on elastomers
 shares similarities and differences with simple oxidation from
 atmospheric oxygen.   Ozone damage, usually in the  form of
 cracking,  tends to be more of a surface  phenomenon than simple
 oxidation.   It is  greatly  accelerated by mechanical stress,  which
 produces  fresh surface areas  at  crack boundaries.   Simple
 oxidation,  on  the  other hand,  is slower;  it occurs more in the
 bulk  of a  material,  and it is less affected by  the degree of
 stress (Mueller and  Stickney,  1970).   At  pollutant concentrations
 and stress  levels  normally encountered outdoors  (and in many
 indoor environments),  the-elastomer hardens  or  becomes  brittle
 and cracked, losing  its  physical  integrity.   High  humidity and
 mechanical  stress  greatly  affect  the  formation, depth of
 cracking,  and,  in  automotive  tires, the adhesion piles  (Davies,
 1979; Wenghoefer,  1974)  (CD,  p.  8-3).
     03 affects natural rubber and other elastomers in a dose
 related fashion.   Dose  is  defined in materials research as the
product of concentration and  duration  of exposure.  The
 importance of O3 dose was demonstrated by Bradley and Haagen-Smit
 (1951), who used a specially  formulated 03 sensitive natural
rubber (NR).  Samples exposed to 20,000 ppm cracked almost
 instantaneously, and those exposed to  lower concentrations took a
proportionately longer time to crack.   Cracking occurred at a

-------
                              X- 43
 rate  of  0.02  to 0.03  ppm-hr over  the entire range of
 concentrations.
      Haynie et  al.  (1976)  exposed samples of a tire sidewall to
 O3 at concentrations  of  0.08  and  0.5 ppm  for 250  to 1000  hr under
 10-20 percent strain.  From these and other data,  they estimated
 that  at  the 03  standard  of  the  time  (0.08  ppm,  1-hr average),  and
 at the annual NOX standard  of 0.05 ppm, it would  take  2-5 years
 for a crack to  penetrate cord depth.   In  addition to stress,
 factors  affecting the cracking  rate  include atmospheric pressure,
 humidity,  sunlight, and  other atmospheric  pollutants.   Veith and
 Evans (1980)  found a  16-percent difference in cracking rates
 reported from laboratories  located at various geographic
 locations  (CD,  p. 8-48).
     Ozone has  been found to  affect  the adhesion  of piles
 (rubber-layered  strips)  in  tire manufacturing.  Exposures to 03
 concentrations  of 0.05 to 0.15  for a few hours  significantly
 decreased  adhesion in  an NR/SBR blend, causing  a  30 percent
 reduction  at  the highest 03 level.  This adhesion problem
 worsened at higher relative humidities.  Wenghoefer (1974)  showed
 that 03   (up to 0.15 ppm), especially  in combination with  high
 relative humidity (up  to 90 percent),  caused  greater adhesion
 losses than heat and NO2 did, with or without high  relative
 humidity (CD,  p. 8-49).

     2.   Textile Fibers  and Dyes
           The effects of 03 on dyes have been known for nearly
 three decades.  In 1955,  Salvin and Walker  exposed  certain  red
 and blue anthraquinone dyes to  0.1 ppm 03 and noted fading which
until that time had been thought to be caused by NO2.  Subsequent
work by  Schmitt  (i960, 1962) confirmed the  fading action  caused
by 03  and the  importance  of relative humidity in the absorption
and reaction of vulnerable dyes.  Later Beloin  (1972, 1973) noted
the acceleration in fading of certain dyes at an 03 concentration
of 0.05  ppm and a relative humidity of 90 percent (CD, p.  8-49).

-------
                              X- 44
      Both the type of dye and the material in which it is
 incorporated are important factors in the resistance a fabric has
 to 03.   Haylock and Rush (1976,  1978)  showed that anthraquinone
 dyes on nylon fibers were sensitive to fading from O3  at a
 concentration of 0.2 ppm for 16  hrs,  while Haynie et al. (1976)
 and Upham et al. (1976)  found no effects from .O3  concentrations
 of 0.1  to 0.5 on royal blue acetate,  red rayon-acetate,  or plum
 cotton.   Field studies by Nipe (1981)  and laboratory work by
 Kamath  et al.  1982  showed a positive  association  between O3
 levels  and dye fading of nylon materials at an 03 concentration
 of 0.2  ppm and at various relative humidities.  In summary, dye
 fading  is a complex function of  03  concentration,  relative
 humidity,  and the presence of other gaseous pollutants.   At
 present,  the available research  is insufficient for quantifying
 the  amount of  damage to  fibrous  materials  attributable to O3
 alone.  Anthraquinone dyes incorporated  into  cotton and  nylon
 fibers appear  to be the  most sensitive to  03  damage  (CD,  p.. 8-
 50) .
     The  degradation of  fibers from exposure  to 03  is poorly
 characterized.   In  general,  most synthetic  fibers  such as
 modacrylic  and polyester  are relatively  resistant;  and cotton,
 nylon, and  acrylic  fibers  show variable  sensitivities to  the  gas.
 Ozone reduces the breaking  strength of these  fibers, and  the
 degree of reduction  depends  on the  amount of  moisture present.
 Under laboratory conditions,  Bogaty et al.  (1952)   found a 20
 percent loss in breaking strength  in cotton textiles under  high-
moisture conditions  after exposure to a  0.06  ppm concentration of
03 for 50 days.  They equated these conditions to  a 500- to 600-
day exposure under natural conditions.  Kerr  et al.  (1969)  found
a net loss of 9 percent in breaking strength  of moist cotton
fibers exposed to 03 at a concentration of 1.0 ppm for  60 days.
The limited research in this area indicates that 03 in  ambient
air may have a minimal effect on textile fibers, but additional
research is needed to verify this conclusion  (CD,  p. 8-50).

-------
                                X-45
       3.  Paints
       The effects of O3 on paint are small in comparison with
  those of other factors  (Campbell st al., 1974) .  Past studies
  have shown that, of various paints, only vinyl and acrylic coil
  coatings are affected (Haynie et a!., 1976) , and that this impact
  has a negllgible effect on the useful life of the material
  coated.   Preliminary results of current studies have indicated a
  statically significant effect of 03 and  relative humidity on
  latex house paint,  but final results are needed before
  conclusions can be  drawn.

       Pigments in artists'  paints have also  been tested under
  controlled  conditions  for  3  .onths  at an average  exposure  level
  of  0.4 ppm  of o3.  While fading  occurred  in anthraquinone-based
  Pigments, no  quantitative  information on  dose-response
  relationships is available.


      4.  Conclusion

      Among the various materials studied, research has narrowed
 the type of materials most likely to affect the ec      from
 increased 03 exposure.   These include easterners and textile
  ro,                                      '             «•  -
 probably the most economically important.   Hhile the limitations
 of McCarthy et al.  (1983)  preclude the ^^ ^^^ °
               the
 d           >        gures  ndicate the magnitude °f p«
 damage from exposures to 03  in  ambient air  (CD,  p.  a-46) .

 D.   Effects on  Personal  Comfort and Well-Being
     The Clean  Air Act requires that secondary NAAQS  for a
 pollutant specify a  leve! of air quality which is adequate to
 protect public  welfare from any known  or anticipated  adverse
 effects,  section 302 (h) includes personal comfort  and well-beina
 ,n referring to effects on welfare.  Those effects  of a
on humans which are not identified as being adverse health
effects but do affect personal comfort and well-being are covered
by th» Provlslon.   symptoms are defined generally as subjective

-------
                              X- 46
 evidence  of  disease  or  physical disturbance  but  are  not
 necessarily  adverse  health  effects.
      Symptomatic  effects  associated with human exposure to O3  and
 other photochemical  oxidants  may  contribute  to a reduction in
 personal  comfort  and well-being.  Similar, but not identical,
 symptoms  have been reported for clinical 03  and  community-
 photochemical oxidant exposures.  Eye irritation, for  example,  is
 commonly  associated  with  ambient  photochemical oxidant levels  of
 about 0.10 ppm but does not occur during controlled  O3  exposures
 at much higher levels than  found  in ambient  air.  Symptoms such
 as nose and  throat irritation, chest discomfort, cough and
 headache  have been reported at >  0.10 ppm O3 in  epidemiology
 studies (Hammer et al., 1974; Makino and Mizoguchi,  1975;  Okawda
 et al., 1979) and at >  0.12 ppm in controlled studies  (McDonnell
 et al., 1983; Avol et al.,  1984;  Kulle et al., 1985).   Impairment
 of athletic  performance in  high school students  may  have been
 caused by irritative symptoms associated with 0.12 ppm oxidants
 (Wayne et. al., 1967).  Other symptoms commonly  reported in
 clinical 03 studies are throat dryness,  difficulty or pain during
 deep  inspiration, chest tightness, substernal soreness  or  pain,
 wheezing,  lassitude, malaise,  and nausea.   CASAC recommended that
 these effects be considered health effects in developing a basis
 for the primary standard for O-,.

 E.  Related Welfare Effects Information and  Issues
     In its public meetings of April 20-21,  1986, December 14-15,
 1987, and December 14-15,  1988 on the CD,  staff paper,   and NAAQS-
related analyses, the Clean Air Scientific Advisory Committee
 (CASAC)  discussed  preliminary new information relevant to the
possible need for additional standard(s)  to protect public health
 from exposure to 03.   At the time  of the meetings much  of the
 information was unpublished and not incorporated into the CD.
 Subsequently this new information base has been reviewed and
 largely incorporated by ECAO into the CDS.   For a more detailed

-------
                              X- 47
 discussion and review of individual studies the reader is
 referred to that document.
      The purpose of this section  is to provide an overview of the
 key issues of concern for the secondary standard:  relationships
 of various air quality indicators,  crop loss estimates,  averaging
 times and forest response to 03.  This section will summarize new
 analyses that attempt to address  these issues and discuss key new
 studies  that relate to the  issue  of averaging time(s).   The
 review of new data  in this  section  and in ECAO's Summary of
 Selected New Information on Effects of Ozone on Health and
 Vegetation will provide input to  staff conclusions and
 recommendations in  Chapter  XI.  For convenience of readers,
 citations for new research  reviewed in this  section are  included
 in as part of the  reference list at the  end of this staff paper.
 The CDS  provides a  more current reference list and discussion of
 03  welfare  studies  published  since  closure of  the  CD.
      Past research,  as  well  as current scientific opinion
 indicate  that the current air quality  standard is  not fully
 protective  of all vegetation  (Reich and Amundson,  1984;  CASAC,
 1986).  Yet  because of  uncertainties in the  data  base for crops,
 and the paucity  of  data  regarding forest  response  to O3/   the
 averaging time,  the level of  the  standard and  even  the need for  a
 separate  secondary  standard remain  difficult  issues to resolve.
      The  issue of averaging times and  form of  the  standard has
 proven to be  one  of the most  difficult issues  associated  with the
 review of the O3 secondary standard.  EPA's previous
recommendation  (CD)  that serious consideration be given to
setting both  a 1-hr and a longer-term secondary standard was
based on two  findings:
     •    That forest trees, because they are perennials,
          experience chronic 03 stress  due to exposure  to long-
          term O3 concentrations
     •    That the relationship between peak values and seasonal
          averages is generally not  predictable with any  degree
          of confidence; therefore the  1-hr standard may  not

-------
                              X- 48
           adequately reduce the probability of a high chronic
           exposure.
      While CASAC members (CASAC, 1987)  strongly endorsed the
 judgment that repeated peaks are critical in eliciting plant
 response,  the Committee's views regarding the need for a long-
 term standard were less clear.   Although some members expressed
 concern that long-term, low level 03  exposures could  adversely
 affect vegetation,  there was reluctance to suggest a  long-term
 standard in the range suggested by the  data (0.04-0.06 ppm)
 because of the obvious conflict with  background levels in many
 areas of the country.   The CASAC challenged EPA to identify  a
 single standard formulation that would  offer protection from both
 repeated peaks of  concern and Long-term exposures.  While monthly
 standard and was specifically mentioned for consideration, it was
 acknowledged that  this exposure statistic  would not be ideal in
 terms of fully capturing extreme situations or biological
 response.
      In  response to these CASAC comments several analyses were
 undertaken  and were presented at the  December,  1987 CASAC
 meeting.  While CASAC  found  the results of  the  new analyses
 interesting  and encouraged  EPA  to pursue the  analysis  of
 alternative  exposure  indicators;  with  Corvallis, they rejected
 EPA's  recommendation to  continue the  standard review until the
 analyses on  alternative  exposure  indicators is  complete.  Rather,
 they  endorsed  the idea of making a decision on  retaining  the 1-hr
 standard with  the information currently available.  There was
 some difference  of opinion among the committee members as to
whether the  current standard or a lower level would be adequate
to protect vegetation.  EPA has recently completed an initial
draft  (Lee et al.,  I988c) of the analysis of alternative exposure
 indicators requested by CASAC.  These results are considered
preliminary and will be undergoing peer-review.  In addition to
the analysis, a few new studies have been completed which look at
various aspects of the exposure dynamics issue.  The analysis and

-------
                              X- 49
 the new studies,  as well as CASAC's response to them  at the
 December 1988 meeting,  are discussed in Section E.3.

      1.   Air Quality Analyses
      Air quality  analyses and,  in particular,  relationships
 between alternative air quality indicators,  are critical in the
 selection of an averaging time  for the secondary standard.   While
 previous analyses have  indicated that relationships between peak
 and mean statistics are generally not stable or predictable with
 any degree of confidence (Johnson et al.,  1986; Lefohn,  1984;
 Heck et  al.,  1984a,  1984b;  Larsen and Heck,  1983)  none of the
 analyses specifically examined  the type of compromise indicator
 that was suggested by CASAC (CASAC,  1986).   An analysis  by
 McCurdy  (1987)  is the latest  effort to examine relationships
 between  various types of air  quality indicators.   These  results
 have been provided in separate  reports and summarized in Appendix
 A.
      For purposes of  this discussion,  it is  interesting  to  note
 that after examining  several  short-term peak,  multiple peak, and
 long-term average indicators, McCurdy found  that all  three
 exposure patterns of  interest are  highly correlated with two
 monthly  forms of  the  standard:  maximum monthly mean of the  1-hr
 daily maximums  and maximum  monthly mean of the 8-hr daily
 maximums  (SP, p.  A-22).  The first  indicator performed better
 than the  second,  but  because they are  highly correlated  with each
 other (r  =  .95),  either  one could be used.    It appears from  this
 analysis  that if  a single secondary 03 NAAQS which provides
 reasonable protection from  both repeated peaks  and long-term
 exposures of concern  is desired, it should be  the maximum monthly
mean of the daily maximum 1-hr averages (max monthly mean).  It
 should be noted that this exposure statistic is largely  based on
purely statistical concerns, as there  is little or no data for  a
one month exposure period.  The same is likely to be true for
other similar compromise exposure statistics.

-------
                             X- 50
     2.  Crop Loss Estimates
     As stated above, the max monthly mean appears to be  an
appropriate statistic in terms of the protection it affords  from
repeated peaks and long-term average concentrations of concern;
however, there is little or no effects data for a one month
exposure period.  This makes it difficult to arrive at crop  loss
estimates for a monthly indicator.
     In an effort to arrive at crop loss estimates for a  monthly
indicator, Larsen has analyzed new exposure indicators in his
existing crop loss model.  This model (Larsen and Heck, 1984) was
adapted from the Lar sen-Heck lognormal plant, injury model (Larsen
and Heck, 1976) to estimate the impact of O3 on crop yield.   This
model uses the "effective mean" 03 concentration (defined in the
next paragraph) to adjust for the greater effect of peak  O3
concentrations.  While calculation of the effective mean  is '
somewhat cumbersome for standard-setting, the Larsen-Heck crop
reduction model does allow one to look at other exposure
indicators and the crop loss estimates associated with them.
     The "effective mean" 03 concentration can be defined by
comparing it with the arithmetic mean.  Estimates of crop loss as
a function of an arithmetic mean 03 concentration essentially
assume that crop loss is proportional to the summation of all
(seasonal daytime)  1-hr average O3 concentrations.   Some studies
suggest (Larsen and Heck, 1984)  that peak 1-hr average 03
concentrations cause a much greater than proportional effect,
suggesting that effects may be proportional to a new dose
parameter called air pollution "impact," a parameter that is the
summation of all (seasonal daytime) 1-hr average 03
concentrations raised to the 2.66 power.  The "effective mean" is
the 2.66th root of impact divided by the number of hourly O3
concentrations that are exponentiated:
     me = [(Ech 2'66)/n]

-------
                              X- 51
 The arithmetic mean can be calculated by merely replacing 2.66
 with 1 in this equation, thus showing the parallel construction
 of the arithmetic and effective means.
      Larsen et al.  (1988)  discussed several crop reduction dose-
 response mathematical models and concluded that the lognormal,
 Weibull, and Box-Tidwell models are all closely related.  When
 fitted to data from the National Crop Loss Assessment Network
 (NCLAN), all three  models produce similar dose-response plots.
 The Larsen-Heck dose-response mathematical model was used to
 estimate percent crop reduction for soybeans:   the above
 "effective mean" O3  concentration was the "dose" and the
 lognormal model was  used to estimate the "response."  This model
 was used to estimate soybean percent crop reduction at each
 agricultural site in the National Aerometric Data Bank (NADB)  for
 each year of years  1981-1985.
      Larsen et al.  (1988)  calculated fourteen  O3 concentration
 parameters for each  site year of data.   The potential ambient
 standards that would  limit crop  reduction to 5,  10,  15 or  20
 percent -of agricultural  sites  are summarized in  Table X-8.  The
 three  of  these parameters  that correlate best  with crop reduction
 are  the  effective mean O3  concentration  '(I percent  of  the
 variance  unexplained), an  arithmetic  mean O3 concentration  (4
 percent  unexplained), and  the maximum l  month  mean of  daily
 maximum  1-hr  03 concentrations (15 percent unexplained).
 Potential  O3 standards can be selected to achieve a desired
 response.   For  instance, if no more than  15 percent crop
 reduction  were  desired at  any agricultural NADB  site,  a second
 highest daily maximum 1-hr 03 standard of 0.099 ppm could be
 used.  A maximum monthly mean standard (of l-hr daily maximum) of
 0.072 would be  required to achieve the same protection.  Table X-
 8 indicates that 65% of agricultural NADB sites achieved this
potential  standard in 1981-1985 as opposed to 40% of sites that
achieved the second high daily max standard of 0.099 ppm.

-------
Table X-8.  Potential Ambient Ozone Standards that Would Llimit Soybean Crop Reduction to 5,  10, 15, or 20 Percent
Ambient ozone standard that Percentage of agricultural NADB Percentage of variance
would limit soybean crop sites that achieved this potential unexplained by
reduction to stated percentage standard in years 1981-1985 regression of soybean
croo reduction attains*:
Ozone Parameter 5% 10% 15% 20% 5% 10% 15% 20%
Second highest daily maximum .059 .079 .099 .119 2 10 40 74
l-h ozone cone (ppm)
Maximum 1-month mean of daily .048 .060 .072 .085 5 26 65 91
maximum l-h ozone cone (ppm)
Maximum 1-month mean of daily .042 .052 .062 .073 5 26 64 89
maximum 8-h ozone cone (ppm)
Maximum 3-month mean of daily .034 .044 .054 .064 3 15 48 85
maximum 8-h ozone cone (ppm)
Number of days/yr that u-h 6 25 - - 17 61
daily max. cone exceeds
.08 ppm
Summer daytime arithmetic mean .035 .045 .055 .065 9 36 70 93
ozone cone (ppm)
Summer daytime effective mean .042 .052 .063 .073 14 45 77 94
ozone cone (ppm)
this ozone parameter
42
15
17
24
27
4
1
                                                                                                                                  ><.
                                                                                                                                  u

-------
                              X- 53
      As  mentioned earlier the "effective mean" 03  concentration
 adjusts  for the greater effect of peak O3 concentrations.
 Several  sites  may have identical  arithmetic mean values,  but the
 higher the  peak concentration at  one of these sites,  the  higher
 will  be  the effective mean value  and the expected  effects at that
 site.
      The exponent of  2.66 on  the  concentration term for the
 effective mean was derived from injury studies and then applied
 to yield studies without validating  its applicability to  yield
 (CD,  6-159).   Larsen  et al.  (1988) did an additional  analysis on
 this  issue  and concluded that crop reduction is probably  closely
 analogous to growth suppression,  leaf dry weight reduction,  and
 percent  leaf injury;  thus,  the average exponent of 2.66 for these
 three effects  was  assumed to  be a  good approximation  to use in
 this model  for crop reduction.  Also,  because  of the  low
 variability of 03  concentrations measured  at agricultural sites,
 estimated effects  would not change much if an  exponent  far  from
 2.66 was  used.
     The  approach  taken by Larsen  et  al.  (1988)  differs from that
 of Lee et al.  (I987a,b;  1988a,b,c).   Larsen's  estimated crop
 reduction was  calculated  from  a lognormal  model  that  expresses  O3
 exposure  as a  function  of the  effective mean times the  exposure
 duration, i.e.,  the total impact  (Larsen and Heck,  1985).   By
 definition,  the  estimated crop  loss is  totally determined by the
 effective mean once the exposed duration is fixed.   Since there
 is no biological variation in the data,  (as in Lee et al.,
 1987a,b;   1988a,b)  correlations between the estimated crop loss
and the exposure indices are, in fact, measures  of association
between the (transformed) effective mean and the other  exposure
indices.   Thus,  it is difficult to tell whether mean indices are
better correlated with plant response than other exposure
indicators based on Larsen et al (1988).

-------
                              X-54
      3.  Averaging Times
      a.  NCLAN/CERL Reanalysis
      The discrepancy between the seasonal mean exposure indicator
 used in the NCLAN studies and the repeated peak exposures
 identified in the CD as being most important for plant response
 make it difficult to evaluate the vegetation effects data base
 for 03.   Accordingly,  OAQPS requested that the 03  team at
 Corvallis Environmental Research Laboratory (CERL) conduct a more
 detailed analysis of the crop yield response data  from the
 National Crop Loss Assessment Network (NCLAN)  and  from the
 Corvallis laboratory.
      Lee et al.,  (1987a,b)  reanalyzed several  NCLAN data sets
 (soybean,  wheat,  cotton,  and alfalfa)  in an effort to determine
 which exposure index produces the most accurate exposure response
 relationship.   Although the most widely used exposure index to
 assess  O3  effects  on crops  is the 7-hr seasonal  mean,  current
 information about  agricultural  crops  suggests  that short-term,
 high  concentration exposures appear to be  more detrimental  to
 crop  production than  long-term,  low concentration  exposures.   The
 seasonal mean  ignores many  factors known to  affect vegetation
 growth  including phenological response, peak concentrations,
 length of  episodes  and  days  between peaks  (CD,  1986).
      Lee et  al. (1987a,b) examined several exposure  statistics
 that  emphasize  peak concentrations more than lower  concentrations
 as alternatives to the  seasonal mean.  These include:   1)
 summation of all concentrations above  a cut-off level,  2) number
 of hourly concentrations above a  cutoff level, 3) summation of
 all concentrations raised to a power greater than 1.  in
 conjunction with 3), an exponential weighting scheme over time
was used to account for differences in response at various stages
 of phenological development; this is described as phenologically
weighted cumulative impact statistics  (PWCI).  Exposure indices
that included all the data performed (24 hr) performed better
than the indices that used only 7 hours of data.  The 7-hr
seasonal mean was never "best" and was near optimal in only 5 of

-------
                               X-55
  14 cases.  The exposure indices that emphasized peaks performed
  better than those that gave equal weighting to all
  concentrations; indices that accumulated the exposures performed
  better than those that averaged the exposures.
       Because of the extreme importance of this type of analysis
  in resolving the issue of averaging time, and form of the
  standard for the secondary standard,  Corvallis has conducted
  increasingly more detailed assessments  of alternative exposure
  statistics using various  weighting schemes for relating exposure
  NCi^T/eSP°nSe'   ^ a  m°re  eXtensive ^respective analysis of
  NCLAN data,  Lee  et  al.,  (.I988a,b)  fit 24 common and  589  general
  phenologically weight  cumulative impact (GPWCI) functions to the
  response  data from  seven crop  studies.   The criteria  established
  for determining  "best" exposure indices were those that  displayed
  the smallest residual sums of  square error when the yield
  response  data were regressed on the various 03 exposure indices
  using the Box-Tidwell model.
      The  "best" exposure index was a GPWCI with sigmoid
 weighting.  Cumulative indices (with concentration thresholds)
 performed as well as the GPWCIs while mean indices did not
 perform as well.   The authors concluded  that,
      "While no single index was deemed "best"  in relating o,
      exposure to  plant  response,  the top performing indices were
      those indices that (1)  cumulate the hourly O3  concentrations
      over  time,  (2)  emphasize concentrations of 0.06  ppm and
      higher either by continuous sigmoid weights or by discrete
      0-1 weights  of  the threshold indices and 3) phenologically
     weight  the exposure such that  greatest weight  occurs  during
     the plant growth stage.  When  assessing the impact of o, on
     Plant growth, these findings illustrate the importance of
     repeated peaks,  and the time of increased sensitivity in
     assessing the impact of O3  on plant  growth."
Although peak concentrations should be given greater weight, the
authors suggested that lower concentrations should also L
included in the calculation of an exposure index

-------
                              X- 56
      Tingey et al. (1988) is essentially a condensation of the
 paper by Lee et al. (I988a,b)  and therefore the conclusions are
 basically the same.  However,  the paper does show the importance
 of exposure duration and the limitation of the seasonal mean to
 specifically incorporate duration.   For example,,  the mean does
 not distinguish among exposures to the same concentration (mean)
 but of different durations (e.g.,  10,  50,  or 100  days).
      The most recent analyses  by Lee et al.  (I988c),  developed in
 response to a request by CASAC (December,  1987),  evaluates
 selected ambient O3 air  quality indicators  and  estimates  the
 exposure levels associated with agricultural losses.   The results
 of this initial draft analysis are  discussed in detail in Lee et
 al.  (1988c).   This section will highlight key points  regarding
 the  methodology and the  results.   Ideally,  the  selection  of the
 most appropriate exposure indices would be  derived from
 scientific  principles  and be validated  by experimental data
 specifically  designed  to identify the most  appropriate exposure
 indices.  Such an approach, however, would delay resolving the
 selection of  exposure  indices  for years  until the necessary
 experiments were  designed and  conducted.  Since the likelihood of
 the  Agency  funding  another major crops research program in the
 near  future is  small,  EPA has  developed  an alternative approach
 of conducting  a retrospective  analysis of existing plant-response
 data.  While recognizing  the limitations of the original NCLAN
 data, CASAC has requested further analysis of the exposure
 dynamics issue through retrospective analyses.  The major
 limitation of such analyses is that the specific studies were  not
 designed.for developing exposure indices or testing exposure
 hypotheses; consequently, the usability of the data is somewhat
 limited.
     Plant yield and hourly O3  monitoring data were obtained for
twelve NCLAN field experiments  (soybean, winter wheat, corn,
 sorghum and cotton).  The twelve NCLAN studies represented five
species with differing cultivars for a total of 17 individual
cases.  Crops were grown according to standard agricultural

-------
                               X-57
  practices and exposed to a range of concentration in open-top
  chambers according to NCLAN protocols.  These crop yield data
 ml
 model  (defined in Appendix A of Lee et al.,  I988b)
      Selection criteria for determining the  "best" exposure  index
 was based on the minimum residual sum of squares  (Rss, for each
 of the 17 cases.  Ranking and selection of exposure indices was
 performed separately for each case.  For each individual case
 the relative performance of an index was measured as the ratio of
 its RSS to the minimum Rss.   For overall oomparlson
 and cultivars,  the exposure indices were evaluated according to
 three criteria:   (l,  the average score calculated as the mean
 relative RSS.s;  and  (3,  percent variation  (i.e.,  range/mean score

      Based  on the  criterion of  minimum RSS,  no single  exposure
 index performed  "best" for all  seventeen cases.   The best  overall
 fits  are  obtained  using  indices that:
      1.   cumulate hourly concentrations overtime,
     2.   place greater weight on concentrations of 0.06 ppm or
          higher either by continuous slgmoid weight or by
          discrete o-l weights, and
     3.    weight hourly concentrations according to the
          phenological stage of plant development.
     The results indicate that while the generalized phenological
cumulative impact (GPWCI, indices best related plant respons! to

             the
             there
                  Uded 3 •1»»"— ***- integrated index
as an at      "
indices th,      9U3llt     nar'      c™ul"-e censored
indices that integrated concentrations of 0.06 (or 0.07, ppm or
higher (SUM06 and SUM07,  also performed well, suggesting that
ambient 03 levels below O.OS ppm are  important in triggering
  a"
       b      r                                   Lefohn et al.
      .b)  and Lee et al.  (I987a,b;  I988b,  who used NCLAK data and

-------
                              X- 58
 cumulation indices with sigmoid and allometric weights in
 demonstrating the importance of peak concentrations in
 determining plant response.
      The integrated exposure indices (SUMO6 and SUM07)  used to
 characterize experimental  exposures are  functions of exposure
 duration and concentration.   Correspondingly the exposure levels
 associated with various yield losses calculated from experimental
 data  are functionally  related to  exposure "seasons."  An implicit
 assumption here is that exposures of varying duration,  with equal
 values  of SUM06 (or SUM07) cause  the same biological response.
 According to Lee et al.  (1988c) experiments replicated in time
 and/or  space,  differing in exposure duration,  should produce
 identical predicted relative  yield  losses using SUM06 or SUM07.
 Therefore,  the discrepancy between  exposure "seasons" and between
 experimental and ambient averaging  times are accounted  for in the
 integrated indices.  Thus, these  results indicate  the integrated
 indices  capture the key components  of exposure,  are  adequate
 descriptors  of plant response, and  are simple  and  easy  to
 implement from a regulatory perspective.
     The  magnitude  of  the O3-induced yield  losses varied  among
 sites, years  and species/cultivars.   The NCLAN  studies  used  in
 this analysis  represent both 0;! sensitive and tolerant
 species/cultivars,  and multiple years, and  experimental  sites.
 Crop yield  losses  for  16 NCLAN studies are  contained  in  Table
 3.3.  The predicted  relative yield  losses for the four  exposure
 indices across  the  16 cases are shown for the mean, the  25th,
 50th,  75th, and  85th percentiles.
     As can be  seen  in Table x~9, median crop losses  of  18.9% are
 expected  in agricultural sites experiencing 03 exposures with a
HDM2 value of  0.12 ppm.  Lowering the concentration to  0.10 ppm
would result in an estimated median crop loss of 12.8%, whereas
going to  0.08 ppm would result in an estimated median crop loss
of 7.9%.  Table X-10 presents the exposure levels associated with
mean and percentiles of predicted relative yield loss of  5 to 30%

-------
                           TABLE X-9
    rPRvo       a?d mean Predicted relative yield losses
    (PRYLs) associated with various levels  of the four
    ras
      HDM2
                        Percentiles nfL
50th
1.7%
4.2%
7.9%
12.8%
18.9%
32.6%
48.7%
MEAN
6.3%
9.9%
13.7%
17.8%
22.3%
33.0%
48.6%
- 75th
9.8%
15.9%
22.1%
28.8%
34.8%
46.0%
58.3%
                                                  15.0%
                                                  21.7%
                                                  28.0%
                                                  33.8%
                                                  39.7%.
                                                  50.5%
                                                  60.4%
Source:  Lee et al.
.(1988c)

-------
                            "ABLE X-10.
                                                              I
Exposure levels  associated with predicted relative yield losses
(PRYLs) of  5 to  30%  for the four exposure indices, HDM2, M7,
SUM06, and  SUM07,  for the 16  NCLAN studies.  Separate
calculations are performed using the 50th, 75th, and 85th
percentiles, and the mean PRYLs for the 16 NCLAN studies.

A)  Exposure levels  with Mean PRYLs of 5 to 30%.

                      Mean Predicted Relative Yield Loss

HDM2 (ppm)
M7 (ppm)
SUMO 6 (ppm)
SUM07 rDDnrt
5%.
0.032
0.027
9.7
7.5
10%
0.061
0.045
17.1
14.4
15%
0.087
0.059
23.7
20.9
20%
0.110
0.070
29.8
26.8
25%
0.131
0.080
34.3
32.3
30%
0.149
0.089
38.8
37.2
B)  Exposure levels with  SOtli Percentile PRYLs of 5 to 30%,

                50th

HDM2 (ppm)
M7 (ppm)
SUMO 6 (ppm)
SUM07 ronun
_5%
0.065
0.042
14.1
11.6
10%
0.089
0.057
21.3
17.9
15%
0.107
0.068
28.0
- 24.3
20%
0.123
0.077
33.1
30.5
25%
0.137
0.085
37.5
34.9
»-t j-n-ijj
30%
0.152
0.093
42.2
39.2
C)  Exposure levels with 75th  Percentile PRYLs of 5 to 30%.
75th Percentile Predi

HDM2 (ppm)
M7 (ppm)
SUMO 6 (ppm)
SUMO 7 fDDirO
5%
0.023
0.028
10.2
6.9
10%
0.041
0.039
16.6
13.0
15%
0.057
0.048
22.4
18.5
cted, Relative Yield To-**:
20%
0.073
0.056
27.6
23.7
25%
0.089
0.066
31.3
28.8
30%
0.104
0.075
33.7
32.6
D)  Exposure levels with 85th Percentile  PRYLs  of  5  to 30%.

                85th Percent!lg Predicted Relative Yield Loss
	53	10%	15%	20%	25%	30%
HDM2 (ppm) <0.020    0.026    0.040     0.055     0.070    0.087
M7 (ppm)    0.026    0.035    0.045     0.054     0.062    0.069
STJM06 (ppm)   3.7      7.5     10.9     13.8     16.7     19 6
SUMQ7 fppm)	2^2	S^J	8.7     11.7     14.7     17.R


Source:  Lee et al.  (1988c)

-------
                                X- 61
  for the four indicators across all sites,   if  EPA 
-------
                               X- 62
       In  regard  to  the  exposure  period  of  interest,  an air quality
  analysis of  83  non-urban  site-years of ambient O3 data  indicate
  an averaging time  of five months from  May to September  (153  days)
  should be used  to  capture 90% of the hourly concentrations of
  0.06 ppm or  higher.  This period includes the EPA-defined O3
  season for most sites, represents the  period when the second
  highest  daily maximum  1-hr concentration generally occurs, and
  coincides with the agricultural season for the majority of crops
  grown in the United States.

      b.   New Studies
      Although only a limited number of studies have been
 conducted with the specific  objective of  developing or evaluating
 various  exposure indices,  several studies have attempted to use
 existing exposure response data  for evaluating a  range of
 exposure indices.   The  results of these retrospective analyses
 have  provided useful concepts and the  conclusions are in general
 agreement. However,  because  the studies  were  not specifically
 designed  to evaluate various  indices,  the  differences in and
 among exposure treatments  may be relatively small.   As a result,
 the power of  these  studies is Less  than desirable.
      In a small  retrospective analysis, Lefohn et al.  (I988a)  has
 come  to similar  conclusions as Lee  et al.  (I987a,b) using
 different NCLAN  data sets  with the Weibull and linear models.  As
 in Lee et al.  (I987a,b), he is comparing the use  of several
 indices of exposure  in  describing the relationship between O3 and
 reduction in  agricultural  crop yield.   Prior to these studies, no
 attempt had been made to determine which exposure response models
 best fit  the  data sets  examined.   Lefohn,  et al.  (I988a) used
 hourly mean concentration data based on 2-3 measurements per hour
 to develop indices of exposure from soybean and winter wheat
 studies conducted in open-top chambers at the Boyce Thompson
 field site.  The efficiency of using seasonal mean and cumulative
 indices (i.e., number of occurrences equal to or above specific
hourly mean concentrations, sums  of all hourly mean

-------
                               X-63
  concentrations equal to or above a selected level, and the
  weighted sum of all hourly mean concentrations,  to describe the
  relationship between exposure to 03 and reductions in the yield
  of agricultural crops has been evaluated.   In most cases  the
  Weibull model,  a functional form used extensively by the NCLAN
  program,  tends  to fit the data based on cumulative indices better
  than on seasonal mean indices.   This conclusion  is similar to
  what Lee  et  al.  (1987a,b)  found with different models and data
  sets,   in addition,  two  other recent papers  (Hogsett  et  al.,  1988
  and  Musselman et al.,  1988) support the  same conclusion  that  1)
  mean indices are  not  among  the  best exposure indicators  and 2)
  the  preferred (yield  best statistical fit to the data, exposure
  indices cumulate  the  exposure impact over the growing season and
 preferentially weight the peak concentrations.  While the
 exposure  indicators developed in these analyses can be
 potentially complex, Hogsett et al.   (1988)  has recently found
 that there are other simple indices such as summation of all
 concentrations above 0.06 (SUM06, and 0.08  ppm (STO08) which are
 almost as good as the complex PWd statistics.  These potential
 exposure statistics were  evaiuated in greater depth in a  new
 analysis (Lee et al.,  I988b)  previously  discussed on p. x-48

 abou/th ^7 ^ Lef°hn  "  a1'  (1988a>  haS ^-ated  Discussion
 about the  interpretation  of  the  data (Runeckles,  1988;  Parry and
 Day,  1,8.,.   Both commenters felt that the data presented was
 insufficient  to  support the conclusion that peak-weight exposure
                   did point out that peak-weighted indices
           t       as wel1 as mean indioes-  The «—•*«•
           the compilation of two years of wheat data with
markedly different exposure durations into a single model.  i»
response, Lefohn et al.  (lM8b,  stated that the wheat data
support the need to include a cumulative component in an exposure
       "       "
          that
<.,    .   _,                      *«-=* ...» more relevant to use in
the standard setting process than seasonal means,  which ignore
the length of the exposure period."

-------
                             X- 64
     These conclusions on the importance of cumulative peak
concentrations in causing plant response are consistent with  the
data presented in the CD as well as other recent studies.  Wang
et al.  (1986) characterized 03 exposure of three tree species as
the number of daily occurrences above 0.08 ppm and 0.12 ppm over
a four month growing season.  The authors concluded that O3
significantly impaired the growth of hybrid poplar in the absence
of visible- injury.  In a three year study with quaking aspen,
Wang et al.  (1986) found that plant growth was reduced 12 to  24%
although the current national air quality standard was exceeded
in only one of those years.  Adomait et al. (1987) characterized
O3 exposure of white bean as the cumulative 03  concentration
above a threshold of 0.08 ppm for a one month period.  In
addition, Reich and Amundson (1986) stated that the O3 induced
decrease in growth of several tree species was directly related
to reduced photosynthesis, which was impaired by the cumulative
O3 dose.
     Although the CD and the weight of the new evidence since the
CD review (with the exception of papers reporting results from
NCLAN)  seem to suggest the importance of peaks, another issue of
concern is how to treat the peaks in the development of exposure
response indicators.  While Lefohn et al.  (1988a)  recommended
that a weighting scheme which gives greater weight to higher
concentrations be used in developing exposure indicators,  others
support the removal of some dose from each hourly 03  value so
that low levels which are less likely to effect plant growth are
eliminated (McCool et al., 1986).   Basically,  three weighting
approaches have been used:  1)  a concentration threshold in which
the concentrations,  number of occurrences or hours above the
threshold cumulated; 2)  an exponential weighting in which all
concentrations are raised to a specific exponential power and 3)
a sigmoid weighting in which all concentrations are weighted with
a multiplicative weighting factor (which depends on
concentration).   The functional  weighting approach using either
exponential or sigmoid weighting seems preferable to the

-------
                               X- 65
  threshold approach because it considers the contribution  of  all
  concentrations in eliciting a response.
       While several studies (Lee et al., I987a,b and 1988b;
  Hogsett et al., 1988; Lefohn et al., I988a; Musselman et  al.
  1988) have focused on developing new exposure statistics, there
  have been several attempts to evaluate the statistic used most
  often in the current 03 data  base - the 7-hr seasonal mean.
  According to smith et al.  (1987)  the open.top cnMber method
  which predictive models such  as  NCLAN are based do not reflect
  the episodic nature  of O3  pollution or  some  of the more  important
  environmental conditions that influence plant  growth.  The
  authors  concluded  that the failure  of the  NCLAN predictive model
  to  account for such conditions might explain why they  found no
  effect of ambient 03 on  yield of field grown soybeans  in New
  Jersey while  the NCLAN model predicted  losses  of around 19%
      Heagle et al. (1986,  1987) examined the seasonal mean
  statistic in  studies on  soybean and tobacco.   Both  studies added
 03 in constant or in variable amounts which were proportiona! to
 ambient 03 concentrations,   mterestingly enough, while there
 were greater fluctuations and higher peak concentrations with the
 proportional method than with  the constant addition method  the
 two methods of addition gave similar seasonal mean
 concentrations.  Regression of yield response on 0,
 concentrations showed  no signlficant differences between  the  two
 fc                                                     e  * the
 fact  that  the  7-hr  seasonal  mean,  even  with the proportional
 addition,  may  fail  to  accurately reflect the elevated  exposures.

                *""
                                              ***" "»
seted                                                «
selected is because it was less sensitive to variations  in 03
patterns (cure, 1986) .  Therefore, the conclusion in these
studies that differences in peak 03 concentrations do not
     Rawlings et al. (1988) conducted additional analyses of the
soybean (Heagle et al.  a986,  and tobacco (Heagle et al. ,  1987)

-------
                             X- 66

                                              and
  data suggested that the
  contest, the 7
                               and that
 limited.             «*«-»«, the -best"  exposure index is

                              <1985)  COmP«ed various
                             rr •  The
additiona! $2.225 billion

     The information reviewed  in this section suggests a
^w  •                        exposure that the plant

                                              -
    4.  Forest Risk Assessment

-------
                               X- 67
  for 03.  While the potential contribution of O3 and other
  atmospheric pollutants has been receiving increasing scientific
  and regulatory interest, the numerous and complex interactions
  between 03,  environmental factors, and health and productivity of
  forests malce it difficult to distinguish 03  effects from natural
  forest processes (Peterson and Violette, 1986).  Although the
  nature and magnitude of the effects and the  precise contribution
  of various atmospheric pollutants is not well understood  there
  is evidence  that some sensitive forest types are negatively
  affected by  O3.   The  effects data,  summarized in Chapter X of
  this staff paper,  is  somewhat  scarce at this point in time and
  tends  to focus more on controlled  studies of seedlings  or
  saplings.  The results of these controlled laboratory or
  greenhouse studies are not easily  extrapolated  to field
  conditions, thus making current extrapolation (e.g., seedlings to
 mature trees, stands  to populations) difficult.
      a.  Overview of  Forest Risk Assessment
      While significant new research is underway to improve our
 knowledge of the exposure-response relationships of various tree
 species and 03  and to develop our extrapolation  modeling
 capabilities,  at this point in the standard review we must assess
 a data  base that has  considerable uncertainty associated with it
 Standard statistical  procedures cannot quantify  this type of
 uncertainty because of the absence of valid statistical  saxnples
 across  the relevant populations.   This type of uncertainty,
 however,  can  be represented using  appropriate decision analytic
 procedures to elicit judgmental probabilities (Walsten and
 Whitfield,  1986;  Peterson and Violette,  1986).   EPA decided  to
 use this  approach as a tool to better quantify the range of
 scientific  judgment and uncertainties regarding 03-caused risks
to forest resources.
     This section win  briefly describe the methods and
prenminary results of . study whioh ^^^ ^ ^^
around estates of forest-tree growth decrement and foliar
       due to 03  exposures.   This information  is  meant to

-------
                              X- 68
 supplement scientific information and data by presenting an
 interpretation of current scientific knowledge regarding O3
 caused forest effects.   It is extremely important to recognize
 that the results presented here are not a substitute for
 continued scientific research;  clearly such research is needed
 and the results of this work confirm that fact.   However,  because
 the magnitude of forest damage  caused by O3 is likely to remain
 uncertain in the near term,  policy  analysts and  regulators may
 find information regarding the  uncertainty among forest experts
 useful in formulating policy options and better  focusing research
 efforts.
      b.   Protocol Summary
      For  judgmental  probabilities regarding forest  damage  to be
 useful to policy makers  they must satisfy three  criteria
 (Peterson and Sueker, 1987):
      1) Judgments should be  made over  the appropriate forest
 response  parameters  for  policy  assessment - in this  case the
 endpoints of  concern are growth  decrement and foliar injury,
      2) Judgments should be  obtained from experts who are  likely
 to possess interpretations of O3 associated forest effects that
 span  the  range  of respected  opinion, and
      3) Judgments should be  relatively stable over the  course  of
      the  study.
      To satisfy  these criteria,   a protocol  for probability
 encoding  was  developed by the investigators and reviewed by  EPA.
 While  detailed  information on the protocol  is contained  in the
 report  (Petersen  et  al.,  1987) it is useful to summarize several
 of the basic  assumptions made in the protocol:
      1) Total above  ground growth decrement and foliar injury
were selected as  response parameters.  Several other response
parameters were considered, but rejected as being beyond the
 scope of this study.
     2) While numerous O3-exposure measures were  reviewed,  two
were selected: the maximum 3-month mean for 24- and 7-hour (0900-
 1600) daily averages.  The selection of exposure statistics for

-------
                               X- 69
 this analysis was extremely difficult.  While new data  indicate
 the importance of repeated peaks  in eliciting plant response  (at
 least for crops), several experts expressed concern about their
 ability to reiate their judgments to a repeated peak statistic
 not commonly .used in the effects literature.  Rather, they seemed
 more comfortable with a mean statistic, which was defined for
 purposes of the study as including the episodic as well as the
 low-level exposures seen by the trees.   These very real concerns
 strongly influenced the selection of the exposure statistics for
 this analysis.
      3)  Trees in each O3 treatment "population" were assumed to
 be exposed  to 03 exposures represented  by the 0.035, 0.055 and
 0.085 03 maximum 3-month means  for the  24- and 7-hour daily
 averages.  The concentrations  were chosen to express the range of
 °3 exposures over which growth  decrement or foliar injury could
 be detected in o.s. forests.  Trees in  the "control" population
 were assumed to be exposed to concentrations ranging between
 0.020 and 0.030 ppm 03 to represent naturally occurring
 background conditions.
     4)   The following assumptions were made about the conditions
of the forests in the hypothetical studies:
     a.    The forests were mixed/monoculture type  and of
          uneven/even aged composition.
     b.    Forest  trees were exposed to one of  the  03  exposures
          from germination  to the  time growth measurements were
          taken.
     c.    Ozone sensitive genotypes were distributed  in forests
          according to  the experts' beliefs.
    d.   Except in the use of plantation trees, trees were
         entirely wild type genetically.
    e.   Site conditions, including slope, aspect, elevation,
         soil type, air temperature and humidity,  and pest and
         pathogen populations were distributed across the
         control and experimental populations as the experts
         believed they were in  the real  world.

-------
                              X- 70
      f.   Growth effects were measured on all trees  (no sampling
           problems) at all locations.  Measurements were assumed
           to reveal effects on all age classes within a true
           population.
      g.   Effects concerned only the species in question and not
           second-order effects involving other species.

      c.  Scientific -Experts
      In order for the judgments to be useful to decision makers,
 they must be obtained from knowledgeable experts in the area of
 03  effects on forests.   The experts were selected to provide
 judgments regarding O3  associated forestry  effects spanning the
 range of respected opinion.  Establishing the range of opinion
 and selecting the experts to be encoded may appear to be very
 subjective;  however,  the field was narrowed considerably by
 focusing on those individuals whose careers had been devoted to
 conducting air pollution research (with a focus on 03)  in the
 area of forest effects.   Recent reviews of  03 associated forest
.effects and  references  to the published literature also
 facilitated  the selection of  experts.   Several  offices  within  EPA
 (OAQPS,  OPPE,  ORD) 'collaborated, to develop  a set  of potential
 candidates from which the five  experts  were  selected.   Names of
 participating experts are provided in Table  X-ll  to facilitate
 the  interpretation  of the range  of responses by other forest
 scientist  and policy makers.  All  experts selected by EPA agreed
 to participate  with the understanding that their  judgments  would
 remain  anonymous.  Particular responses or judgments for experts
 A through  E  are identified  only by randomly  assigned code
 letters.
     d.  Credible Judgments
     To ensure  that an expert's judgments were relatively stable
 and accurately  represented his uncertainty,   a number of steps
 were followed.   First, a draft of  the protocol was developed and
 reviewed by the  staff at EPA and the consultants prior to the
 encoding session.  Following the revision of the protocol,  two

-------
                            X- 71
                            Table X-ll

                      Forest Response Experts
     Expert

     Sam Linzon

     Samuel McLaughlin

     Paul Miller

     John Skelly

     David Tingey
             _.-            	
Ontario Ministry of Environment

Oak Ridge National Laboratory

USDA Forest Service

Pennsylvania State University

U.S. EPA,  Corvallis Environmental
Research Laboratory
Souce:   Peterson et al., 1987

-------
                              X- 72
 separate meetings to encode the judgmental exposure-response
 relationships were held with each expert.  The first meetings
 were conducted during November-December 1986; the second during
 February-March 1987.  During the first meeting,  which lasted 6 to
 8 hours,  experts reviewed the protocol,  discussed relevant
 literature and research into factors determining exposure-
 response functions and estimated probability distributions for
 selected tree species.   At the second meeting, the results of the
 first session,  which had been forwarded  to the experts by mail,
 were carefully reviewed and an additional probability encoding'
 session was completed.   Draft final  results were distributed to
 the experts and discussed in a telephone conference with each
 expert.
      e.   Results
      The  findings  suggest a wide  range of opinion regarding tree
 sensitivity to  03  among  credible  forest  scientists.  Three
 different sets  of  results were developed as  a result of  the
 analysis.   The  first set,  discussed  in sections  3.1  through 3.6
 of  the report,  present  152  cumulative  probability distributions
 for growth  decrement estimated by Experts A  through  E.   These
 judgments dimension  the  uncertainty  over  growth decrement in  19
 different tree  species due  to  six different  03 exposures.
 Cumulative  probability distributions across  tree  species and
 experts reflect  different underlying probability models.  Some
 cumulative  probability distributions are virtually linear,  others
 reflect the normal cumulative probability model,  and others a
 "hockey stick" distribution.  This variability prevented fitting
 a single probability model to the cumulative probability
 distributions developed  in this study.  For purposes of
 description, individual  judgments within any given distribution
were linked using cubic  spline interpolation.
     Figure X-3 summarizes these judgements across experts for
eastern white pine, deemed to be the most sensitive species by
Experts A, B, C and E at all 03 exposures.  The vertical axes in
the figure represent percent growth decrement relevant to

-------
                     . fc
o
c
1-1
o
n

rt
n
<-i
01
o
3
10 —
f) f*)
•6-8
s!
S ^
c °
I"
c 3
• S
ft £.
5 |
V 3
| I 1 h
1
**
p
u •
in j
fo <
?] ^

                             PERCENT GROWTH DECREMENT

                        *   S  3  8  S      -  s   *
            9-
  s
  «* §
  p
  b
 o
 I
r>
 O
 a.
 O    ^
b    ^
tn    ~
%
A

»O



O
                                                             S  8   S  g  g  g
                                                                                    OQ
                                                                                    C
                                                                                    1
                                                                                    
-------
                              X- 74
 background conditions.   The horizontal axes represent the six
 different O3  exposure levels.   The vertical barss  at  each 03
 concentration represent each expert's  maximum credible confidence
 interval,  assumed to be 98%,  for  growth decrement.   The symbols
 within  each bar indicate respectively  the  .01,  ,,25,  .50, .75,  and
 .99  probability fractiles.   Thus, in Figure X-3,,  the maximum
 credible  interval for Expert C's  judgment  of growth  decline  at
 exposures  characterized by  the  0.035 ppm O3  7-hour average
 statistic  ranges  between a  lower  bound (F=0.01) growth decrement
 of 0% to an upper bound (F=0.99)  growth decrement of 15%.  The
 three intermediate symbols  represent growth decrement at
 cumulative probability  fractiles  .25,  .50,  and  .75.   Figure  X-3
 illustrates a  number of important points:
     •     First,  there  is a wide  range of uncertainty expressed
           in the  experts' judgments.   A and  E are consistently
           the  most confident, c the least.
     •     Second,  uncertainty tends to  increase as O3 exposures
           increase.
     •     Third,  there  is no agreement  across experts that growth
           decrement  due to O3 will occur (within the maximum
           credible interval of 98%)  until O3 exposures associated
           with the 0.085 ppm 7 hour average  statistic are
           reached.
     •     Fourth,  2  of  4 experts believe that there will be
           little,  if any, growth decrement at the 0.035  7-hr
           average.
     The second set  of results presents exposure-response
functions  for O3 caused  growth decrements over the range of 03
exposures  considered in the study.  The experts were unanimous in
their belief that exposures of 0.035 ppm 03 7-hr average would
not cause measurable growth decrement in the tree species
considered.  These exposure-response functions are summarized in
Figures 3.15 through 3.30 of the report.  Figures 3.31 and 3.32
 (Peterson et al., 1987)  presents upper  and lower bound judgments
across experts, providing the largest possible credible interval

-------
                               X- 75
  and the broadest interpretation of uncertainty.   As can be seen
  for eastern white pine,  this interval  is often quite large.   For
  example,  extreme risk estimates range  from 0  to  65% growth
  decrement at 0.055 ppm 03, 24-hr average.
      The  third set of results focus on the judgements  on foliar
  injury  to forest  trees,  Although  the  experts  indicated
  considerable uncertainty with respect  to growth and yield
 effects,  their lack of confidence  in risk  estimates for  foliar
 injury were even more pronounced.
      Peterson et al.   (1987) was designed to explicitly present
 the scientific uncertainty in estimates of 03  induced forest
 decline.  it is intended to augment scientific research by
 reframing effects data so that non-scientists  can better evaluate

               X
 does       h                                         ••     *
 does  not,  however,  replace effects  research as it did  not
 generate any  new  hypotheses or data.  What  it  does is  confirm
 what  many  felt was  the case - that  there .is a  wide range  of
                           f°rest
                                             regarding the  role
      and forestry effects.  Further scientific research is
needed before confident estimates of growth decrement can be
attributed to o3.  Much of this research will be conducted during
the next several to five years under NAPAP.s Forest Response
Program and EPA.s research program,  the Effects of TropLpheric
Ozone on Forest Types.                                F"*pneric

-------

-------
                              XI- 1
  XI.  Staff Conriln.
       Secondary .Standard
       Drawing upon the evaluation of scientific information
  contained in the CD and CDS, this section provides preliminary
  staff conclusions that should be considered by the Administrator
  in selecting the pollutant indicator,  form, averaging time and
  level of the O3  secondary standard.  The primary focus of this
  discussion will  be  vegetation and ecosystem effects,  as this data
  base  provides the best support for the secondary standard for o,
  Materials damage and  effects on personal comfort and  well-being
  will  be  covered  in  the final section.   Risk and  benefit analyses
  coverzng vegetation effects  are currently underway as part of a
  continuing program  to  assess secondary standard  effects.

      A.   Pollutant  Indicator
           On February  8,  1979,  the chemical designation  of the  0
 primary and secondary  standards was changed from photochemical
 oxidants to 03 (44 FR 8202).   EPA changed the designation of the  -
 standard to 03 since the Federal Reference Method  (FRM) for
 determining compliance specifically measured 03 as a surrogate
 for total oxidants,  and because a substantial vegetation effects
 research base has established 03 as being chiefly responsible for
 the adverse effects  of photochemical air pollutants, largely
 because of its relative abundance compared to other photochemical
 oxidants.
     The  toxicities  of the peroxyaoetyl nitrates,  of hydrogen
 peroxide  (H202,, and of formic acid are less well documented than
 the  toxioity  of 03.  These oxidants have  been the focus  of
 considerably  less  research because  the  levels at  which they occur
 in the ambient  air,  even  in urban areas,  appear to  warrant  much
 less concern  (CD,  p. 7-1).
     Peroxyacetyl  nitrate, perhaps the best studied  of the non-0,
oxxdants,  is produced by photochemical reactions similar to those
that produce o3.  controlled exposures of plants to PAN have
indicated that PAN can cause foliar injury and yield losses in

-------
                             XI-  2
 sensitive  cultivars  of  leafy vegetable  crops  (p.,  X-22) .   However,
 a  comparison  of  PAN  concentrations  likely  to  cause  either visible
 injury  or  reduced yield with measured ambient concentrations (CD,
 Chapter 5)  indicates that  it is  unlikely
 that PAN related effects occur in plants in the U.S.  except in
 some areas of California and possibly in a few other  localized
 areas.   Because  phytotoxic concentrations  of  PAN  and  the  other
 photochemical oxidants  generally occur  at  significantly lower
 ambient concentrations  and are less widely distributed than O3,
 the focus  of  this standard review will  be  the effects of  03.
     The question of whether 03 can serve as  an abatement
 surrogate  for controlling other  photochemical oxidants is
 addressed  by  Altshuller (1983) who concluded  that "the ambient
 air measurements indicate that O3 may serve directionally but
 cannot  be  expected to serve  quantitatively as a surrogate  for  the
 other products."  This  conclusion-appears  to  apply to the  subset
 of photochemical products of. concern -  03,  PAN, PPN,  and H2O2 -
 identified  in the CD, even though Altshuller  (1983)  examined the
 use of  O3 as an abatement surrogate for all photochemical
 products.   Lack of a quantitative,  monotonic  relationship between
 03 and other photochemical oxidants is discussed in Chapter 5 of
 the CD  in which average PAN/03 ratios for different sites and
 years vary  from 9 to 3.  In  addition, no single measurement
 methodology can quantitatively and reliably measure O3 and other
 photochemical oxidants, either individually or in ambient air
mixtures.
     In  spite of the above limitations,  it is generally
 recognized that control of ambient 03 levels currently provides
the best means of controlling photochemical oxidants  of potential
welfare concern  (03/  PAN,  PPN, and  H202) .   This recognition along
with a  controiled-exposure, welfare data base which implicates
 only O3 among the photochemical oxidants at levels commonly
reported in ambient air, supports the recommendation  that 03 be
retained as the pollutant indicator for controlling ambient
 concentrations of photochemical oxidants.   Unless significant

-------
                              XI- 3
  additional evidence which demonstrates welfare effects from
  exposure to ambient levels of non-03 oxidants becomes available
  it ls the staffs conclusion that 03 remains a reasonable
  surrogate for protection of public health from exposure to
  photochemical oxidants.

  B.  Form of the Standard and Averaging Time(s)
       The current secondary 03 NAAQS  is expressed  as an hourly
  average which is the concentration not to be exceeded on more
  than 1  day per year on average.  During the  last  standard review
  the decision  was made  to change  from the deterministic to the
  statistical form of the standard.  The deterministic form,  which
  permitted  only a single hourly exceedance of the  standard level
  in any given year,  did not adequately deal with the variations in
  03 concentrations which are largely due to the random nature of
  meteorological  factors  affecting formation and dispersion of o
  in the atmosphere,  in  addition, EPA  further modified the
  standard so that one expected exceedance would be given a daily
  interpretation; that is, a day with two hourly values over the
 standard level counts as one exceedance of the standard level
 rather than two.  it is recommended that the statistical form
  (i.e., a number of expected exceedances allowed per year)  be
 retained for the secondary standard.
                   "" """* aPPr°priate "eraging time to protect
            and ecosystems from exposure to 03, questions
 immediately arise concerning which  components  of an exposure are
 most  important in causing vegetation  responses.   The studies
 conducted thus far offer  some useful  information in assessing the
 answer to this question,  but to date  research has not yet cllarly
 defined which  components  of  exposure  are the most critical in
 eliciting plant responses  (CD, p. 6-7) .  in lignt of  these
 uncertainties, the selection of an averaging time which
 correlates well with the effects of concern is a  difficult tart.

to the m" "t       Ut"e C°nSenSUS " ^ S0ienti"= ~-»ity as
to the most appropriate summary statistic,  but researchers are

-------
                             XI- 4
 currently using many different exposure statistics in their
 studies,  thus making comparisons between studies; extremely
 difficult.  This section assesses the information available on
 which to  base a judgment regarding the averaging time for the
 secondary standard.
      Much of the research over the last three years has been
 driven by these concerns related to the issue of averaging
 time(s).   In particular,  the  discrepancy between the seasonal
 mean  exposure indicator used  in the NCLAN studies and the
 repeated  peak exposures identified as being most important for
 plant response have  motivated several retrospective analyses
 which have attempted to investigate alternative  exposure indices
 using existing experimental data.   Based on the  weight of the new
 evidence  from analyses  of  plant response data from NCLAN and
 other research programs,  EPA  concludes  that exposure indices that
 "best"  related to plant response were those indices that
 cumulated hourly concentrations over  time  and gave greater
 weights to higher concentrations (Lee et al.,  1987a,b;  1988b;
 Hogsett et al., 1988; Lefohn  et al.,  1988).   These results
 support the  conclusions of the  CD  and the  CDS  on  the  importance
 of peak concentrations  and the  cumulative  impact  of  O3  in
 relating  plant  response to exposure.
      EPA  has  responded  directly  to  CASAC's challenge  (CASAC,
 1986)  to  identify a  single standard formulation that would offer
 protection from both repeated peaks and  long-term  exposures  of
 concern in a  series of analyses designed to identify the most
 appropriate exposure indicator.  The most recent analysis  (Lee,
 et al., 1988c) proceeded along two  lines:  (l) regression
 analysis  of plant response data against various exposure indices
 to determine  indices that "best" depict biological response, and
 (2) ambient air quality analysis of ambient 03 data to identify
 indices that correlate well with various exposure patterns.
While the results of these analyses have been discussed in some
detail previously (SP, p. x-47), the key conclusions also provide
valuable  insight regarding the issue of averaging times.

-------
                              XI-  S
      Among the indicators Lee et al.  (19880, analyzed was the
 second highest daily maximum l-hr concentration  (HDM2) because it

       10SS
  th    h    aPPr°Xlmation °f the =^ent air quality standard and
  the 7-hr seasonal mean (M7) , because it is the statistic most
  often used in the current data base for crop loss.  According to
  Lee et al.  (19Mo, ,  neither the single pea)c ^^ ^ ^  »
  term averages are biologically relevant.   The HDM2 and M7 indices
  do not adequately describe the temporal variations in exposure
  and do not  relate well to plant response.   Thus,  a major
     '           n0nCUnUlatiVe indices is «>e exclusion of exposure
 '     The  importance  of  this  factor  in  determining piant response
 *s readily apparent  when examining  yield  loss  results  for the
 wheat cultivar VONA, which is extremely sensitive to 03.   m  the
 VOKA wheat results,  there is greater variation in the   redicted
as                                     "°<~*tive indices such
as HDM2 and
                (F.gures 3.1, 3.2, Lee et al., 1988c) were used
 than when yield loss was estimated using the SUM06 and SUM07
         3.3, 3.4, Lee et al., 1988c) exposure indices.  The lower
         YTr VariabUity in PRYLS *« the cumulative indices
        and 07,  are used are iargely explained by differences in
      ortn
      or the M7.   Larsen et al.  (1983)  also points to the
 l^tations of the present !-hr O3 NAAQS  in controlling  crop
 reductxon  because it  forces some sites to control more than is
 necessary  ln order to protect against  a given  level  of crop lL
 S^lany,  cure et al.  (19M) have alluded to  some limitations of
 the M7  such as the fact that the  seasonal  mean characterizations
 of O, exposure were much less sensitive than the  l-hr max to
yearly concentration variations in O3 patterns.  This latter
pomt supports the conclusion of Lee et al.  (I988c) „.,„ hh
         speoifioally ^^ expQsure ^ £.£ -    he

                                            concentrations to the

-------
                             XI- 6
      The results of Lee et al.  (1988c)  indicate that while the
 generalized phenological cumulative impact (GPWCI)  indices best
 related plant response to 03 exposure,  there were other  indices
 that  were near optimal.   These  indices  included sigmoid-weighted
 integrated index (SIGMOID centered  at 0.062 ppm) which are much
 too complex to implement in the air quality management system of
 air pollution control.   The cumulative  censored indices  that
 integrated concentrations of 0.06  (or 0.07) ppm or  higher (SUM06
 and SUM07)  also performed well,  suggesting that ambient  O3  levels
 below 0.08 ppm are  important in triggering plant; response and
 should be included  in  an exposure index.   The cumulative censored
 indices are adequate descriptors of exposure which  relate well  to
 biological response, and are simple and easy to implement from  a
 regulatory perspective.   As stated  earlier, these results support
 the conclusions  reached  by Lefohn et al.  (1988)  and Lee  et  al.
 (1987a,b;  1988b) who used NCLAN  data and cumulative indices with
 sigmoid and allometric weights  in demonstrating the importance  of
 peak  concentrations in determining  plant response.   In addition,
 the results indicate that  fair to strong associations exist
 between the cumulative censored  indices (SUM06  and  07) and  the
 peak  and mean  indices.   The integrated indices  SUM06 AND  SUM07,
 are strongly related to  M7, and  somewhat less related to  HDM2.
 Thus,  the  relationship between SUM07 and HDM2 falls  just  below
 the level  defined by the authors (Lee et al.,  1988c) as
 indicative  of a  strong association.
     These  results suggest the cumulative indices SUM 06  and  SUM
 07 correlate well with a long-term  index (M7)  and a short-term
 index  (HDM2) and relate well to biological response.  Thus  these
 indicators have potential for setting a standard that protects
against adverse effects  from repeated peak and long-term
exposures.  While these  indicators appear to be promising,  EPA
staff concludes that the preliminary nature of the methodology
and results make a decision to change the averaging time and  form
of the standard premature at this point in time.  EPA further
believes that the analysis should be reviewed further by the

-------
                              XI- 7
  d^"""!!!^1117 and CASAC before the resuits are used ^
                                   standard form and averaging time
       The analysis, although preliminary,  does provide  CASAC with
  an additional piece of information regarding the crop  loss
  expected at alternative 1-hr standard levels.  As pointed out
  previously, these results are derived differently than Larsen et
  «1.  (1988,  in that actual air quality and plant  response data
  were used in deriving exposure-response functions.  Nevertheless,
  the  two analyses  represent different  approaches  to estimating
  crop !oss for alternative exposure indicators and provide EPA
  With information  that was not  available previously,   nilu the
  form of the 1-hr  standard has  definite limitations,  EPA staff
  concludes that establishing a  1-hr averaging time standard in an
  appropriate range  (level  of the standard is discussed in the next
  section,  represents the best staff recommendation that  could be
  made to the Administrator at this time to  close out  the review of
  the scientific data.  »ith this portion of the review complete?
  and after considering CASAC 
-------
                            XI-  8
 economic  losses  because  of the cost and time  involved  in  such an
 effort.   The views  of CASAC and  the Administrator regarding  the
 need for  such an analysis would  have to be taken into
 consideration.
     Secondly, the  option of continuing the review may appeal to
 those who do not find the current form of the standard to be a
 suitable  air quality control target and would prefer to use  one
 of the cumulative indicators suggested by Lee et al. (I988c)  once
 the analysis has been finalized.  Additional time for  review and
 revision  of this analysis would  allow the scientific community an
 opportunity to review the alternative indicators and move toward
 a consensus regarding selection  of the most appropriate exposure
 indicator.•  The  liability of this alternative is that  it
 postpones action on the secondary standard and thus, fails to
 utilize new and  existing data to assess the most appropriate
 exposure  indicator or the protection afforded by the current  1-hr
 standard.
     While evidence exists to support the hypothesis that for
 agricultural crops,  cumulative, repeated peak exposures appear to
 be more important than long-term, low level concentrations,  there
 is still considerable uncertainty regarding the O3  distribution
 patterns that affect trees.   Some CASAC members objected to the
 attempt in the previous staff paper to draw similarities between
 the responses of trees and crops to 03  and to the  suggestion that
 the exposure pattern of interest for crops might be the same  for
 trees - that is,  repeated peaks over time.  Perhaps there are no
 lessons from the years of research on crops that can be applied
 to forestry.  On the other hand,  it seems incumbent upon staff to
 examine the  total data base,  look at similarities  and differences
between crops and trees,  and lay out plausible hypotheses that
will spur debate on the key issues which science has not yet
 resolved.   It is with this in mind that staff examines  the data
 on exposure  of trees to 03 and  notes  a  good many studies  that
report the exposure period in terms of  hours  above  given
thresholds.

-------
                              XI- 9
       Mclaughlin et .1.  (19aa) reported that patterns of annual
  growth in which pine along the Cumberland Plateau reflected a
  loss in capacity for short-term recovery and eventual loss of
  tree vigor.  The primary cause of the observed decline was
  chronic exposure to elevated concentrations of O3,  possibly
  accompanied by low levels of S02  and other pollutants.   Reports
  of oxidant injury on white pine in rural Virginia (Hayes and
  SXelly,  1977,  and on the Cumberland Plateau (Mann et al.   1980)
  indicate  that  the injury appears  to be associated with  total   '
  oxidants  concentrations of 0.08 ppm or higher.   Hayes and  SXelly
  (1977) reported  concentrations which equalled  or exceeded  0  08
  ppm  for 104  hours at  one of the rural sites they monitored in
  western Virginia,  skelly and Johnston (1978, also reported that
 .03 concentrations during July of 1977 were above  0.08 ppm  30
  percent of the time at  the same location.  In addition
  relatively new evidence with slash pine Hogsett et al. (1985)
  suggests that regimes containing peaX 03 concentrations  elicit a
 greater response than regimes containing mostly lower
 concentrations over similar time periods.  Also, Wang et al
  (1936,  conceded that the 20 days  when 03 exceeded concentrations
 of 0.08 ppm and the 1 day when the concentration exceeded 0 12
 ppm was responsible  for significantly impairing the growth  of
 hybrid  poplar.
      While assessment of the above  evidence  suggests  the
otrees                                                    °
of trees to 03, further research is needed to confirm or deny
this hypothesis.  As Peterson et al.  (1987, point out, there is
currently considerable uncertainty regarding the effe^s
forest trees,  in addition, there is considerable uncertainty
regarding the O3 distribution patterns that effect trees
consequently,  EPA staff recommend that until these scientific
uncertainties can be resolved,  no separate secondary standard
'~ '— **-
              d        ~  — **- - that the secondary
         should be based on protection of vegetation

-------
                             XI- 10
      C.   Level of Standard
      A critical factor in determining the level of the standard
 is deciding which effects are to be considered adverse.  There
 are two important dimensions to this question.   The first is what
 should be measured.   The second is  how big does the change in the
 selected measure have to be to be considered adverse.   Answers to
 the question of what should be measured fall,  broadly  speaking,
 into one of two groups,  physical measures and economic measures.
 The most commonly proposed physical measure for commercial crops
 is a change in yield.   In addition,  the bulk of the data
 available since the  last standard review uses  growth and yield
 loss as  the response measure.   While a  change  in yield loss is
 the most commonly proposed physical effect  measure  for commercial
 crops, a broader range of effects could be  considered  adverse.
      It  is common to segregate  the  effects  of air pollutants on
 vegetation,  into "injury" and  "damage"  categories  (Guderian et
 al.,  1985).   Injury  encompasses  all  measurable  plant reactons,
 such as  reversible changes in metabolism, reduced photosynthesis,
 leaf necrosis,  leaf  drop,  altered quality or reduced growth that
 does not influence all effects that  reduce  the  intended  use or
 value of the  plant or plant system.   Value  or intended use  must
 be  viewed from  the perspectives  of both;  (l) value to  society  or
 human welfare and  (2) value to the organism itself  (i.e.,
 reproduction) or the surrounding  community.  From this
 perspective damage is equivalent  to an  "adverse" effect.  In this
 context,  yield  loss  is easily identified as an adverse effect.
     Concerning the magnitude of  the change that is  of concern,
 one  could pick some absolute value  (e.g., tons) or percentage
 (e.g., l%, 10%); but there is an  element of arbitrariness in the
 selection of any particular numerical value.  Clearly,  the  answer
 to the question of what constitutes  an adverse effect is a policy
 judgment.  Previous attempts to answer this question by  imposing
 a specified percentage yield decrement to be avoided were
 criticized by members of the scientific community at the 03 CASAC
meeting of April 1986.  Some CASAC members argued that 10 percent

-------
                              XI- 11
  loss in yield may not be a meaningful criterion for all crops and
  that economic significance needs to be considered when
  determining the adversity of a given effect.  For example, a 10
  percent decrement on a low valued crop does not have the same
  welfare effect as a 10 percent decrement on a high valued crop
  An economic measure of adversity can serve to weight yield
  effects according to their relative impact on the well-being of
  agricultural producers and consumers.
       For these reasons,  a  measure of adversity such as  percent  or
  absolute decrement in crop value is useful in addition  to  a
  physical effect measure  such as  percent  yield and  growth
  decrement.   However,  the range of  economic effects resulting  from
  03 reductions will not be uniform across agricultural producers
  consumers, income groups, and geographic areas.  Thus, a single'
  economic measure of adversity may be less  appropriate
 than a more extensive analysis that considers the'distribution of
 economic effects.
      Clearly, the answer to the question of what constitutes an
 adverse effect is a policy judgment.  The definition of "adverse"
 ultimately involves an interplay among scientists and society
 represented by the Administrator,  while the role of the
 scientist is to identify or describe the effect and explain why
 it  is adverse,  the Administrator  must determine if he is willing
 to  accept the impact/consequence  of the  effect or take action
 (i.e.,  expend money and resources)  to limit the consequences of
 the  adverse  effect.   Information  on both  physical effect measures
 and measures  of economic  welfare  should be  useful in making this
 policy  judgment.                                          y
     Two  national models  currently  exist that  have  used NCLAN
 crop-response data, along with adequate aerometric  and economic
 information, to estimate the magnitude and distribution of
agricultural economic benefits associated with the control of o,
 (Adams et al., 1984 and Kopp et al., 1985).   Both of these models

           :tfits as the change in producer and —
          after O3  control.   These models can be used to

-------
                             XI- 12
 calculate economic measures  of adversity (such as percent change
 in farmer profits),  as  well  as to describe  the distribution of
 economic  impacts.
     A major  shortcoming  of  the Adams  et al.  (1984)  and Kopp et
 al.  (1985)  models  is that neither model  accounted for  on-going
 U.S. agricultural  subsidy policies.  Agricultural subsidies tend
 to distort prices  such  that  they  do not  reflect the  true value of
 additional production and thus make the  value  of  increased
 production difficult to assess.   The forum  of  agricultural price
 support policy can have a substantial  impact on the  benefits
 realized  from any reduction  in 03.  Measures of economic benefits
 must be based on realistic assumptions concerning the  features of
 agricultural  policy  and benefit measures must  be  interpreted as
 being  conditional upon  the existence of the policy being  assumed.
     The  original Kopp-Adams measures of benefits  may  be
 misleading  because they ignore  crop surpluses  caused by
 agricultural  subsidies.   McGartland (1987) argued  that increased
 production  due to decreased 03 could exacerbate crop surplus
 problems, and that the  Adams et al., 1988 and Kopp et  al.,  1988
 models  substantially  overestimated the economic benefits from  03
 control.  Madariaga  (1988) argued that additional  surpluses  are
 not likely  to occur under  current provisions in the law.   The
 Food Security Act of  1985  allows for subsidy reductions and
 requires  acreage restriction increases to counter additional
 surplus production.  However, despite the provisions of this Act,
 future  farm policies and the resulting economic effects on
 agriculture are uncertain.
     Because of CASAC's interest in economic measures of
 adversity, EPA conducted an illustrative analysis  (Table XI-l)
which presents national benefit estimates for nine crops, three
 air quality scenarios, and three measures of benefits using a
modified version of the Kopp et al.,  1988 model.  The nine crops
that have been studied were selected based on economic importance
and data availability.  The three crops soybeans,  corn, and wheat
 alone account for more than one-half of the value of all U.S.

-------
                              XI- 13
  sales of agricultural crops.  The estimates in Table xi-i
  represent benefits from reducing seasonal ambient 03 levels  (7-hr
  means) to 60, 45, and 30 ppb under three different assumptions
  regarding agricultural policies,  in response to CASAC's
  recommendations, three separate measures of economic benefits are
  presented.   Measure (i),  the original Adazas/Kopp approach
  assumes no  agricultural  market distortions.   Measure (2)  the
  approach suggested by McGartland,  accounts for market distortions
  but assumes no acreage restriction increases or other policy
  adjustments to counter additional  surplus production.   Measure
  (3),  the approach suggested by Madariaga,  is a cost  savings
  measure  of  benefits that takes both market distortions  and  policy
  adjustments  to increased surpluses into  account.  The benefit
  estimates for  soybeans, peanuts, barley,  sorghum, oats,  and
  alfalfa were calculated as  the simple change in producer and
  consumer surpluses after 03 control,  since the prices of these
  crops are not  supported.
      Given present agricultural policy, the benefits of
 controlling o3 are partly dissipated by the costs of the
 additional excess production of farm crops.  This can be seen by  -
 comparing measures l and  2  in Table xi-i.  An alternative policy
 that effectively prevented  an increase in excess production would
 permit the realization of greater net  economic benefits from o,
 control.   This can be  seen  by comparing measure 3  with measure 2
 in  the table.
      The  total  benefit estimates  presented in Table XI-l can be
 used as measures of the agricultural welfare  impacts  resulting
 from glven 03 changes.  Adversity may better charactered using
 reduced economic  value  and physical effect  decrements.
 Nevertheless, the choice of  a specific threshold level of
 economic benefits that  is to be considered  as significant, is
 still somevhat  arbitrary and open to subjective judgment.
 Further, distributional concerns may be of importance.  Results
 fron the modified Kopp et al. model indicate that the
distribution!  impacts on  agriculturaZ  producers and consumers

-------
                                  XI- 14
      from reducing 03 will  be  non-neutral  and will  depend  on farm
      policies.   For example, corn,  wheat,  and cotton producers would
      benefit substantially  from O3  reductions if producer  prices  for
      these crops were fixed by agricultural  policies,   in  contrast,  it
      was  estimated that  soybean producers  would actually lose profits
      after O3 reductions since increased production would  cause large
      price declines.  The big  gainers  in this case  would be soybean
      consumers.   Any use of economic information to assess the welfare
      impacts from changing  03  levels should take these considerations

      Table XI-1.  U.S. Agricultural  Welfare Benefits from Reducing
                  Rural Ambient Ozone (7-hr seasonal means)  to 60,
                  45, and 30 ppb for Three  Alternative Benefit Measures
                  (millions  of  1986  $)

                                       Crop Benefits
Ozone

•



Reduction Benefit
(PPb)
60
60
60
45-
45
45
30
30
30
Measure
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
Soybeans
4
4
4
248
248
248
927
927
Corn
0
0
8
77
30
314
415
105
927 1019
Wheat
0
0
3
169
101
435
426
259
1000
Cotton
6
6
9
639
617
642
1351
1250
1107

Other
Crops*
N/A
N/A
N/A
N/A
N/A
N/A
103
103
103

•
Total
10
10
24
1133
996
1639
3222
2542
4156
NA = Not Available
*0ther Crops include: peanuts, barley, sorghum,  oats, and alfalfa.
     into account.  In part, because the views of CASAC regarding the
     use of economic information in determining adversity are so
     divergent, EPA plans to provide the Administrator with
     information on physical effects and economic measures of

-------
                              XI- 15
  adversity to use at his discretion in reaching a final decision
  on the O3 secondary NAAQS.
       During the last SP review the only approach available for
  analyzing estimated crop loss for alternative averaging times was
  that of Larsen et al.  (1988).  The preliminary results of Lee et
  al.  (1988C)  provide an alternative approach for evaluating
  alternative  exposures  indices and the crop losses associated with
  them.   It should be noted that while  Lee et al.  (I988c)  is
  confmed to  a  subset of  the NCLAN data,  three of  the crops
  studied (corn,  soybeans,  and wheat) account for more than one-
  half of the  value of agricultural production in the  U.S.   In
  addition, the  cultivars  selected  were the  dominant cultivars  in
  the crop  growing regions at  the time of  the study.   Therefore
  these results provide the  information necessary to assess  the
  appropriate range for a l-hr  secondary standard.
      CASAC's past comments regarding the range for the l-hr
  standard reflect a division of opinion among the Committee
 inembers.  At the April 1986 meeting the Committee indicated that
 the current standard was not protective of vegetation, but no
 alternative level was suggested.   At the December 1987 meeting
 there was support expressed for the current 0.12  ppm standard as
 well  as  for 0.10 ppm.  Based on CASAC's  written comments as well
 as  comments made at  the December 1987  meeting,  a  range of 0.08 -
 0-12  ppm appears to  capture the range  of  opinion  expressed.
that                          °f  ^  St  a1'  (1988C>'  "        •
that the upper-end  of  this  range offers very  little  protection
for vegetation.  As can be  seen  in Table X-9  (SP, p  X-52)
median crop losses of  18.9% are  expected in agricultural sites
h^rTlln9 °3 eXP°SUreS With a HDM2 value of 0.12 ppm.  Thus,
half of the crops in this study  experienced losses of 18 9% or
greater upon attainment of the current  standard,  if EPA decides
to use a more protective strategy such as the 75th percentile
rather than the median PH^s, the crop loss in agricultural

"lie xTT ^ °'12 ^ ^^ C°Uld g° - ^ - 34.8%
(Table X-9,  SP, p.  x-52).   in addition,  a clear majority of CASAC

-------
                             XI- 16
 members at the December 1988 meeting felt that the current
 standard was not protective of vegetation.  Therefore, staff
 concludes that 0.06 ppm - 0.10 ppm is an appropriate range for
 the 1-hr standard.
      Lowering the HDM2 to 0.10 ppm would result in an estimated
 median crop loss of 12.8%,  which could go as high as 28.8% when
 using the 75th percentile of PRYL's (Table X-9,  SP,  p. x-52).
 Similarly,  lowering the HMD2 to 0.08 would result in an estimated
 median crop loss of 7.9% which would go as high as 22.1% when
 using the 75th percentile of PRYL's.   EPA concludes  that in order
 to be protective of anything other than median crop  losses,  the
 bottom end  of the range needs to be lowered to 0.06  ppm.   This
 would allow for a 15.9% or  greater yield loss  in  25% of the crops
 at 0.06  ppm as compared to  22.1% or greater loss  in  25% of the
 crops at 0.08 ppm.
      While  these results are preliminary,  they do  provide
 additional  information  on which  to base  a  decision on  the level
 of  a  1-hr standard.  CASAC  comment on the  methodology  and the
 results  of  this  analysis help determine  how these  results should
 be  factored into the range.   In  the judgment of EPA  staff, a
 range  of 0.06-0.10 ppm  takes  into  consideration the  results of
 Lee et al.  (I988c) as well  as CASAC comments regarding  the upper
 end of the  range.

 D.  Staff Conclusions
     Based upon  the assessment in Section XI.A-E, the staff
 conclusions regarding the welfare effects of O3 are as follows:
     1.  In consideration of the large base of welfare
 information attributing effects to 03 exposure and the limited
evidence which demonstrates welfare effects from exposure to
ambient levels of non-O3 photochemical oxidants,  there appears  to
be little evidence to suggest a. change in chemical designation
from 03 to photochemical oxidants.
     2.  Because the bulk of the data available since the last
standard review uses growth and yield loss as the response

-------
                              XI'17  luce
  measure, this parameter could be utitt»«d to evaluate different
  levels of the secondary standard. Alternatively a broader range
  of effects could be considered adverse-including:  foliar injury
  premature senescence, reduced photosynthesis, altered carbon    '
  allocation and reduced plant vigor, ^Although there is no precise
  relationship between these effects ana-growth and yield
  reduction,  these more subtle effects'^ often early warning
  signals of  potentially harmful effe«:*=on plant vigor.   In
  addition, economic measures of advtesffcy such as changes in
  producer and consumer surpluses  resting afterO3  control may  be
  useful  in evaluating alternative levels of the  secondary
  standard .                            ve
       3.   The  weight of  the  recent evidence seems to suggest  that
  long-term averages,  such as the  7-ho«*l seasonal mean, may not  be
  adequate  indicators for relating exposure and plant response
       4. Repeated peak concentrations Accumulated over time are
  the most critical element in determining plant response for
  agricultural crops. Exposure indicates which emphasize peak
  concentrations and  accumulate concentrations over time probably
 provide the best biological basis fofostandard setting
 Unfortunately, the analysis currently available on such
 indicators is preliminary.   EPA ..«, CASAC advice and comment on
 the methodology as well as  the suggested exposure indicators
 (SUM06 and 07) .                   .  ,a .,
      5.   There is currently a laokrsrcsxposure.response
 information  on forest tree  .M.ct.rodM  addition, there  is a
 broad range  of uncertainty  among  statists regarding  o, effects
 imoorTt  treSS'. COnSe9Uent1^  «»»**•  •»  -nsensus on the most
 important averaging  time for pereftnNls  or  on the precise role  of
 03 vs. other pollutants in causing --fSMst decline.  Therefore
offt
of forest trees is not warranted, 'hat
     «.  Given the lack of effects. data on forests and the
preliminary nature of the Lee et al. (19e8c)  results regarding
selection of the appropriate exposure statistic for crops  th!

-------
                             XI- 18
 EPA staff concludes  that it  may be  premature  at  this  point in
 time to  change  the form of the  standard  and averaging time,   it
 is  our judgement  that  a 1-hr averaging time standard  in the  range
 of  0.06  - 0.10  ppm represents the best staff  recommendation  that
 could be made to  the Administrator  at this time  to  close out the
 review- of the scientific data.  With this portion of  the review
 complete,  and after  considering CASAC's  views on all  issues,  the
 Administrator will be  in a position to make a regulatory decision
 on  how and when to best act  on  the  1-hr  standard.   This is
 consistent with CASAC  comments  (December 1987 and December 1988
 meetings)  urging  EPA to consider a  1-hr  averaging time  and act  on
 the existing state of  science rather than extend the  review  until
 a more exhaustive assessment  is made of  alternative averaging
 times.
     Alternatively,  EPA could continue the standard review.
 Additional time for  review of information on alternative  exposure
 indicators would allow  the scientific community the opportunity
 to move toward a consensus regarding selection of the most
 appropriate exposure indicator.   The liability of this
 alternative is that  it  postpones action on the secondary  standard
 and thus fails to utilize new and existing data to assess the
 most appropriate exposure indicator or the protection afforded by
 the current 1-hr standard.
     7.  There appears to be no threshold level below which
 materials damage will not occur; exposure of sensitive materials
 to any non-zero concentration of O3  (including natural background
 levels)  can produce effects if the exposure duration is
 sufficiently long.  However,  the slight acceleration of aging
processes of materials which occurs at the level  of the NAAQS is
not judged to be significant or adverse.   Consequently, the staff
concludes that materials data should not  be used  as a basis for
defining an averaging time and concentration for  the secondary
standard and that the secondary standard  be based on protection
of vegetation.

-------
                             XI- 19
      8.   Effects on personal comfort and well-being,  as defined
 by  human  symptomatic effects,  have been  observed in clinical
 studies at 03 levels  in the  range  of  0.12 - 0.16  pp.  and at
 somewhat  lower levels in extended  exposure epidemiological
 studies.  These  effects include nose, and throat  irritation,
 chest discomfort, cough, and headache.   In addition,  cough,  chest
pam on deep inspiration, chest tightness, wheezing,  lassitude
mala.se, and nausea have been reported in controlled O3 exposure
stud.es.  CASAC recommended that these effects be considered
health effects in developing a basis for the primary standard for

-------

-------
                    A- 1




              APPENDIX A: AIR QUALITY
                             ,.°i

                                  -   °8-

suss s-.i'Ss's.s.'gr.ss's •a? %»

EStap'S^Srsa^asia-







researchers in investioatino o ?™~ long-term mean used by NCLAN
(Heck and Tingey? Jlvi?  All of ?h!fS -°2 .a9Jlcu""ral crops

-------
                               A- 2

                              -Air Quality),  but data
                of me  F1rst ivo Air
      The standard is attained when the expected number  of  days
              y0^ of exceedances is determined by a
 and ™«     J0rmula explained in Appendix H of 40 CFR 50
 and repeated here in simplified form to remove a term:
      1-Hour ExEx =   y *
                          n
 where:     v = number of measured 1-hour daily maximum values
               >o . 12 ppm
           N = number of days in the 03 season
           2 = number of days assumed to have a 1-hour dailv
               maximum <0.12 ppm
           n = number of days having a valid 1-hour daily
                maximum O3  value

 This  formula,  and  its application,  is described in an EPA
 guidance  document  (Curran,  1979).

      The  second air quality indicator of interest here is one of
 the variant approaches to  defining  the characteristic highes?
 concentration  (CHC) .   For  our purposes,  the CHC is that daiS
 San^n/i? <3Uality ValU!  tJ?3t is exPe<*ed  to be exceeded iomore
 than  one  time  per  year (during the  -03 season/'  actually)  on

 ?ro™ T;    ^ ar%a "UmbSr °f Ways  ^°  choose the CHC,  ranging
 from  using  fitted  statistical  distributions to a tabular "look-
 ai^XS0?? (°HrHan'  i979)'   A modified version of the look-up
 used^v «i!t lf\   ™ beCaifSe - it; is the  simPle«t method and  isP
 used  by most states  in analyzing O3 air  quality  fAMTB   19811
 The CHC under  this  approach  is theVl highest" ob^ved daily
 maximum value  where  n equals  the number  of  years of valid data.2


 qualityedatae f°°tnote on p' IV~1 regarding more recent O3 air


     2A valid year of data is one that contains a valid dailv
maximum one-hour average for 75% of all  the days  in  the  state-
defined 03 season.   A valid day has data for 75% of the hours

-------
                                A-  3
             ui                   £ta,sthe  CHC  is  the.4th-

  of data, it is the second-highest value?      "  *****  1S  °ne
  B.
  50 stages?     neauutv'Sata*!5"031 *""  (MSAs> in «»
                                                          with
  for a  MSA.   This  «as dne in    MSAS               * available
                                  "
       1-  Expected Exceedances  of  the  Existing  ozone  NAAQS


 had .ore ?han onflxpectefexcLf^  <45'1% °f th°Se With
 Thus, a signifiSanrro               Par "* ™* °*  NAAQS

 between 9 am and 9
number  of  states,       lt                     "6"^  ln
because exceedances  are rare  in
experience       concen                ™        " thSSS MSAs
                            on
daily lives,  plople  ivino in thi« S?^ 3? they g° about their
doing so, but *any people ?iiin« In evf*3 *?* Jh? fiatsntiai  for
areas never experienceo3 le"^?! greate? SSf^o ,%y polluted  °3
discussion of this, see Paul, et al  (!|86?        PP™'  F°r a

-------
                              A- 4
 and 1987.   over 18 million people live in MSAs that had between
 10 and 20  expected exceedances.                          Between

 H,*  ?art  °f the cumulative frequency distribution of l-hour ExEx
 ^  JJ!   °™ ln ??ble.A"*' along with ^lar information for
 two other  air quality indicators that will be discussed later
 The median number of l-hour expected exceedances is 1.1;  however
 there is a lot of variability in the data.   Although not  SSSrtly
 observable in Table A-l,  37% of  the MSAs  investigated4  have 0
 exceedances.   The mean of the sample,  almost 5 exceedances, is
 much greater than the median due to the very high number  of
 expected exceedances in the worst areas.   The worst MSA is Los
 Angeles, with about 125 exceedances.

      Expected exceedances data for the 1979-1987 time period were
 investigated in a sample  of the  224 MSAs  to determine if  a
 temporal trend existed in the l-hour ExEx indicator.  A MSA was
 included in this trend analysis  if it  had 3 or more valid years
 of data in the 1979-1987  period.   Under this criterion, 128 MSAs
 had sufficient data for trend analysis.   An area was  considered
 to have a  trend in its expected  exceedance  index if the slope
 coefficient for the temporal  variable  was significant at  p< 10 5
 Of the  128 areas,  only l  (
-------
                                     A-5
                                  Table A-l

            CUMULATIVE FREQUENCY DESCRIPTIVE STATISTS ^cnr
                WITH PEAK AND MULTIPLE-HOW  of J£ A?OUALm
                          INDICATORS IN URBAN AREAS
 Statistic
 Mean
 Std.  Dev.

 Minimum
 Median
 75th  Percent.

 90th  Percent.
 95th  Percent.
Maximum

Sample Size
 Expected
 Exceedances
 of 0.12 ppm
 1-Hour Daily
^Maximum
                   0.0
                   1.1
                   4.0

                  10.0
                  14.8
                 125.3

                   224
2nd-High
1-Hour

Maximum
    m]
                           .039
                           .119
                           .140

                           .167
                           .180
                           .340

                           224
                                                                 Observed
                                                                 Exceedances
                                                                 of  0.08 ppm
                                                                 8-Hour Daily
                                                                 Maximum
                             0
                             9
                            18

                            29
                            38
                           150

                           224
Note:  Std. Dev. = Standard Deviation
       Percent.  = Percentile

Source:   McCurdy and Atherton (1988).

-------
                              A- 6

 quite bad in 1988 over much of the United States,  and using 1988
 data in the regression analyses might change many  of the
 significant down-trending results.

      Additional information on trend in the l-hour ExEx indicator
 is available.   EPA's Monitoring and Data Analysis  Division (MDAD)
 undertakes air quality trends analyses of all NAAQS pollutants on
 an annual basis (MDAD,  1988).

      MDAD's trend analyses used 242 sites to investigate if a
 trend existed  in the number of expected exceedances over the ten-
 year time period 1977-1986.  The composite average  of the expected
 number of exceedances decreased 38% during that time.   1979
 1980,  and 1983  values are higher than the other years.

      2.    Second-Highest  1-hour and 8-hour Daily Maximum Air
           Quality Indicators

      Cumulative frequency descriptive data for these two air
 quality  indicators appear in  Table  A-l,  previously  introduced
 The  Table indicates  that  the  median 2nd-high 1-hour design value
 for  the  sample  of 224  MSAs  is  slightly lower than the  current 0,
 NAAQS  level.  The mean  is higher than the current standard,  0.126
 ppm,  because of very high design values  seen in a few  MSAs.   For
 instance,  10% of the MSAs have a 2nd-high daily max >0.167 ppm.
 The  worst 5% have a  design value >0.180  ppm.  The highest design
 value  sites are located in MSAs near  or  adjacent to the  New York,
 Los  Angeles, and Chicago  major metropolitan  areas.

     Looking at the  temporal trend  in the 2nd-high  daily max
 indicator, the  composite  average  of this  indicator  decreased 13%
 between  1979 and 1986  (MDAD,  1988).   This decrease  is  modest in
 the  post-1983 time period, and  1983 2nd-High  values  are  higher
 than adjacent years' values.   Thus, the  trend in this  indicator
 is not monotonic.  If trend in  individual  MSAs is investigated,
 rather than a composite average,  it is obvious that  there  is  a
 cyclical  pattern in  2nd-high daily max at  most sites.

     As  seen in Table A-l, there  are  many expected  exceedances of
 an 8-hour daily maximum >0.08 ppm.  There  also is wide
 variability in  their number, from 0 for  63 MSAs (28% of  the
 sample) to 150  for the worst MSA. The mean is about  15,  and  the
 standard  deviation of the sample  is larger than the mean.

     3.  Alternate 8-Hour Daily Maximum Air Quality  Indicators

     In previous 03 Staff Papers, the ShrDM indicator of interest
 focused on a concentration level of 0.08 ppm.  (See Table  A-l.)
Recent interest  has  focused upon alternative concentration
 levels, such as  0.06, 0.10, and  0.12 ppm.  Descriptive
 statistical data for these concentration  levels appear in  Table
A-2.

-------
                                      A-7
                                     Table A-2
  Statistic
Observed
Exceedances
of 0.06 ppm
8-hour Daily
Maximum
                                      Observed
                                      Exceedances
                                      of 0.08 ppm
                                      8-hour Daily
                                      Maximum
Observed
Exceedances
of 0.10 ppm
8-hour Daily
M ax imum
Observed
Exceedances
of 0.12 ppm
8-hour Daily
Mean
Std. Dev.
Minimum
Median
75th Percent.
90th Percent.
95th Percent.
Maximum
Sample Size
No. of MSAs >1
% of Sample
47.4
24.9
0
48
61
77
87
183
204*
197
96.6
14.9
20.9


18
f\f\
29
38
' 150
224
161
71.9
. -.
4.2
11.8

0
1
4

10
14
111
224
49
21.9
IIUA i mum
1.4
6 8
\j % %j
n
\j
0
o

2
5-
74
224
30
13.4
NOTE:  Std7~Dev7
       Percent.
Standard Deviation
Percentile
 greatly affect the statistics shown ?or

Source:   McCurdy and Atherton (1988).
                           indicator
                                                                     W°uld

-------
                              A- 8
      Note the great differences among the alternative 8-hour
 indicators in the mean number of days above the  various
 concentration "outpoints."   The average  MSA has  3  times as many
 days with an 8-hour daily maximum average over 0.10  ppm as it
 does over 0.12 ppm.   It also has about 3.5 times as  Sany days
 above 0.08 ppm as it does over 0.10  ppm.

      Finally,  Table A-2 shows that the average MSA has about 3

 HowverS  ?h?Y ^^ W*th an  8"hOUr °M >'06 PPm than da^S >'08 PP*-
 However,  this ratio is suspect because 8hrDM>.06 data are missing
 for  many  California MSAs.   Where the data are available,  the mean
 value for 8hrDM>.06 indicator would  be much higher,  as would the
 ratioo   (The median statistics are messed up for the same
 reason.)   Relationships among  the various 8-hour air quality
 indicators are more fully discussed  in subsequent  sections of
 this  Appendix.

      Note the  last two lines  of  Table  A-2.   They indicate the
 number and relative  portion  of MSAs  having  one or  more 8-hour DM
 average greater than the  cutpoint concentration  shown.   Only 22%
 of the sample  has  one or  more  8hrDM>.10 ppm, for instance,  while
 72% has one  or more  8hrDM>.08  ppm.

      4.   Longer-term Air  Quality  Indicators
                                                                »
      Two  longer-term indicators  of interest  are  the  max monthly
 and 3-month  mean  indices.  The daily maximum averaging time  used
 for these  indicators  is 1-hour and 8-hour, respectively.
 Cumulative frequency  descriptive  information for these two
 indicators is  shown  in Table A-3.

      There is  moderate  sample variability in the two  indicators
 as evidenced by the  small difference between the mean and median
 values and the  small  standard deviations relative to their means.
 The large  3-month means of the top 5% MSAs in the sample  should
 be noted.  The median value is 0.055 ppm.  The worst area has a
maximum 3-month mean  of 8-hour daily maximum averages of  0.132
ppm—higher than the  current 1-hour daily maximum NAAQS!   (The
area  is Los Angeles.)

     5.  Relationships Among Air Quality Indicators  in Urban
         Areas

     As an introduction to relationship analyses among air
quality indicators, we consider Figure A-l.  Depicted are linear
and non-parametric correlation coefficients among the air quality

-------
                                    A-9
                                 Table A-3

                                 ^
 Statistic

 Mean
 Std.  Dev.

 Minimum
 Median
 75th  Percent.

 90th  Percent.
 95th  Percent.
Maximum

Sample Size
                           Maximum
                           Monthly
                           Mean for
                           1-Hour
                           Daily
                           Ma xi mums
                            .071
                            .017

                            .025
                            .069
                            .079

                            .086
                            .099
                           .178

                            224
Maximum
3-Month
Mean for
8-Hour
Daily
Maximums
 .056
 .012

 .016
 .055
 .061

 .067
 .072
 .132

 222*
Note:  Std. Dev. = Standard Deviation
       Percent.  = Percentile
Source:  McCurdy and Atherton (1988).

-------
                                       A-10
                                        Figure A-l

                    CORRELATIONS AMONG SHORT- ANO LONG-TERM AIR QUALITY
                           INDICATORS IN MSAs  (USING  2nd  HIGH)  Q
                                   Exposure Pattern of Interest
               Peak,
               Short-Term
Multi-Peak,
Short-Term
Long-term
Average
.89
1 	 —
1
Expected
Exceed, of
1-Hour
Daily Max.







I
\&
2nd-
1-Ho
Da 11
Maxi
'





High
ur ^r
Tium
^


firt
.80
on
. oU
.70




Number
v Days> .
•^ 8-Hour
Maximum
/*"
,88
.88

\





of fi,
08 ppm *oc
Daily 4. "85 ;



— ^ Maximum
Month. Mean
_v of 1-Hour
Maximum "
3-Month Mean
» Of 8-Hour
Daily Max.
s



s
\
.95
.94


^ 	 1
                                                     Daily Max.
Source:  McCurdy and Atherton (1988).




         are depicted.   A dotted line is used when  the  coefficient  I Tell than'
         |./5|.   All  coefficients  are significant at  p <.05.

-------
                               A- 11

                                       depict

                    Kf •    ,,
                                                             -
stratification is descred  n  ordy (if 8?^     USed *°r the
All coefficien   dpte  in     «             F-
Significantly different  than  2e«"t p< of       9Ure A'2)  are

-------
                                       A- 12
                                         Figure  A- 2
                                                            AIR  QUALITV
                                   Exposure Pattern of Interest
               Peak,
               Short-Term
 Multi-Peak,
 Short-Term
 Long-term
 Average
                                                .70
                                                .74
   2nd-High
   1-Hour
   Daily
   Maximum
   V
Expected    .71
Number of
8-Hour Daily
Maximum
                                           .88
                                           .88
                                    .80
                                    .75
   y
Maximum  "
3-Month Mean
Of 8-Hour
Daily Max.
                               .95
                               .94
           Maximum
           Month. Mean
           of 1-Hour
           Daily Max.
Source:  McCurdy and Atherton (1988).

-------
Proportion  (*)  Of Sites Exceeding  Value
on Horiz. Axis
                                                                    *» m TJ
                                                                    < X 3D
                                                                    ss°
                                                                    » n
                                                                    mm
                                                                        o
                                                                        3D
                                                                    Vg2
                                                                    o x _
                                                                    OJ CD I-H
                                                                      m 2
                                                                    u jo
                                                                    •D  -O
                                                                    3 O m
                                                                   O O
                                                                   jj ^>
                          rn
                                                                   m n TJ

                                                                   T3S
                                                                   X J> CD
                                                                   02^
                                                                   3D m

                                                                   0 X 3D
                                                                   **• o m
                                                                   n c: *>
                                                                   r~ 33 tn

                                                                   ^55
                                                                  •xPS
                                                                   t« -< m
                             (a
                             c:
                             T
                             CD
                             I
                             CO
                                                                   in
                                                                  3D
                                                                  D
                                                                  t/D
                                                                     3> 2
                                                                     X G1

-------
             Proportion  (X)  of Sites  Exceeding Value on Horiz. Axis
                      ru
                      o
      t_>
>• m     T	
< x         3 :
CD T3  M.     I <
•> CD  O 4.   *J
Q> O       ^^
to  it-
(D  CD
(0  Q.
                                                                                             i
                                                                                            i—•
                                                                                            4*

-------
Proportion  (%)  of Sitea Exceeding Value  on  Horiz.  Axia
. M.
M.O
O
—J^^
IX)
o
(I)
o
Ik
o
Ul
o
0)
o
1
VI
o
GO
o
(O
o
-
                                                        o
                                                        o
                                                              c

-------
                              A- 16
 „>.  . The. Fel^t:L?nfhiPf am°ng certain air quality indicators are
 graphically depicted in Figures A-3  through A-5.   Figure A-3
 relates the number of 8-hour daily maximum averages >.08 ppm (x
 axis)  to five alternative 1-hour standards:  0.08,  0.09,  o 10
 0.11 and 0.12 ppm.   since the x-axis variable is an interval-
 scale expected exceedance variable in this case,  the curves
 follow a negative exponential shape.

      An example or two may aid the reader in reading Figure A-3
 and subsequent curves in this Section.   The  Figure indicates that
 if a 0.12  ppm 1-hour daily maximum standard  is met by all sites
 in the data set,  10% of all MSAs might have  13 or  more days with
 an 8-hour  daily maximum >.08 ppm,  on average.   At  least  one area
 might  have as many as 30 days.  Fifty percent of the MSAs might
 have 4 or  more days with an 8-hour ExEx  indicator  >.08 ppm,  on
 average.   If the l-hour daily maximum air standard was lowered  to
 0.10 ppm and attained by all areas in the data set,  10%  of the
 MSAs may have 6 or  more days over ShrDM  > .08  ppm,  on average.
 If a 0.08  ppm daily max 1-hour standard  were attained, no area
 would  be expected to have any day with an 8-hour ExEx value > 08
 ppm.

     Figure A-4 is  like Figure A-3 except that the 8-hour average
 cutoff concentration is now.0.06  ppm.  Note  that there are many
 more expected exceedances of 0.06  ppm than 0.08 ppm  for  any l-
 hour daily maximum  standard (compare  A-4  with  A-3).   The ratio  is
 anywhere between  8:1 and 20:1  for  the various  l-hour  standards
 analyzed.

     It  should  be noted that there is no  0.11  ppm  l-hour daily
 max  standard depicted  in Figure A-4.   It  so  closely  followed  the
 0.10 ppm curve  that  it  was  deleted for the sake of clarity.
 Thus,  it can be said that there  is no practical difference
 between  0.10  and  0.11 ppm 1-hour NAAQS with  respect to reducing
 the  number of days with a 8-hour daily maximum  >.06 ppm.

     In order to reduce the  expected  number  of days with an 8-
 hour 03 daily max >.06 ppm down to, say,  20 per year in  the worst
 10% of urban  areas would  require a 1-hour NAAQS of about 0.09
 ppm.  To further reduce these days to 10 per year may require a
 1-hour 03 NAAQS of 0.08 ppm.

     Figure A-5 depicts  the  relationship between alternative  1-
 hour daily max  standards  and the number of days with an  8-hour
 daily max  average >  .10 ppm.  If the current 03 standard would be
 attained everywhere, the  10% worst area probably would have only
 slightly more than 1 day per year with an 8-hour daily max >.10
 ppm, on average.  There would be less than 1 day per year  on
 average in 98%  of all areas  if an 03  NAAQS of 0.11 ppm or lower
were attained everywhere.

-------
                     A- 17

 tested.)                      y ^xmum and' cannot be







 ^covered that   *
                     °'12 PE"° ^     «y maximum

    2.  are + 42-129% for the o.io pprc standard.


    3-  ^^pofs L'r^fpp^!"9 an intervai


    4-  u^Tthr^re't^^^^na^Lr

   tS
Thus, there may be no
      Say,
'-   -  "'              '"
                           us-'"

-------
                              A- 18

 nn                  !:h°Ur daily max >'°*  PPm-   Additional work is
 underway to refine these uncertainty estimates,.

      Numerous ratio analyses  were undertaken among the urban O,
 air  quality indicators.   For  instance,  the mean ratio of the 2nd-
 highest  8-hour daily max to the 2nd highest l-hour daily max is
 0.811 (s.d.=.082),  with  a range for the 224 area data set of
 0.467 to 0.940.   These statistics are similar  to those found bv
 other investigators.                                           *

      Other  relational analyses  focused upon exceedances of
 various  8-hour outpoints vis-a-vis exceedances of the current O,
 NAAQS.   One such  analysis fitted  curves having the general form
 8hr>(c)  = a * (l  ExEx ** b),  where (c)  represents the cutpoint
 concentration of  interest (e.g.,  .08  ppm,  etc).  The fitted curves
 for  R (c >.06)  are  depicted in  Figure A-6.   While the curves show'
 the  general relationship between  the  two  air quality indices for
 alternative cutpoints, they are misleading at  ExEx=0 (at x=0)
 At ExEx=0,  the predicted number of exceedances of the alternative
 8-hour cutpoints  as  depicted  in Figure  A-6 are all 0.   In
 reality,  that is  only true for  R(.12);  at  an ExEx or x of 0,  the
 observed data indicate the following  exceedances:

                Rf.06)         Rf .081         Rf ;iO)         R ( . 1-2 )
 Minimum           o              o             o             o	
 Median          37              4             o             o
 Mean            36.1            5.0           0.3            o
 Maximum         102            23             6              o

 Thus, the equations  in general  underestimate the  number  of 8-hour
 exceedances  for the  0.06, 0.08, and 0.10 ppm cutpoints  for 0
 exceedances  of  the  1-hour daily maximum NAAQS.8
      While a statistic such as R2 is not readily available for
non-linear equations using the PC statistical package available
to us, one index of goodness-of-fit of the equation might be to
divide the "explained" sum-of-squares due to the regression
equation by the total sum-of-squares.  Using this index, the
following values are obtained:

     R(.06): 66%    R(.08): 83%    R(.IO): 96%    R(.12): 98%

Another measure of the fitting ability of the equations is to
investigate the 95% confidence intervals for the two unknown
parameters, a and b.  The predicted intervals were compared to
the estimated a and b values to give the following relative
confidence intervals for a and b, respectively:

R(.06): ±19%, ±51%   R(.08): ±16%, ±7%   R(.IO): +13%, +3%
R(.12): ± 21%, ±3%

-------
                          Figure A-6
                                              «»«
  0)
  >
  c.
  3
  o

  c
  o

  0)
  D
  r-1
  (0
  C.
  o
 0)
 0)
 o
 c
 10
 •o
 0>
 01
 o
 X
 LJ

 T3
 0)
 +J
 u
 0)
 a
 x
 UJ
O


CO
        200
        160
 160
 140
 120
 100
80
60
      40
      20
                 30       60        go       120


                 1-hour Expected  Exceedances
                                                150

-------
                              A- 20
 alt«rJ£??Vo J*    J *? that you can have many exceedances of
 alternative 8-hour daily maximum averages without equaling or
 exceeding the current NAAQS.  Attaining the 1-hour NAAQS is no
 nr?tSn n*at altefnativ* 8-hour daily maximum, in the r^nge of
 £v * H^°-?8HPpm Y11^ be atta^ed.  This conclusion is reinforced
 an Ah       analysis of specific days in 21  MSAs having l-hour
 JnH f^°UJ eX?a!dfnCeS °f the cutP°ints mentioned above (McCurdy
 and Atherton 1988).   Results of  this analysis  appear in Table A-
    If a relative frequency viewpoint of probability is taken  the
 results indicate that if a day exceeds  the current NAAQS
 concentration level  of 0.12 ppm  then it is very likely (> 95%)  to
 exceed a 0.06 or 0.08 ppm 8-hour daily  maximum' (DM)  cutpoint.   it
 On  thoth                          -        -ur     cupn.
 On  the  other  hand,  a  day  can  experience an exceedance of various
 8-hour  DM outpoints without exceeding the 0.12 ppm  1-hour DM
 i™  i*      Y  ^  °f days  exceedin9  an 8-hour DM cutpoint of 0.06
 ppm also  exceed a 1-hour  0.12 ppm concentration.  This  increases
 to  51-s  for  a  0.08 ppm cutpoint and  to 91% for a 0.10  ppm
 cutpoint. Thus, it is very possible to have days exceeding the

    **                             without e
     6.   Consecutive Daily Exceedances of the Current 1-Hour
          Standard Level

  •v nA number of CASAC members were interested in knowing how
likely it is to have two or more days in a row with a violation
of the current 1-hour O3 standard level of 0.12 ppm.  An analysis
or tnis multiple-day phenomenon was undertaken for most of the
MSAs in our sample of 224. See Table A-5.  The total number of
days in the sample with a 1-hour DM >.12 ppm is 1,002.  About 58%
of these "violations" are single-day episodes.  Almost 17% are
two-day episodes, while about 12% are three-day episodes.  Four-
day episodes constitute about 7% of violation-days,  and greater-
than-four-day episodes are 6% or so of violation-days.  Thus
almost 75% of days having a 1-hour DM >.12 ppm occur singly or in
pairs.                                                     J

     Perhaps a brief explanation of the data appearing in Table
A-5 is in order.   The table provides information on both episode
length (the left-hand column)  and how often that episode occurs
in an individual MSA (the second column) .  In addition  it
provides data on how many MSAs had a particular episode-
length/frequency combination.   For instance,  look at the 2 -day
episode length portion of the table.  The first row indicates
that 28 MSAs only had one 2-day episode.   The next row indicates
that 11 MSAs had two 2-day episodes, and so forth.  The fourth
column simply is the product of the number of MSAs times the
episode-length/frequency combination.  (For example,  for the first
row of  2-day episodes:   2 x 1 x 28 = 56.)

-------
                                    A- 21
                                 Table A-4
 Percentage of days that are:

 1. >.12 ppm 1-hour DM* &
   >8-hour outpoint shown

2. >8-hour outpoint shown .&
   >.12 ppm 1-hour DM

3. >8-hour outpoint shown  &
   <.12 ppm 1-hour DM
                                    Alternative  8-Hour Average Daily
                                       Maximum Outpoints  (in ppm)
                                    0.06
99.7
26.3
73.7
 96.7
51.3
48.7
                           58.6
                           91.0
                            9.0
     Ua7ly Maximum

-------
                                   A-  22
                                Taole A-5

                ESTIMATED FREQUENCY OF DAILY OZONE EPISODES BY
                            LENGTH OF THE EPISODES
Length of Frequency
Episodes of Episodes
(Days) in an MSA
1
2


2
3
4
7
9
3

1
2
3
4
5
6
4

1
3
4
>4 (All)
Number of
Occurrences of
That Frequency
579


28
11
3
2
1
1


13
5
1
1
1
1


7
1
2
8
Total Percent
Episode- of Total
Pays Episode Davs
579 57.5

16.6
56
44
18
16
14
18

12.3
39
30
9
12
15
18

7.2
28
12
32
62 6.2
TOTAL
                                                      1002
100.0

-------
                               A-23
                                                  the >4 category
  detail for all years of datl indi^^n* 4 ePlsodes in more
  the country have episodes o? tSat ?ena^h   ^nly a few areas of
  South coast Air Basin olcaUfS?nia^;>,  h°SS areas are the
  region.  The South Coast recency Ls h J   greatef New ^ork City
  to 30 days, and lengthlc^S™ days are rellS^T^* °f Up
                  ^^
 different  states
                 s
 valid ambient air quality dat   stuctY' 88 monitoring  sites with
 .the following SAROAD land use codes':ldentlfled as having  one  of

      32 - rural agricultural

      41 - strictly remote

      45 - other remote

 are  located     hia    sCesSsureauV^0^^ °f
 Statistical  Area  (MSA)    Thi£ fL? 5«!2 "de£lned Metropolitan
 use  coding classification  SrS ar2al%r   ^C°n5raVene the land
 as are forested areas •  tru 1 v^L?         f°Und within an MSA,
 except in unusLlclrc^llJncll    T^T.f'  howeYer'  are not'
 urban analyses  reported  in  ?h?f ' o   ?®     S Used for the non~
 (1987).  see ?ha?P?eport  ?or  idd???oia? ^ described ^ "cCurdy
 included, year  of analvsif  r  addltlonal information  on sites
highest i-hour  and s^hoSr avlrages?"""96 by State'  arid month  of
short-term peak indicaors"     *    r  lscussl°" Begins with

-------
                              A- 24

      1.   1-Hour ExEx Indicator

      A one expected exceedance indicator was calculated for each
 nonurban monitoring site.   For the 86 sites:

      •    70% attained the 0.12 ppm 03  standard  in  the  years
           analyzed.                                     J

      •    76% had fewer than  1.5  expected exceedances of  a 0 12
           ppm daily maximum one-hour value.

      •    the largest number  of expected exceedances  at a site
           was 27.3  days.9

 Cumulative frequency distribution statistics for the  one  expected
 exceedances air quality indicator is depicted in Table  A-6.   (The
 other indicators shown in  the Table will be  discussed below )   As
 can be seen,  there  is a lot of variability in the data.

      The non-urban  distribution is  very  different than  the urban
 distribution  (compare Tables  A-l  and A-6).   There are many fewer
 expected exceedances of the current 03 standard  in  rural/remote
 areas than in urban  areas.  This is  to be expected,  given  the
 urban origin  of precursor  pollutants that ultimately  result  in
 high  ambient  O3  concentrations.


      2.   Second-High Daily Maximum

      Cumulative frequency  distribution data  for  the 2nd-high
 daily max  indicator  in  rural/remote  areas  is  presented  in  Table
 A-6.   There is  not much  variability  in the data,  especially  when
 compared to that  seen  for  the  same  indicator  in  urban areas.
 While non-urban areas have  similar,  but  somewhat  lower,  mean 2nd-
 high  "design  values," there is  a  large difference in this  air
 quality  indicator for the worst areas.   The maximum 2nd-high
 indicator  seen  in rural/remote  areas  is  0.16  ppm.   It is  0 37  for
 urban  areas.
      This site is Kern County,  CA and it is very different than
most of the other rural/remote sites.  It is the "outlier," so to
speak, for many of the air quality indicators investigated.
However, the County is definitely an agricultural area.  Johnson
et al. (1986) indicates that it has the second-highest amount of
crop acreage and harvested acreage of the 84 counties in the
sample (two counties have two sites).  Thus, the site belongs in
the sample and it indicates how much variability in air quality
is possible at rural sites.

-------
                                  Table A-6
Statistic
^ — "™ ""^ """^ •""-"^—^•••••™^1

Mean
Std. Dev.

Minimum
Median
75th Percentile

90th Percentile
95th Percentile
Maximum
Sample Size

Expected
Exceedances
of 0.12 ppm
1-Hour Daily
Maximum 	

1 0
i .0
3 4
w «T
o n
U . VJ
0 0
\J 9\J
1.4
* * «
3.5
6 8
*s • W
27.3
86

2nd-High
1-Hour
Daily Max.
m"~ 1
	 ... j
™"™^^~"**^^'~"^^-^^i*^^^.^^M

.11
/"»*•»
.02

.06
.11
T O
.12
.14
1 C
.15
.16
o^r
00
Expected
Number
of Daily
Maxs >
f\f>
.08 ppm

21.1
18.1

0
17
29
48
58
132

85*
                                      °ne
Source:   McCurdy (1987)
                                                             value for

-------
                              A- 26
      3.   8-Hour Expected Exceedances

      Cumulative frequency data  for  this  indicator  are seen in
 Table A-6.  There is  relatively  a  lot  of  variability in this
 indicator.   Note the particularly large  number  of  expected
 exceedances for the  worst site  (Kern  County,  CA).

      The  data  indicate  that  it  is possible  for  rural  and  remote
 areas to  experience  many days with  a  maximum  8-hour average >0  08
 ppm.   The median area,  for instance,  has  17 days over that
 cutpoint  concentration.

      Although  not shown here, the number  of 8-hour expected
 exceedances drop dramatically as  the  cutpoint concentration
 increases.   For instance,  the median  non-urban  area has 2  days
 over  a 0.10 ppm daily max and no  days over a  0.12  ppm cutpoint
 (McCurdy, 1987).   The mean values for these two indicators is 5.7
 and 1.4,  respectively.


      4.   Longer-term Air  Quality  Indicators

      Data for  the  max monthly and 3-month mean  indicators  appear
 in Table A-7.   The 1-hour monthly mean is significantly higher
 than  the  8-hour 3-month mean for  any percentile level  chosen.
 The two distributions are statistically significantly  different
 at p<.05, using a  paired  t-test.


      5.  Discussion  of  the 3-Month Mean and Other Multiple-Hour
         Indicators

     The 3-month  indicator used here is different from that often
used to characterize O3 exposure to  agricultural crops.  Most of
these long-term studies,  such as  those conducted by the National
Crop Loss Assessment Network (NCLAN),  use a "growing season"
fixed 7-hour (9am  -  4pm)  daily average exposure statistic.

     This would be a difficult standard formulation to implement
because of variability  in growing.season across the country.
While a three month period seems  biologically relevant because it
approximates the average growing  season for most crops in the
U.S.,  there is  considerable uncertainty as to whether the fixed
daily daylight period captures the time interval of greatest
plant sensitivity.  It certainly does  not capture the time period
of maximum 03 in a day for many rural  and remote sites (McCurdy,
1984).  This is depicted  in Figure A-7,  which compares cumulative
frequency distributions of the 8-hour  daily maximum, 7-hour fixed
time period, and 1-hour daily maximum  peak air quality

-------
                              A-27
                          Table A-7
 Statistic

 Mean
 Std. Dev.

 Minimum
 Median
 75th Percentile

 90th Percentile
 95th Percentile
 Maximum

 Sample  Size
                     Maximum
                     Monthly
                     Mean for
                     1-hr Daily
                     Max. Ave.

                       .066
                       .012

                       .048
                       .067
                       .073

                       .079
                       .086
                       .118

                       86
   es: 5td. Dev. = standard Devi at ion
*tt!e
 Maximum
 3-Month
 Mean for
 8-hr Daily
.Max. Ave.

  .054
  .009

  .029
  .054
  .059

  .063
  .067
 .086

  85*
                                     °"e Slte h» • ->««1n« value for
Source:  McCurdy (1987)

-------
               A-28
              Figure   A-7

CUMULATIVE FREQUENCY DISTRIBUTION OF THREE PEAK
           AIR QUALITY INDICATORS
                          • 9ain-4pm Fixed Time
                                 Period

                          -•8-Hour Daily Max.

                          ••••1-Hour Daily Max.
                             n ~ 86
.04
.10
.12
  I
.14
  I
.16
     Ozone Concentration Level  (  ppm )

-------
                              A- 29

                           ere
                    peak ai:
  parohe                 sn   sss £est r°ng an «»»^«
  means are significant? dilfSSnt Tat po.80)T     ' '      indicators
                                    80)

  fro*        S
  values  respectively a;o9%46%   d  «n,°  °-°f3 Ppm-   These
                                 %
  hour                          onSr1 m"ns-  T"ird-guarter 7
  Mccurdy  (1985). The Sean ?w  ?h»"rf  afeas.a« analyzed in

 Undoubtedly the dif4rnc   arS ,  ?• S^ndfrd deviation value.
            iLTt
 uncover a  better exposure statistir         oer indicators may
 perspective .(Lee et af?  1988)^          3 Plant dose-
 6'   AJeai10"^1^ Am°ng Mr Qualit* indicators in Non-Urban
the current  standard in non^urbfn Jr-«   Jndlcat^s that attaining
 -
                               n  r-«
8-hour daily maximums win no  be ^IriTc^r^T^^ that
these "attaining  areas, » as many as^L? JJSS^'    ^ct  in 10% of
hour daily maximum > o 08 ppm   This n™2£  S   °°2Ur With an 8~
0.10 ppm l-hour daily maxiX' standard    *  ^°PS tO 8 days for a

-------
                                      A-30
                            Figure  A-8

              CORRELATIONS AMONG SHORT-TERM, MULTIPLE-PEAK  AND
             LONGER-TERM AIR QUALITY INDICATORS IN NON-URBAN AREAS
                        Exposure Patterns of Interest
               Peak,
               Short-
               Term
Multiple-Peak
Short-Term

    .63
                 V
               2nd-High
Expected
Number of
^?ur  £..:7.L..\ Days  .08 ppm
Daily   \      X fl-Ho
               Daily
               Maximum
8-Hour Daily
Maximum Ave.
                                   .85
                                            Longer-Term
                                            Average
                                                   .76
      y
   Maximum
v  3-Month Mean
'  of 8-Hour
   Daily Max.
                          ,76
           Maximum
           Monthly Mean
           of 1-Hour
           Daily Max.
                                                                .90
Source:  McCurdy (1987a).

         The sample size for all  indicators  is 86.   A  dotted  line is  used when
         jne correlation coefficient  between  the  two variables  is less  than
        10.75.1  All coefficients  are  significant  at  p<.05.

-------
 n>
 Q.

 fu
 r+

 Cu
to



3


<-i-
ZT

It)


Q.


r+
Cu



ro
Proportion  |%)  Of Sites Exceeding Value on Horiz. Axis
                                                                                       09
                                                                                             c
                                                                                             -J
                                                                                             n>

-------
 O>
 Q.
O)
£U
rt-
0>
fD
Q.
Q)
fu
fD
Proportion  (%)  of Sites  Exceeding Value  on Horiz.  Axis
                                                                                    £
                                                                                           (0
                                                                                           -j
                                                                                           O)
                                                                                           I
                                                                                           t-«
                                                                                           o
                                                                              >

                                                                              N)

-------
 O>
 Q.
 Cu
                   Proportion  <*>  of Sites  Exceeding  Value  on Horiz.  Axis
 rt>
 1/1
n>

QL
H-
Cu
O)
r*-
 D 2
 0) Q)
 M- X
 I—> M-
«< 3

3C 3
Q)
X Q)
M- |
3 Z
C O
     ~J (D
     Q) Ol
     (O D
     (D
     10 O
    *D CD
       C
       •3

-------
fD
Q.
CU
c+

ft)
fD
l/l
O)
Q.
Ql
CU
t/1
n>
Proportion  (%)  of Sites  Exceeding  Value on  Horiz.  Axis
                                                                                             >
                                                                                          (Q


                                                                                          n>

-------
 ct-
 O;
 rt>
 O.
 Oi
 fD
 00
0)
CL
O>
Qi
in
fD
      nj

      QL
Proportion  <*)  of Sites Exceeding

                                                                                          n
                                                                                                to
                                                                                                c
                                                                                                -J
                                                                                                I
                                                                                                I—»
                                                                                                CO

-------
                              A- 36
                            •

      The curves shown in Figure A-9  are  almost  identical to those
 shown in Figure A-3.   This  may indicate  that  air  quality
 distributions are practically  the  same for  urban  and  non-urban
 areas if the characteristic highest  concentrations10 are
 similar.

      The proportion of non-urban sites exceeding  various- max
 monthly  means given attainment of  alternative 1-hour  standards is
 depicted in  Figure A-10.  It indicates that  10%  of non-urban sites
 attaining the 0.12 ppm 03 standard could have a max monthly mean
 of  0.073  ppm or higher.   If a  0.10 ppm 1-hour standard were
 attained,  the mean for the  highest 10% of non-urban sites may  be
 >0.068 ppm.                                                  J

      Figure  A-ll  depicts  the relationship between attaining
 alternative  1-hour daily  max standards and  the maximum 3-month 8-
 hour  daily max indicator  in non-urban areas.  Attaining  a 0 12
 ppm 1-hour 03  standard might keep all but 10% of  non-urban  areas
 under a  0.06  ppm  3-month  mean.  The worst 10%  3-month mean drops
 to  0.056  ppm  for  a 0.10 ppm standard and to 0.052 ppm for a 0.08
 ppm standard.

      Figure A-12  indicates  what may happen to short-term  peaks
 when  exceedances  of a  8-hour daily maximum average are regulated.
 Attempting to  reduce short-term peaks by reducing the number of
 days  with  an  8-hour daily maximum average >0.08 ppm is
 inefficient.   It  takes a  large reduction in expected exceedances
 of a  0.08  ppm  daily max to  effectuate a small reduction in  the 1-
 hour  design value.

      A slightly more efficient relationship is shown in Figure A-
 13.   This  Figure  relates alternative max monthly mean standards
 to 1-hour  design  values.  Going from a 0.08 ppm to a 0.07 ppm  max
 monthly standard may reduce the design value from 0.125 to  0.118
 ppm in the worst  10% of non-urban sites.   A drop to a 0.06  ppm
 max monthly standard might result in a very large reduction in l-
 hour  daily maximum CHC.  For the worst 10% of non-urban sites
 the CHC could  be  as low as  0.102 ppm.
     10CHC;  in this case,  the  2nd-highest  1-hour  daily maximum
design value.

-------
                             A- 37
      Summary

86 (or urban and non-urban ar^l, r*speCtl"liy.*iySSS "" 22" and

-------

-------
                                 B-l
        APPENDIX B:  GLOSSARY OF PULMONARY TERMS AND SYMBOLS**
  Air
                                   aiveoiar
Airway conductance (Gaw) ,  Reciprocal of ai
                                                             .  T0


                                           airway resistance.  Gas
     mouth and the alveoli
                                                        to airflow

                                                  opening at the
     contrasted with  AIR SPACE
                                                                or
                                               bronchioles.   To  be
    of allegyor

    idiosyncratic hypersensitiviti4s
    cells lining the

    tissue, and a layer
                                 a
                                                            State
                                                assoclated with
                                                   o£ epithelial
                                                the lungs.   The

Adapted from Appendix A, Volume v of the CD

-------
                               B-2
 Asthma:  A disease characterized by an increased responsiveness
      of the airways to various stimuli and manifested by siowina
      of forced expiration which changes in severity either
      spontaneously or as a result of therapy.  The tlrm asthma
      may be modified by words or phrases indica^ng III Biology
      factors provoking attacks, or its duration.        etioiogy,

 Breathing pattern:  A general term designating the
      characteristics of the ventilatory activity,  e.g.  tidal
      volume,  frequency of breathing,  and shape of  th2 volume" time


 Bronchiole:   One of the finer subdivisions of the  airwavs  less
      than l  mm in diameter,  and having no cartilage fni?s wSl.

 Bronchiolitis:   Inflammation of the bronchioles which may be
      Jf^6 °r ?5ronic'   If the eti°l°9Y is known,  it should be
      stated.   If permanent occlusion  of the lumens is present
      the term bronchiolitis  obliterans may be used.

 Bronchitis:   A non-neoplastic disorder of structure or function
      or  the bronchi resulting from  infectious or noninfectious
      irritation.   The  term bronchitis  should  be modified bv
      appropriate words  or phrases to .indicate its  etioloqv  its
      chronicity,  the presence of associated airways  dysfunction .
      or  type  of anatomic  change.  The  term chronic bronchitis
      when  unqualified,  refers to a  condition  associated  with
      prolonged  exposure to nonspecific bronchial irritants  and
      accompanied by mucous hypersecretion  and certain  structural
      alterations in the bronchi.  Anatomic  changes may include
      hypertrophy of the mucous-secreting  apparatus and epithelial
      metaplasia,  as well  as more classic  evidences of  epltnellal
      inflammation.  In  epidemiologic studies,  the presence  of
      cough or sputum production on most days  for at  least three
      months of  the  year has sometimes  been  accepted  as a
      criterion  for  the  diagnosis.

Bronchoconstrictor:  An agent that causes a reduction  in the
      caliber  (diameter) of airways.

Bronchodilator:  An agent that causes  an increase in the caliber
      (diameter) of airways.

Bronchus:  One of the subdivisions of the trachea serving to
     £?J!hJy,air1t5Land-fr2m the.lun
-------
                               B-3
 Carboxyhemoglobin (COHb):   Hemoglobin in which the iron is
      associated with carbon monoxide.  The affinity of hemoglobin
      for CO is about 300  times greater than for O2f        y-"»in

 Chronic obstructive lung  disease (COLD):   This term refers to
      diseases of uncertain etiology characterized by persistent
      slowing of airflow during forced expiration.  It is
      recommended that a more specific term,  such as chronic
      obstructive bronchitis or chronic obstructive emphysema, be
      used whenever  possible.   Synonymous  with chronic obstructive
      pulmonary disease (COPD).

 Closing capacity (CC):  Closing volume plus  residual volume
      often  expressed as a  ratio of  TLC,  i.e.  (CC/TLC%).

 Closing volume (CV):   The  volume exhaled  after the expired qas
      concentration  is  inflected from an alveolar plateau during a
      controlled breathing  maneuver.   Since the value obtained is
      dependent on the  specific  test technique,  the method used
      must be  designated in the  text,  and  when necessary
      specified by a  qualifying  symbol,  closing volume is often
      expressed as a  ratio  of the VC,  i.e.  (CV/VC%).

 Conductance  (G):  The  reciprocal o£  RESISTANCE.   See AIRWAY


FEVt/FVC:  A ratio of timed (t = 0.5, 1, 2, 3 s)  forced
     expiratory volume  (FEVt) to forced vital capacity (FVC).
     The ratio  is often expressed in percent  100  x  FEV^/FVC    It
     is an index of airway obstruction.

Forced expiratory flow  (FEFx):   Related to some portion  of  the
FVC curve.  Modifiers refer to the amount of the  FVC  already
exhaled when the measurement is made.  For example:

          FEF75% =   instantaneous forced expiratory  flow after
                    75% of the FVC has been exhaled.

          FEF200-i200 =   mean forced expiratory flow between 200
                         ml and 1200 ml of the FVC  (formerly
                         called the maximum expiratory flow rate
                         (MEFR).

          FEF25-75% =    mean forced expiratory flow during the
                        middle half of the FVC [formerly called
                        the maximum mid-expiratory flow rate
                         (MMFR)].

          FEFmax =   tne maximal forced expiratory flow achieved
                    during  an FVC.

-------
                               B-4
 Forced expiratory volume (FEV) :   Denotes the volume of gas which

            capacity, e.g. (FEV^/VC) X  100.  Forced

 Forced vital capacity (FVC) :   Vital capacity performed with a
      maximally forced expiratory effort.        *°nnea witn a


 Functional residual capacity  (FRC) :  The sum of RV and ERV (the
 Gas  exchange:   Movement of oxygen from the alveoli into the
      pulmonary capillary blood as carbon dioxide enters the
      ^r'i  Jr0m th? bl°°d-   In  broader  terms'  ^e exchange of
      gases  between alveoli  and lung  capillaries.        •
             H   TraPPin
-------
                              B-5
Lung
                                                        «-
Maxina.y
     synonymous with nxi
        expiratory f low rate
                            rate (MMFR or
                   (V , :  volume of air ,reathed in one minut£_
     (fa) -  sie VENTIWTION           ( T) a"d breathing frequency
    various inorganic salt  sue         atr
                                                             the

      ease?Y VirUS' ni— 9-ism, or etiologic agent causing
                                                action of
   ingredient of photchecal so"3 a"  N°" in the air' •»

-------
                               B-6
 Pulmonary edema:  An accumulation of excessive amounts of fluid
      in the lung extravascular tissue and air spaces.

 Pulmonary emphysema:  An abnormal, permanent enlargement of the
      air spaces distal to the terminal nonrespiratory bronchiole
      accompanied by destructive changes of the alveolar £a?ls" and
      ™H??"VHV10US/ibr0sis-  The term emphysema may be
      ™J« *  by words or phrases to indicate Its etiology, its
      anatomic subtype, or any associated airways dysfunction

 Residual volume (RV) :   That volume of air remaining in the lungs

      be fndTS^1'6^1^1011-  The meth°d of measurement should
                                                 <  * appropriate
 Resistance flow (R) :   The ratio of the flow-resistive components
      .of pressure to  simultaneous flow,  in cm H20/liter pi? ?ec
      Flow-resistive  components of pressure are' obtained by
      subtracting any elastic or inertial  components,  proportional
      respectively to volume and volume acceleration  ?Mos? f low
      w??iS^nCeS 1^thS resPirat°ry Astern are  nonlinear,  varying
      with the  magnitude and direction  of  flow,  with lung volume
      and lung  volume history,  and possibly with  volume
      acceleration.   Accordingly,  careful  specification of the
      conditions of measurement is necessary;  see AIRWAY
Respiratory  frequency  (fR) :  The number of breathing cycles per
     unit of time.  Synonymous with breathing frequency  (fB).

Specific airway conductance  (SGaw) :  Airway conductance divided
     TGV       * Volume at which ifc was measured.  SRaw = Raw x
                       C°lorless 
-------
                          B-7
The method of







lung tissue (p,  ^t^S-" RreS=iSRtan°°f
                                           ^* ~~
(chest,  Whereit
 ventilatory flow rate
 is the product of SS
             conditions
                                           designates

                                             V°lume>
           V  =
           V  =
              ExPired volume  per  minute (BT.PS) ,


              Inspired volume per minute  (BTPS) .
 ALVEOLAR) .
                                                        to
                            r
gas.  Alveolar ventilton is
because when a total  volSme of
spaces, the last part does no J
but occupies the dead space  to
inspiration.  Thus the ?o?Sme of
expelled completely
volume of the
                                           (see VENTILATION,
                                             »
                                      reP.lac^d with fresh
                                       tOtal ventilation
                                         the alve°lar
                                  SXelled fro™ the body
                                             with the next
                                             actually

                                                  from the

-------
                                          Appendix C
                  UNITED STATES ENV,RONMENTAL PROTECT,ON AGENCV


                                 WASHINGTON. D.C. 20460
                                    May 1. 1989
                                                                          Office OF
                         K-


                      Protectlon
   Washington.  DC    20460

   Dear  Mr. .Reilly:

                                                                     »
    CASAC  did  not  r«.»rh

position paper recommendation  tha?" the' r^l ,°n  endo^ment  of  the  staff

-------
                                        -2-

     margin of safety Is intended  to  provWe orSo     "^   *« you are  aware.
have not yet been  uncovered  by research ,  «E  * ?   *?tn* adverse effects which
a  matter of disagreement. ^^11^^!^    ^T medlcal ^cancels
development of a  standard wfth .  ^ItSS^?  * f* subcommi«ee favored
annual  distribution  of ozone  conLSS)i?SSilly»,iIObu8t1. Upper bound °" the
expected exceedance form of the  s?anda d    Wht th  r"  ld^ncg  On  thc curr«"t
adv,ce on  what form  the  Agency should  consijlr     Comrr,!ttee offers  no further
           which alters  the
  staff position paper.  Within CASAC        Y
  felt  that healthy individuals ^
  any  of the responses c^^
  nuld  to  moderate  respiratory  symptoms) in thp
  members   believed   that  adverse  S wou?d
  mduced  more  severe effects  fle
 severe respiratory symptoms.   The
 influenced  by  recognition   {h»t
 represents  a  blending  of scientific  and
 appropriate to inform^u of
                                                           are or are  not adverse
                                                 presentat"°n  of  this issue in  the
                                                          °p!nF°n; SOme me^ers
                                                          T™  ^^  !nduced
                                                          decrement in FEV  ™
                                                          ™  Paper' whlle  alfe^
                                                     ^  w:Perience^  until  ozone
                                                                    m°derate  "'

                                                             health  effect  !ssue
                                                                      we  fee, it
                                              potentlal  for  effects  arising from
                                                   " -
based  on  recent  controlled  human
phalSUCh,  Pr°lon8ed  «posures
Further,  for people  exposed  to
possibility  that  chronic Sfevers
changes have  not  been demonstrated

                                                                        concern
                                                         a"d . toxlcol°gy studies.
                                                         «sp'ratory  impairment.
                                                           °ver  a  lifetime,  the
                                                         C°ncern' aIthouS"
and
on  various measures  of
illnesses  or exacerbation  o
Have  obvious  .imitations  in
                                                                              cite
                                                      Supplemen'
                                                 h         .the effects  of
                                                  osP|tall«tons for  respiratory

-------
                                        -3-

                                                                           >-<-
                  r
                   not  r«.a/-h  ,


  Pff on Paper recSSTm^Swi " of^n'hLr^11100 - ^  endors^nt  of  the  staff








  th«  tPhPm>a"d  1Ve favored  a" W ^.tarf  oT1 a" UP?-lr value of  '«*  *»
data  base  is  very  large and  adequate  fo *y?   |3,/ew hours) to ozone.   The

-------
                                        -4-
                                  aan   w                             -
  seasonal  and lifetime exposure* o  ozone  iTlS^  ^  * mU't!ple  hou''
  expanded  research  effort.   This  musfbe done S th^T^ *"  accele«ted and

  ozone standards  will derive from a  stronger sdent.4 base.      "  COnsIderatlo"s <>f
                                          t.    set
nature    While  th  Comittee  I           to  to^T °f 3 Stf?Ctly sdentlflc
standards  we see  no need,  in view o7 the f alrLdv  extLl  **  Y°U  °"  thc  ozone
rev,ew  the  proposed  ozone  standards ono/ to  t£?^IpCOI?meiltf ProvWed. to
g^tej;.    In  this instance,  the  public  comln.     P"bllcai°n  in the  Federal
opportunity  for the Committee  tc ^provide  anHS tirST1  Wf"  Pr°vWe
ma  be nece                      P       any addlt«onal comments  or
                             e  c  prove  an      tir
 may be necessary.                 P       any addlt«onal comments  or  review that
                 to  contact
                                       Sincerely,
                                       Roger O. McClellan. D.V.M.

                                       Chairman, Clean Air Scientific Advisory
                                        Committee

ROM:ewb

-------
United States

Environmental Protection
Agency
Office of the Administrator
science Advisory Board

Washington, DC 20460
EPA-SAB-CASAC-89-019
May 1989
   EPA
Report of the

Clean Air Scientific

Advisory Committee

(CASAC)
                         _
            f°LQzone: Closure on

            the OAQPS Staff

            Paper (1988) and the

            Criteria Document

            Supplement (1988)

-------
                    ABSTRACT
  Scientific and Technical InforLtion (19881°


    ss^yirer ssss sHTf  -
adequate scientific and fa^iSal^»2i *~ ™°fu**nt"J Provide an
pri-a^ and secondary

-------
                    NOTICE












sclent--i -P-?/^ •m.^.j.      Provide a balancer* CV^^^L.  **'d'=*»»-y.  Tne
jSs^ ^^^^^^^^

-------
                U.S. ENVIRONMENTAL PROTECTION AGENCY

                       SCIENCE ADVISORY BOARD


               LEAN AIR  SCTENTTFTC ADVTSDPV /•»nmTTTfT

                         (STATUTORY MEMBEPfi)         —
 Chairman


 Dr.
 Members

"'
Executive Secret-.ar-y
                                   Scientist, Science Advisory
                   , U.S. Environmental Protection Aaencv
         M Street, SW, Washington, D.C.  20460     Agencv'

-------
          0. S.

         Clean
                      ^iron,..nt.i Protection Agency
 Chairman

 Dr-
                    CASAC Q7n^ Pmn? mm.,,.,..
Brunswick, New Jersey
"'
          Maryland

                                         s  University,   New
                                             Health sciences,
                                         and Public Health,
                                             Futura, Bashington,
                                                      Xnstitute,
                                       Boyce
                                       POCUS me.,   Los  altos,

-------
  Dr-
                                    President-
 Dr-
                school of
Dr. Jerome J. Wesolowski, chief  Air
     «., CaUfornia Oepart^l 'O

Dr. Georae T  Wni F-F

     Laboratorie"' '
     Michigan
                                          ^^
                                                  «°tors Research
                                          Department,  Warren,
Executive
  '

-------
                            REFERENCES


 Abraham,  W.  M.;  Januszkiewicz,  A. J.; Mingle, M.; Welker, M.;
 Wanner,  A.;  Sackner,  M. A. (1980) Sensitivity of
 bronchoprovocation and tracheal mucous velocity in detecting
 airway responses to 03. J. Appl.  Physiol.  Respir.  Environ.
 Exercise Physiol. 48: 789-793.

 Adams,  W.  C.;  Schelegle, E. S.  (1983)  Ozone and high ventilation
 effects on pulmonary function and endurance performance. J. Appl.
 Physiol.:  Respir. Environ. Exercise Physiol. 55: 805-812.

 Adams,  R.  M.;  McCarl, B. A. (1985)  Assessing the benefits of
 alternative  oxidant standards on agriculture: the role of
 response  information. J. Environ. Econ.  Manage.: 12:  264-276.

 Adams,  R.  M.;  Hamilton, S. A.;  McCarl,  B.  A. (1985)  An assessment
 of  the  economic  effects of ozone on U.S.  agriculture.  J. Air
 Pollut. Control  Assoc.: 35: 938-943.

 Adams,  R.  M.;  McCarl, B. A. (1985)  Assessing the benefits of"
 alternative  ozone standards on  agriculture:  The role of response
 information. J.  Environ. Econ.  and  Mgmt.  12: 264-276.

 Adams, R.M.; Hamilton,  S.A.;  McCarl, B.A.  (1984)   The  economic
 effects of ozone on agriculture.  Corvallis, OR:   U.S.
 Environmental  Protection Agency;  EPA report  no.  EPA-600/3-84-090.
 Available  from:  NTIS,  Springfield,  VA; PB85-168441/XAB.

 Adams, W.C.;  (1987) Effects of  ozone exposure at ambient air
 pollution  episode levels on exercise performance.  Sports med.
 (Auckland) 4:  395-424.

 Adedipe, N. 0.;  Barrett,  R. E.; Ormrod, D. P.  (1972)
 Phytotoxicity  and growth responses  of ornamental bedding plants
 to  ozone and sulfur dioxide.  J. Am.  Soc. Hortic. Sci.  97:  341-
 345.

 Adedipe, N. 0.;  Ormrod,  D.  P. (1974) Ozone induced growth
 suppression in radish plants  in relation to  pre- and post-
 fumigation temperatures.  Z. Pflanzenphysiol.  71: 281-287.

Adomait, E. J.;  Ensing,  J.; Hofstra, G.  (1987) A dose-response
 function for the  impact of  O3 on Ontario-grown white bean  and an
estimate of economic  loss.  Can. J. Plant Sci.  67:  131-136.

Air Management Technology Branch  (AMTB, 1981)  Guideline for use
of city-specific  EKMA in preparing ozone SIPs.  Research Triangle
Park, N.C.:  U.S. Environmental Protection Agency  (EPA-450/480-
U £• I  •

-------
 Alpert, S. M.; Gardner, D. E.; Hurst, D. J.; Lewis, T. R. ;
 mfohi^'J?* L; £97^} Effects of exposure to ozone on defensive
 mechanisms of the lung. J. Appl. Physiol. 31: 247-252.

 Altshuller, A. P. (1977) Eye irritation as an effect of
               air pollution> j^ Air Pollute control Assoc. 27:
 Altshuller, A. P. (1983) Measurements of the products of
 atmospheric photochemical reactions in laboratory studies and in
 ambient air-relationships between ozone and other products
 Atmos.  Environ. 17: 2383-2427.
 Altshuller,  A.  P. (1986)   Review paper:  The role of nitrogen
 oxides in nonurban ozone formation in the planetary boundary
 layer over N America,  W Europe,  and adjacent areas of ocean!
 Atmospheric Environment 20:  245-268.

 American Thoracic Society (1985)  Guidelines as to what
 constitutes  an  adverse respiratory health effect, with special
 reference to epidemiologic studies of air pollution. •  Am. Rev
 Respir.  Dis.  131: 666-668.                 '

 Amiro,  B.  D.; Gillespie,  T.  J. ;  Thurtell,  G.  W.  (1984)  Injury
 response of  Phaseolus  vulgar! s to ozone flux density.  Atmos
 T?Titri >-/••> T>   1Q.1->n-7_i">ic                             J   "wiuwo.
Environ. 18: 1207-1215.
Amoruso, M.A. ;  Witz,  G.;  Goldstein,  B.  D.  (1981)  Decreased
superoxide  anion  radical  production  by  rat alveolar macrophages
following inhalation  of ozone  or  nitrogen  dioxide.   Life Sci. 28:
Aranyi, c. ; Vana,  S.  c. ;  Thomas,  P.  T. ;  Bradof,  J.  N. ;  Fenters
J. D.; Graham, J.  A.; Miller,  F.  J.  (1983)  Effects  of  subchronic
exposure to a mixture of  03, S02, and  (NH4)2 SO4 on  host defenses
of mice. J. Toxicol.  Environ.  Health 12:  55-71.

Astrand, P.O.; Rodahl, K.  (1977)  Textbook of work physiology.  New
York, N.Y.:  McGraw-Hill,  Inc.

Avol, E. L. ; Linn, W. S.;  Shamoo, D. A.;  Spier,  C.  E. ;  Valencia,
L. M.; Venet, T. G. ; Trim, S.  C. ; Hackney,  J. D.  (1987)  Short-
term respiratory effects  of photochemical oxidant exposure in
exercising children. J. Air Pollut.  Control Assoc.  37:  158-162.

Avol, E. L.; Linn, W. S.;  Shamoo, D. A.;  Hackney, J. D.  (1988)
Seasonal ozone reactivity  in Los Angeles  residents.  Presented
at: 8 1st annual meeting of the Air Pollution Control Association-
June; Dallas, TX.  Pittsburgh, PA:  Air Pollution Control
Association; paper no. 88-122.6.

-------
                                                 .• "**»•*.
                                                    ro  -soc
Am. Rev. Respir  D!S  131:^19-622?    exercisi^ adolescents
                                                               f

air pollutionin adolescents.       roSeSjnai off e^S,°f a^ient
specialty conference on the  evaluation 5 55  °f .an international

                                -



-------
  Barrv'  B-  E:'*.Mlller'  F.  J.;  Crapo,  J.  D.  (1983)  Alveolar
  rSi   5   S   ^ury caused  by inhalation  of  0.25 ppm of ozone. In:
  Lee,  S.  D.;  Mustafa, M. G.;  Mehlman,  M.  A.,  eds.  International
  symposium  on the  biomedical effects  of  ozone and  related
  photochemical oxidants; March 1982;  Pinehurst,  NC.   Princeton
  NJ:   Princeton Scientific Publishers, Inc; pp.  299-309.  (Advances
  in modern  environmental toxicology:   v.  5).

  Barry,  B.  E.  (1983) Morphometric  and  morphologic  studies of the
  effects  of asbestos, oxygen,  and  ozone  on the lung  [Ph.D.
  thesis]. Durham,  NC: Duke University.

  Bartlett,  D.,  Jr.; Faulkner,  C. S. II;  Cook,  K.  (1974) Effect of
  chronic  ozone exposure on lung elasticity in younq  rats    J
  Appl. Physiol.  37: 92-96.                               *

  Bates, D.  V.;  Bell, G. M.; Burnham, C. D.; Hazucha,  M.;  Mantha,
  J.; Pengelly,  L.  D.; Silverman, F. (1972) Short-term effects of
  ozone on the  lung. J. Appl. Physiol.  32: 176-181.

  Bates, D.  V.;  Hazucha, M.  (1973) The  short-term effects  of ozone
  on the human  lung. In: Proceedings of the conference on  health
  effects of air pollutants; October; Washington,, DC.  Washington,
  DC: -U.S Senate, Committee on Public Works; pp. 507-540-  serial
  no. 93—15.

 Bates, D. V.;  Sizto,  R. (1983) Relationships between air
. pollutant  levels and hospital admissions in  Southern Ontario
 Can.  J.  Public Health 74:  117-122.

 Bates, D. V.; Sizto,  R. (1987) Air pollution arid hospital
 admissions in southern Ontario: the acid summer haze  effect
 Environ. Res. 43:  317-331.

 Bates/ p. V.; Sizto,  R. (1989) The Ontario air pollution study:
 identification of the causative agent.  In:   International
 symposium on the health effects of acid aerosols:   addressing
 obstacles in an emerging data base; October 1987;  Research
 Triangle Park, NC.  EHP Environ.  Health Perspect.   79: 69-72.

 Beckerson,  D. W.;  Hofstra, G. (1979)  Response of leaf diffusive
 resistance  of radish,  cucumber,  and soybean to O3  and SO, singly
 or in combination. Atmos.  Environ. 13: 1263-1268".

 Bedi,  J.  F.;  Folinsbee,  L. J.; Horvath,  S.  M.; Ebenstein, R. S.
 (1979) Human exposure to sulfur dioxide  and ozone: absence of a
 synergistic effect. Arch.  Environ. Health 34: 233-239.

 Bedi,  J.  F.;  Horvath,  S. M.; Drechsler-Parks, D.  M.   (1988)
 Reproducibility of the  pulmonary function response of older men
 and women to  a 2-hour ozone exposure.  JAPCA  J.  Air Pollut
 Control  Assoc. 38: 1016-1019.

-------
Sis
                                     of dyed
                                             pollutants-
Bennett,  j. p.- oshima, R. j. ;




-------
                                 6
 iM?; T"' Tinge*' D- T- <:L977) A study of the potential ways
 in which ozone could reduce root growth and nodulation of
 soybean. Atmos. Environ. 11:  737-739.

 Boatman  E. S.; Sato, S.; Frank, R. (1974) Acute effects of ozone
 on cat lungs. II. Structural., Am. Rev. Respir. Dis. 110: 157-169.
 B°ck'  N.; Lippman, M'; Lioy' P-' Mun°Zf A.; Speizer, F. (1985)
 The effects of ozone on the pulmonary function of children. In-
 Proceedings of an inter-national specialty conference on the
 evaluation of the scientific basis for an ozone/oxidant standard-
 November 1984; Houston, TX.  Pittsburgh,  PA; Air Pollution
 Control Association; pp. 297-308. (APCA international specialty
 conference transactions: v. 4).                           «-j.aj.uy

 Bogaty  H. ;  Campbell,  K. S.; Appel,  W.  D.  (1952)   The oxidation
 of  cellulose by ozone in small concentrations.   Text. Res.  J.  22:
 8 1—83 .

 Boorman,  G.  A.;  Schwartz,  L. W. ; Dungworth,  D.  L.  (1980)
 Pulmonary effects of prolonged ozone insult in  rats: morphometric
 evaluation of the central  acinus.  Lab.  Invest.  43:  108-115.

 Bormann,  F.  H.  (1985)  Air  pollution  and forests:  an ecosystem
 perspective.  Bio Science 35: Q34-441.

 Bormann,  F.  H. ;  Smith,  W.  H. (1980)  Effects  of  air  pollution on
 forest  ecosystems.  In:  Energy and the fate  of ecosystems.
 Washington,  DC:  National Academy of  Sciences; pp. 308-318.

 Botkin, D. B.;  Smith,  W. H.; Carlson, R. W. ; Smith,  T.  L.  (1972)
 Effects of ozone on white  pine saplings: variation  in inhibition
 and recovery  of  net photosynthesis.  Environ. Pollut.  3: 273-289.

 Bradley,  C. E. ;  Haagen-Smit,  A.  J. (1951)  The  application  of
 rubber  in the quantitative determination of  ozone.   Rubber  Chem
 Technol.  24:  750-775.

 Bryan, C. L. ; Jenkinson, S.  G.  (1987) Species variation in  lung
 antioxidant enzyme  activities. J. Appl. Physiol. 63:  597-602.

 Bureau of the Census  (1982)   The journey to work.  Washington,
 D.C.: U.S.  Department of  Commerce   (Special studies, Series P-
 23 ,  No. 122) .

 Butler, L. K.; Tibbitts, T.  W. (1979) Stomatal mechanisms
determining genetic resistance to ozone in Phaseolus  vulgaris L
J. Am.  Soc. Hortic. Sci. 104: 213-216.

Cameron, J. W. (1975) Inheritance in sweet corn for resistance  to
acute ozone injury. J. Am. Soc. Hortic.  Sci. 100: 577-579.

-------
3S sS318 — "



-------
 Clarke  B.B.;  Henninger,  M.  R.;  Brennan,  E.  (1983)  An assessment
 of potato losses caused by oxidant air pollution in New Jersey.
 Phytopathology  73:  104-108.                                   *

 Clean Air Scientific Advisory Committee Meeting,  held in Research
 Triangle Park,  N.C.  on April  22 and 23,  1986.

 Clean Air Scientific Advisory Committee Meeting,  held in
 Washington,  D.c.  on  December  14-15,  1987.

 Clean Air Scientific Advisory Committee Meeting,  held in Research
 Triangle Park,  December 14-15, 1988.

 Code  of  Federal Regulations  (CFR);  various  titles and sections.

 Coffin,  D. L.;  Gardner,  D. E.  (1972)  Interaction  of  biological
 agents and chemical  air pollutants.   Ann. Occup.  Hyg.  15:  219-
 235.

 Coffin,  D. L.;  Gardner,  D. E.; Holzman,  R.  S.; Wolock,  F."  J.
 (1968) Influence  of  ozone  on  pulmonary  cells. Arch.  Environ
 Health 16: 633-636.

 Coffin,  D. L.;  Blommer,  E. J.; Garnder,-D.  E.; Holzman,  R.  (1967)
 Effect of  air pollution  on alteration of susceptibility to
 pulmonary  infection.  In: Proceedings  of  the third annual
 conference on atmospheric  contamination  in  confined  spaces; May
 Dayton,  OH.  Wright-Patterson Air Force  Base, OH:  Aerospace    '
 Medical  Research  Laboratories; pp.'71-80; report  no. AMRL-TR-67-
 200.   Available from:  NTIS,  Springfield, VA; AD-835008.

 Contant, C. F.,  Jr.;  Gehan, B. M.; Stock, T. H.;  Holguin,  A. H.;
 Buffl-er, P. A.   (1985) Estimation of  individual ozone exposures
 using  microenvironment measures.  In:  Proceedings  of  an
 international specialty  conference on the- evaluation of  the
 scientific basis  for  an  ozone/oxidant standard; November 1984;
 Houston, TX.  Pittsburgh, PA: Air Pollution Control  Association;
 pp. 250-261.  (APCA  international specialty conference
 transactions: v.  4).

 Cosio, M. G.; Hale, K. A.; Niewoehner, D. E. (1980)  Morphologic
 and morphometric  effects of prolonged cigarette smoking  on the
 small airways. Am. Rev. Respir. Dis. 122: 265-271.

 Costa, p. L. (1989) Pulmonary function studies in  the rat
 addressing concentration versus time relationships of ozone.
Presented at: Third U.S.Dutch international symposium on
atmospheric ozone research and its policy implications;  May;
Nijmegen, The Netherlands.  Amsterdam, The Netherlands:  Elsevier
 Science Publishers; p. 733-743 (Studies  in environmental Science
 35) .

-------


-------
                                 10

 De Vos, N. E.; Hill, R. R., Jr.; Pell, E. J.; Cole
 CYM3»i-i+- -i ^a^ •! »»»» j..v._._.!4._      f-           , — • w • , WVS.LC ,
 De Vos  N. E.; Pell, E. J.; Hill, R. R. , Jr. . Cole   R.  H-
                                  of potato
    S?aA A* J"' ^i***^' J' A- Bryant, L. R.  (1983) Effects  of
 combined exposure to ozone and carbon monoxide  in humans.  In:
 Mehlman, M. A.; Lee,S. D.; Mustafa, M. G., eds. International
 symposium on the biomedical effects of ozone and related
 photochemical oxidants; March 1982; Pinehurst, NC.  Princeton
 NJ; Princeton Scientific Publishers, Inc. p. 145-149  (Advances  in
 modern environmental toxicology, V.5).                (Advances  in

 DeLucia, A.J.; Adams, W.C. (1977).   Effects of ozone  inhalation
 during exercise on pulmonary function and blood biochemistry.   J
 Appl.  Physiol.: Respir. Environ. Exercise Physiol. 43: 75-81.

 Dimeo,  M.  J.;  Glenn, M. G'.;  Holtzman,  M. J.; Sheller, J. R. ;
 Nadel,  J.  A.;  Boushey,  H. A.  (1981) Threshold concentration of
 ozone  causing an increase in bronchial reactivity in humans and
 adaptation with repeated exposures. Am.  Rev. Respir. Dis. 124:
 ^4O^^4o •

 Dochinger,  L.  s.;- Bender, F.  W. ; Fox,  F. L. ; Heck, W. W. (1970)
 Chlorotic  dwarf of eastern white pine  caused by an ozone and
 sulphur dioxide interaction.   Nature (London)  225: 476.

 Dochinger,  L.  S.;  Townsend, A. M.  (1979)  Effects of roadside
 deicer  salts and ozone  on red maple progenies.  Environ.  Pollut.
 19 t  229~237.

 Donovan, D. H. ;  Williams,  S.  J. ;  Charles,  J.  M. ; Menzel,  D. B.
 (1977)  Ozone toxicity:  effect of dietary vitamin E and
 polyunsaturated fatty acids.  Toxicol.  Lett.  1:  135-139.

 Dowell,  A.  R.;  Lohrbauer,  L.  A.;  Hurst,  D.;  Lee,  S.  D. (1970)
 Rabbit  alveolar macrophage damage caused by  in  vivo ozone
 inhalation. Arch.  Environ. Health 21:  121-127.

 Drechsler-Parks, D.  M. ; Bedi, J.  F. ; Horvath, S. M.  (1984)
 Interaction of  peroxyacetyl nitrate 'and  ozone on pulmonary
 functions. Am.  Rev.  Respir. Dis.  130: 1033-1037.

Drinkwater, B.  L. ; Raven,  P.  B.; Horvath,  S. M. ; Gliner,  J.  A. ;
Ruhling, R. W. ;  Bolduan,  N. W., (1974) Air pollution,  exercise  and
heat stress. Arch. Environ. Health  28: 177-181.

-------
                              11
                                             2STi &.'"
            Ltional
                                     -  p-

                                                        -
       .  Health 5:  631-642.
Toxicol. Environ. HeSlS^eJf-fls? tO °Z°ne exp°sure' J-

-------
                                 12
              -              R- L-; Kreit' J- w-       ,
        ozone-induced lung function changes in normal  and
 TMru1! "SSJ?" ^ *?? 6ffeCt °f in<*omethacin.  Presented at:
 Third U.S. -Dutch international symposium on atmospheric ozone
 research and its policy implications.  May; Nijmegen, The
 ?S« her*an ds:  Amsterdam:  Elsevier Science Publishers; p. 493-
 499.  (Studies in environmental science 35).

 f?5aJf'r^' L-;  fchwar^2' L-  w->' K°sch,  P. c.; Dungworth, D. L.
 (1981)  Chronic bronchiolitis in nonhuman primates after prolonged
 ozone exposure. Am.J.Pathol. 105: 121-13?T              prolonged

 Evans,  L.  S.; Ting,  I.  P.  (1974)  Ozone  sensitivity of leaves:
 .ni?™?^**,   leaf.wfter content,  gas transfer resistance, and
 anatomical characteristics.   Am.  J.  Bot. 61:  592-597.
       .M; •?.•''  Johnson,  L.  V.;  Stephens,  R.  j. ;  Freeman,  G.
 (1976b)  Cell renewal in the lungs of rats exposed to low levels
 of  ozone.  Exp.  Mol.  Pathol.  24:  70-83.

 Evans, M.  J.;  Stephens,  R.  J. ;  Freeman,  G.  (1976c)  Renewal of
 pulmonary epithelium following  oxldant  injury.  In:  Bouhuys  A
 ed.  Lung cells  in  disease:  proceedings  of a Brook Lodge    '   "'
 conference;  April; Augusta,  MI.  New York,  NY: Elsevier/North-
 Holland  Biomedical Press;  pp. 165-178.

 Evans, M.  J.; Johnson,  L.  V.; Stephens,  R.  J. ;  Freeman  G
 (1976a)  Renewal of the  terminal  bronchiolar epithelium in the rat
 following  exposure to N02 or O3.  Lab. Invest. 35: 246-257.
                         .           T' J-;  Sauder'  L-  R"*
             Adaptation  in human subjects to the  effects of
inhaled ozone after repeated exposure. Am.  Rev. Respir. Dis.  119:
   ™ / j U •
Feder, W. A.  (1978) Plants as bioassay systems for monitoring
atmospheric pollutants. EHP Environ. Health Perspect.  27:  139-
147.
Feder, W. A.; Campbell, F. J.  (1968) Influence of low  levels  of
ozone on flowering of carnations. Phytopathology 58: 1038-1039.

Ferdo, A. (1985)  Ozone micr©environment factors (Memo).  Durham
N.C.: PEI Associates.                                            '

Filipowicz,  C.; McCauley, R. (1986) The effects of chronic ozone
exposure on pulmonary collagen content and collagen synthesis in
rats. J. Appl. Toxicol. 6: 87-90.

-------
                                 13

 Folinsbee,  L.  J.;  Hazucha,  M.  J. (1989)  Persistence of ozone-
 induced changes in lung function and airway responsiveness.
 Presented at:  Third U.S.-Dutch international symposium on
 atmospheric ozone  research and its policy implications; May 1988;
 Nijmegen,  The  Netherlands.   Amsterdam:   Elsevier Science
 Publishers; p.  483-492.  (Studies in environmental science 35).

 Folinsbee,  L.  J.;  McDonnell, w.  F.; Horstman,  D.H. (1988)
 Pulmonary function and symptom responses after a 6.6 hr exposure
 to  0.12 ppm ozone  with moderate exercise.  J.  Air Pollut. Control
 Assoc.  38:  28-35.

 Folinsbee,  L.  J.;  Bedi, J.  F.; Horvath,  S.  M.  (1980)  Respiratory
 responses in humans repeatedly exposed  to low concentrations of
 ozone.  Am.  Rev. Respir. Dis. 121:  43.1-439.

 Folinsbee,  L.  J.;  Silverman, F.; Shephard,  R.  J.  (1977a)  Decrease
 of  maximum  work performance following ozone exposure.  J.  Appl.
 Physiol.: Respir.  Environ.  Exercise Physiol.  42:  531-536.

 Folinsbee,  L.  J.;  Horvath,  S.  M.;  Raven,  P.  B.;  Bedi,  J.  F.;
 Morton,  A.  R.; Drinkwater,  B.  L.;  Bolduan,  N.  W.;  Gliner, J. A.
 (1977b)  Influence  of exercise  and  heat  stress  on pulmonary
 function during ozone exposure.  J.  Appl.  Physiol.:  Respir.
 Environ.  Exercise  Physiol.  43:  409-413.

 Folinsbee,  L.  J.;  Silverman, F.; Shephard,  R'.  J.  (1975)  Exercise
 responses following ozone exposure.  J. Appl. Physiol.  38: 996-
 1001.

 Folinsbee,  L.  J.;  Drinkwater,  B. L.;  Bedi,  J.  F.;  Horvath, S. M.
 (1978)  The  influence of exercise on the pulmonary  changes due to
 exposure to low concentrations of ozone.  In: Folinsbee,  L. J.;
 Wagner, J.  A.; Borgia,  J. F.;  Drinkwater, B.L.;  Gliner,  J.A.;
 Bedi, J.F.,  eds. Environmental stress: individual  human
 adaptations.  New  York, NY:  Academic  Press; pp.  125-145.

 Folinsbee,  L. J.;  Bedi, J.  F.; Horvath, S. M.  (1984) Pulmonary
 function changes in trained athletes  following 1-hour  continuous
 heavy exercise while breathing 0.21 ppm ozone. J.  Appl.  Physiol.:
 Respir. Environ. Exercise Physiol.  57: 984-988.

 Foster, K.  W.; Timm,  H.; Labanauskas, C.  K.; Oshima, R. J. (1983)
 Effects of  ozone and sulfur dioxide on tuber yield  and quality  of
potatoes. J. Environ. Qual.  12:  75-80.

Frager, N.  B.; Phalen,  R. F.;  Kenoyer, J. L. (1979) Adaptations
to ozone in reference to mucociliary  clearance. Arch.  Environ.
Health 34:  51-57.

-------
                                14

 Freeman,  G.;  Juhos,  L.  T.;  Fxiriosi,  N.  J.;  Mussenden,  R.;
 Stephens,  R.  j.;  Evans,  M.  J.,  (1974)  Pathology of pulmonary
 disease from  exposure to interdependent ambient gases  (nitrogen
 dioxide and ozone).  Arch.  Environ.  Health  29:  203-210.

 Freeman,  G.;  Stephens,  R.  J.;  Coffin, D. L.;  Stara,  J. F. (1973)
 Changes in dogs'  lungs  after  long-term  exposure to ozone: light
 and electron  microscopy. Arch.  Environ.  Health 26:  209-216.

 Fujimaki,  H.;  Ozawa,  M.; Imai,  T.;  Shimizu, F.  (1984)  Effect of
 short-term exposure  to  03 on antibody response  in mice.  Environ.
 Res.  35:  490-496.

 Fujinaka,  L.  E.  (1984)  Respiratory  bronchiolitis following long-
 term ozone exposure  in  Bonnet, monkeys:  a morphometric  study
 [master's  thesis], Davis, CA: University of California.

 Fujinaka,  L.  E.; Hyde,  D. M.; Plopper,  C. G.;  Tyler, w.  S.;
 Dungworth, D.  L.; Lollini,  L. O.  (1985) Respiratory  bronchiolitis
 following  long-term  ozone exposure  in Bonnet monkeys:  a
 morphometric  study.  Exp. Lung Res.  8: 167-190.

 Fukuda, H.; Terakado, K. (1974) On the  damage  of plants  due  to
 peroxyacyl nitrates  (PAN).  Tokyo Nogyo  Shikenjo Sokuho:  35-36
 [English abstract].

 Furukawa,  A.;  Kadota, M. (1975) Effect  of ozone on photosynthesis
 and respiration in poplar leaves. Seibutsu Kankyo Chosetsu 13:  1-
 / •

 Gardner, D. E.; Miller, F.  J.;  Illing, J. W.; Kirtz, J.  M.  (1977)
 Increased  infectivity with  exposure to ozone arid sulfuric  acid
 Toxicol.   Lett. 1: 59-64.

 Gerrity, T. R.; McDonnell,  W. F. (1989)  Do functional  changes  in
 humans correlate with airway removal efficiency of ozone?
 Proceedings of: Third U.S.-Dutch international  symposium on
 atmospheric ozone research  and  its policy implications;  May  1988;
 Nijmegen,  The Netherlands.  Amsterdam:  Elsevier Science
 Publishers; p. 293-300.  (Studies in environmental science 35).

 Gerrity, T. R.; Wiester, M. J.  (1987) Experimental measurements
 of the uptake of ozone in rats and human subjects.  Presented  at:
 80th annual meeting of the Air Pollution Control  Association;
June 21-26, 1987; New York, NY. Pittsburgh, PA:  Air Pollution
 Control Association; paper no. 87-99.3.

Gerrity, T. R. (1987) Nasopharyngeal uptake of  ozone in  humans
and animals.   Research Triangle Park, NC: U.S.  Environmental
Protection Agency, Health Effects Research Laboratory; report  no.
EPA/600/D-87/357. Available from:  NTIS,  Springfield, VA; PB88-
 140561.

-------
                   15
      S. I.; Adams, W. c.
      and ambient heat on
      57: 450-456.

                          SSBJMS..



ssar-s,

-------
                                 16
 ™«     ?' (1984> Similarities of host defense mechanisms
 against pulmonary disease in animals and man. J. Toxicol
 Environ. Health 13: 471-478.

 Grose,  E. C. ; Richards, j. H. ; Illing, j. w. ; Miller, F. J. •
 SaVJe3'-D' W*;  Graham' J- A"- Gardner, D. E. (1982) Pulmonary
 host defense responses to inhalation of sulfuric acid and ozone
 J.  Toxicol. Environ.  Health 10: 351-362.

 Grose,  E. C.; Costa,  D. L. ; Hatch,  G.  E. ; Miller,  F. J. ; Graham,
 J.  A.;  Crapo, J.  D.;  Chang, L.-Y; Stevens,  M. A. J  Jaskot, R.  H •
 Tepper,  J. s. (1989)   Chronic exposure to ozone restrictive luno;
 disease.  Research Triangle Park, NC:   U.S. Enviornmental
            A
-------
                              17
  Atmospheric ozone research and '+•  '  i«   • *' 9ran^' ^* ^* / eds
  environmental science  5



  Hackney, J. D. ;  Linn, W.  S.; Law D  C •
          sss.-sssr srs^ssr as
 Health 7: 38?-390           ^ ln huinans-  J- Toxicol. Environ.
    . (Advances in modern environmental toxicology" v'

Available fro,;      Srin?"??! 09^ . "'

-------
                                 18
ohnAo  •   '* Nlewoenner' D- E.; Cosio,  M.  G.  (1980)  Morphologic
2X?   in.muscuiar P^^onary arteries: relationship to cigarette
 **™** dlS6ase' and emPhysema. Am.  Rev.  Resir. D
   2273-18™**  lS6ase' and emPhysema.  Am.  Rev.  Respir. Dis

 Hamburger, S. J. ; Goldstein, B.  D.  (1979a)  Effect of ozone on the
                                                            °the
                  .
 of erythrocytes by concanavalin A.  II:  study of human subjects
 receiving supplemental vitamin E.   Environ.  Res.  19:  299-305.

 Hammer,  D.  I.;  Hasselblad,  V.; Portnoy,  B.;  Wehrle, P.  F.  (1974)
 Los Angeles student nurse study: daily  symptom reporting and
 photochemical oxidants.Arch.  Environ. Health 23:  255-260.

 Hanson,  G.  P.;  Thome,  L. ;  Addis, D. H.  (1975)  The ozone
 sensitivity of  Petunia hybrida Vilm. as  related to physiological
 age.  J.  Am.  Soc.  Hortic.  Sci.  100:  188-190.

 Harkema,  J.  R, ;  Plopper,  C. G.'; Hyde, D. M. ;  St.  George,  J.  A. ;
 Dungworth,  D.L.  (I987a) Effects of  an ambient level of  ozone on
 primate  nasal epithelial  mucosubstances . Quantitative
 histochemistry.  Am.  J.  Pathol.  127: 90-96.

 Harkema,  J.  R. ;  Plopper,  C. G.; Hyde, D. M.  St., George,  J. A, ;
 Wilson,  D  w.;  Dungworth, D. L.  (1987b) Response  of the macaque
 nasal epithelium to  ambient levels  of ozone.  A  morphometric  Study
 of  the transitional  and respiratory epithelium. Am. J.  Pathol.
 128:  29-44.

 Harkov,  R. ;  Brennan, E. (1979)  An ecophysiological analysis of
 the response  to  oxidant pollution.   J.  Air Poliut. Control Assoc.
 Vol. 29:  157—161.

 Harkov, R.;  Clarke,  B.; Brennan, E.  (1979)  Cadmium contamination
 may modify response  of tomato to atmospheric  ozone.  J. Air
 pollut. Control Assoc. 29: 1247-1249.

 Harvard, M. R. ;  Treshow, M. (1973)  Impact of  ozone on the growth
 and reproduction of understory plants in the  aspen zone of
western U.S.A. Environ. Conserv. 2:  17-23.

Hatch, G. E.; Slade, R. ; stead, A.  G. ;  Graham, J. A.   (1986)
Species comparison of acute inhalation  toxicity of zone and
phosgene.  J. Toxicol. Environ. Health  19:  43-53.

-------
                           19
                 and
:  Air Pollution Contro  ssociatio;
soiene

science 3S) .
                                                      on
                P 5H-5W           .             Elsevier
                            (Studies ln environmental




                             ssr
                     achr    d

-------
                                 20
        A ? ' Sllverman' P-; Parent, C. ; Field, S.; Bates, D. V.
 (1973) Pulmonary function in man after short-term exposure to
 ozone. Arch. Environ. Health 27: 183-188.

 Hazucha, M. (1981)  Assessment of ozone -induced hyperreactivitv bv
 histamine in normal healthy subjects.  In: Proceedings of the
 Jes®ar°V*a?ning W°rkshop on health effects of oxidants; January
 1980;  Raleigh,  NC.  Research Triangle Park, NC: U.S. Environmental
 Protection Agency;  pp.  314-327; EPA report no. EPA-600/9-81-001
 Available from:  NTIS,  Springfield, VA; PB81-178832.

 Heagle,  A. S.;  Spencer, S.; Letchworth,  M. B.  (1979c)  Yield
                    Wheat t0 chronic doses of ozoiM. Can. J.  Bot .
 Heagle,  A.  S.;  Heck,  W.  W. ;  Lesser,  V.  M. ;  Rawlings,  J.  Q.  (1987)
 Effects  of  daily ozone exposure duration and concentration
 fluctuation on  yield  of tobacco.  Phytopath.  77:  856-862.

 Heagle,  A.  S.;  Lesser,  V. M. ;  Rawlings,  J.  0.; Heck,  W.  W. ;
 Philbeck, R.  B.  (1986)  Response of  soybeans to chronic doses of
 ozone  applied as constant or proportional additions  to ambient
 air. Phytopath.  76: 51-56.

 Heagle,  A.  S.;  Body,  D.  E. ;  Pounds,  E.  K.  (1972)  Effect  of  ozone
 on yield of sweet corn.  Phytopathology  62:  683-687.

 Heagle,  A.  S.;  Body,  D.  E.;  Neely, G. E.  (1974)  Injury and 'yield
 responses of  soybean  to  chronic doses of ozone and sulfur dioxide
 in the field. Phytopathology 64:  132-136.

 Heagle,  A.S.; Letchworth, M.B.  (1982) Relationships among injury
 growth,  and yield responses  of  soybean  cultivars  exposed  to  ozon4
 at different  light intensities, J. Environ.  Qual. 11:  690-694.

 Heagle,  A.  S.; Philbeck, R.  B.; Knott, W. M.  (1979a)  Thresholds
 for injury  growth, and yield loss caused by  ozone on  field corn
 hybrids. Phytopathology  69:  21-26.

 Heagle,  A.  S.; Heck, W. W. ; Rawlings, J. 0.; Philbeck, R. B.
 (1983)  Effects of  chronic doses of ozone and sulfur dioxide  on
 injury and  yield  of soybeans in open-top field chambers.  Cron
 Sci. 23: 1184-1191.                                          v

 Heagle, A.  S.; Philbeck, R.  B.; Letchworth, M. B. (1979b) Injury
 and yield response of spinach cultivars to chronic doses  of  ozone
 in open-top field  chambers.  J.  Environ.  Qual. 8:   368-273.

Heath,  R. L.  (1980) Initial events in injury to plants by air
pollutants.  Annu. Rev. Plant Physiol. 31: 395-431.

-------
                                  21
                                                  *•
                                                               - c-




Heck, W.w. Personal communication.  March 6, 1986

-------
                                22

 Heggestad,  H.  E.  (1973)  Photochemical air pollution injury to
 potatoes in the Atlantic coastal  states.  Am.  Potato J. 50: 315-
 328.

 Henderson,  W.  R.;  Reinert,  R.  A.  (1979)  Yield response of four
 fresh market tomato cultivars  after acute ozone exposure in the
 seedling stage.  J.  Am.  Soc. Hortic.  Sci. 104:  754-759.

 Hoffman,  G.  J.; Maas,  E.  V.; Rawlins,  s.  L.  (1975)  Salinity-ozone
 interactive effects on alfalfa yield and water relations. J
 Environ.  Qual.4: 326-331.

 Hofstra,  G.;  Littlejohns, D. A.;  Wukasch, R.  T.  (1978) The
 efficacy of the antioxidant ethylene-diurea  (EDU)  compared to
 carboxin and benomyl in  reducing  yield losses from ozone in navy
 bean.  Plant.  Dis.  Rep.62: 350-352.

 Hogsett,  W.  E.; Tingey,  D.  T.;  Lee,  E. H.  (1983)  Ozone exposure
 indices:  Concepts  for  development and  evaluation  of their use.
 In: Assessment of  Crop Loss from  Air Pollutants,  pp.  107-138.*
 W.W.  Heck,  O.C. Taylor and  D.T. Tingey (eds.).  - Elsevier Applied
 Science,  New York.

 Hogsett,  W.  E.; Tingey,  D.  T.;  Holman, S.  R.  (1985)  A
 programmable exposure  control  system for  determination of the
 effects  of  pollutant exposure  regimes  on  plant growth. Atmos.
 Environ.  19: 1135-1145.

 Holguin,  A. H.; Buffler, P. A.; Contant,  C. F., Jr.;  Stock,  T,
 H.; Kotchmar,D. J.;  Hsi, B. P.; Jenkins,  D. E.; Gehan, B.  M.;
 Noel, L.  M.; Mei, M.  (1985) The effects of ozone  on asthmatics  in
 the Houston area. In:  Proceedings of an intermrtional specialty
 conference  on the evaluation of the scientific basis  for an
 ozone/oxidant standard; November  1984; Houston, TX.   Pittsburgh,
 PA: Air Pollution Control Association; pp. 262-280.  (APCA
 international specialty conference transactions:  v.  4).

 Holtzman, M. I.; Cunningham, J. H.; Sheller,  J. R.;  Irsigler, G.
 B.; Nadel, J. A.; Boushey,  H. A.  (1979) Effect of ozone on
 bronchial reactivity inatopic and nonatopic subjects.  Am.  Rev.
 Respir. Dis. 120: 1059-1067.

 Holtzman, M. J.; Fabbri, L. M.; Skoogh, B.E.; O'Byrne,  P.  M.;
 Walters,  E. H.;Aizawa, H.;  Nadel,  J. A. (I983a) Time  course  of
 airway hyperresponsiveness  induced by ozone in dogs.  J.  Appl.
 Physiol.: Respir. Environ.  Exercise Physiol.  55:  1232-1236.

Holtzman, M. J.; Fabbri, L. M.; O'Byrne, P. M.; Gold,  B.  D.;
Aizawa, H.; Walters, E. H.; Alpert, S. E.; Nadel, J.  A.  (1983b)
 Importance of airway inflammation for hyperresponsiveness  induced
 by ozone. Am. Rev.  Respir.  Dis. 127 686-690.

-------
                                 23
 Horsman, D. C.; Nicholls, A. 0.; Calder, D. M. (1980) Growth
 responses of Dactylis glomerata, Lolium perenne and Phalaris
 aquatica to chronic ozone exposure. Aust. J. Plant Physiol. 7:
 511—517.

 Horstman, D.; McDonnell, W.; Abdul-Salaam, S.; Folinsbee, L.;
 Ives,  P. (1988)  Current US EPA research concerning more
 pronounced human exposures to low ozone concentrations.
 Presented at: 81st annual meeting of the Air Pollution Control
 Association; June; Dallas, TX.  Pittsburgh, PA: Air Pollution
 Control Association; paper no. 88-122.5.

 Horstman, D.; McDonnell, W.; Folinsbee, L.; Abdul-Salaam, S.;
 Ives,  P. (1989)  Changes in pulmonary function and airway
 reactivity due to prolonged exposure to typical ambient ozone
 levels.   In:  Schneider, T.,  Lee,  S. D.; Wolters,  G.  J. R.;
 Grant,  L. D. eds.   Atmospheric ozone research and its policy
 implications:   Proceedings of:  The Third U.S.-Dutch
 international symposium on atmospheric ozone research and its
 policy implications; May; Nijmegen,  The Netherlands.   Amsterdam
 The Netherlands:   Elsevier Science Publishers;  pp.  755-762.
 (Studies in environmental science  35).

 Horvath,  S.  M.;  Gliner,  J.  A.; Folinsbee,  L.  J. (1981)  Adaptation
 to  ozone duration  of effect.  Am. Rev.  Respir.  Dis.  123:  496-499.

 Horvath,  S.  M.;  Gliner,  J.  A.; Matsen-Twisdale, J.  A.  (1979)
 Pulmonary function and  maximum exercise responses  following acute
 ozone  exposure.  Aviat.  Space  Environ.  Med.  50:  901-905.

 Howell,  R.  K.;'Koch,  E.  J.;  Rose,  L.  P.,  Jr.  (1979)  Field
 assessment  of  air  pollution-induced  soybean yield  losses.  Agron
 J.  71: 285-288.

 Howell,  R.  K.; Rose,  L.  P., Jr.  (1980)  Residual air pollution
 effects  on  soy-bean seed quality.  Plant Dis.  64: 385-386.

 Howitt,  R.  E.; Gossard,  T.  E.;  Adams, R. M.  (1984a) Effects of
 alternative  ozone  levels and  response data  on economic
 assessments: The case of California  crops.  J. Air pollut.  Control
 Assoc. 34:  1122-1127.

 Howitt, R. E.; Gossard,  T.  E.;  Adams, R. M.  (1984b)  Applied
 Welfare Economics  and Public  Policy.  New york: Prentice-Hall.

 Hucl, P.; Beyersdorf, W.  D.  (1982) The  inheritance  of ozone
 sensitivity  in selected  Phaseolus vulgaris  L. populations.  Can
 J. Plant  Sci. 62:  861-865.

Hunt, W.  F.  ed.  (1985)   National air quality and emissions trends
report, 1983.  Research  Triangle Park, N.C.:  U.S.  Environmental
Protection Agency   (EPA  450/4-84-029).

-------
                                24

 Hunt,  W.  F.;  Curran,  T.;  and Freas,  W.  P.  (1987)   National and
 regional  trends  in ambient  ozone  and nitrogen dioxide
 measurements,  1976-1985.  Durham,  N.C.:  U.S.  Environmental
 Protection Agency.

 Hurst,  D. J.;  Gardner, D. E.;  Coffin, D. L.  (1970)  Effect of
 ozone  on  acid hydrolases  of the pulmonary  alveolar macrophage. J.
 Reticuloendothel.  Soc. 8: 288-300.

 Hurst,  D. J.;  Coffin, D.  L.  (1971) Ozone effect on lysosomal
 hydrolases of  alveolar macrophages in vitro.  Arch.  Intern.  Med.
 127: 1059-1063.

 Hyde,  D.  M.;  Plopper, C.  G.  (1989) Ozone-induced  structural
 changes in monkey  respiratory  system.  Proceedings of:  The Third
 U.S.-Dutch international  symposium on atmospheric ozone research
 and its policy implications; May; Nijmegen, The Netherlands.
 Amsterdam, The Netherlands:  Elsevier Science Publishers;  p.  523-
 532.   (Studies in  environmental science  35).

 Illing, J. W.; Miller, F. J.;  Gardner, D.  E.  (1980) Decreased
 resistance to  infection in  exercised mice  exposed to  NO 2   and O
 3  . J.  Toxicol.  Environ. Health 6: 843-851.

 Jacobson, J. S.  (1977) The  effects of photochemical oxidants  on
 vegetation. In:  Ozon und Begleitsubstanzen im photochemischen
 Smog.  [Ozone and other substances in photochemical  smog]:  VDI
 colloguium; 1976; Dusseldorf,  West Germany:   Verein deutscher
 Ingenieure (VDI) GmbH; pp.  163-173.  (VDI-Berichte no. 270).

 Jacobson, J. S.  (1982) Ozone and the growth and productivity  of
 agricultural crops. In: Unsworth, M.  H.; Ormrod,  D. P.,  eds.
 Effects of gaseous air pollution in agriculture and horticulture.
 London, United Kingdom: Butterworth Scientific; pp. 293-304.

 James,  C. (1980) Economic,  social and political implications  of
 crop losses: a holistic framework for loss assessment in
 agricultural systems. In: Crop loss assessment:; proceedings of E.
 C. Stakman commemorative symposium; miscellaneous publication no.
 7; pp.  10-16.

 Javitz, H. S.; Kransnow, R.; Thompson, C.; Patton, K. M.;
 Berthiaume, D. E.; Palmer, A.  (1983)  Ambient  oxidant
 concentrations in Houston and  acute health symptoms in  subjects
with chronic obstructive pulmonary disease: a  reanalysis of the
HAOS health study.  In:  Lee,  S.D.; Mustafa, M..G.; Mehlman, M.A. ,
 eds.  International symposium  on the biomedical effects  of ozone
 and related photochemical oxidants; March  1982; Pinehurst, NC.
Princeton, NJ: Princeton Scientific Publishers,, Inc.; pp.  227-
 256. (Advances in modern toxicology:  v. 5).

-------
                                25
 SwRl project no. 01-4902          Southwest Research Institute.
 2SS:
                                 .

v. 5).              (Advances ln modern environmental toxicology:
elferti on;reioy rcton6"0?  ar P°"^ion=  its
Arch.  Environ PHeal?S 30? ?l"?2ff  elenentalT sch°°l children.
        ren      ed toauonn

-------
                                 26
 Kagawa, J.; Tsuru, K. (1979b)  Respiratory effects of 2-hour
 exposure to ozone and nitrogen dioxide alone and in combination
 in normal subjects performing  intermittent exercise. Nippon Kyobu
 Shikkan Gakkai Zasshi 17:  765-774.                           ^yoou

 Kagawa, J.; Tsuru, K. (1979a)  Effects of ozone and smoking alone
 and in combination on bronchial reactivity to inhaled
 acetylcholine. Nippon KyobuShikkan  Gakkai Zasshi 17: 703-709.

 Kamath, Y.  K. ; Ruetsch,  S.  B.;  Weigmann,  H.  D.  (1982)
 Microspectrophotometric  study  of ozone fading of disperse dves in
 nylon.   Text.  Res. J. : 53:  391-402.
      '  H*  R*;  vincent'  L-  M.;  Kowalsky,  R.  J. ;  Horstman,  D.  H. ;
 O'Neil,  J.  j.; McCartney,  W. H. ;  Bromberg,  P. A.  (1987)  Ozone
 exposure increases  respiratory epithelial permeability in humans.
 Am.  Rev.  Respir.  Dis.  135:  1124-1128.

 Kehrl,  H.  R. ;  Hazucha,  M.  J. ;  Solic, J.  j.;  Bromberg,  P.  A.
 (1985)  Responses  of subjects with chronic pulmonary disease  after
 exposures  to 0.3  ppm ozone. Am. Rev. Respir. Dis.  131:  719-724.

 Kehrl,  H.  R. ;  Vincent,  L.  M. ;  Kowalsky,  R.  J. ;  Horstman,  D.  H. ;
 O'Neil,  J.  j.;  McCartney,  W. H. ;  Bromberg,  P. A.  (1989)  Ozone-
 induced  changes in  the  pulmonary  clearance  of 99mTc-DTPA in man
 In:  Schneider, T. ;  Lee, S. D. ; Wolters, G.  J.  R. ;  Grant,  L.  D.'
 eds.  Atmospheric ozone research  and its policy implications:
 Proceedings of:   The Third U.S. -Dutch international symposium;
 May  1988; Nijmegen,  The Netherlands.  Amsterdam, The Netherlands-
 Elsevier Science  Publishers; pp.  343-351.   (Studies in
 environmental  science 35) .

 Kehrl, H. R. ;  Hazucha,  M. J.;  Solic, J.  J.;  Bromberg, P. A.
 (1983) The acute  effects of 0.2 and 0.3  ppm  ozone  in persons with
 chronic obstructive  lung disease  (COLD). In: Lee,  S. D. ; Mustafa,
 M. G.^Mehlman, M. A.,  eds. International symposium on the
 biomedical effects of ozone and related  photochemical oxidants;
March 1982; Pinehurst,  NC.  Princeton, NJ: Princeton Scientific
 Publishers, Inc.; pp. 213-225.   (Advances in modern environmental
 toxicology: v.   5) .

Kelleher, T. J. ;  Feder, W. A.  (1978) Phytotoxic concentrations of
 ozone on Nantucket Island: long range transport from the middle
Atlantic states over the open ocean confirmed by bioassay  with
ozone-sensitive tobacco plants.  Environ. Pollut.  17:  187-194.

Kelly, N. A.; Wolff, G. T.; and Ferman, M. A. (1984)  Sources  and
sinks of ozone  in rural areas.   Atmospheric Environment 18: 1251-
 1266 .

-------
                             27
                               -
Respir. 2: i-l.               er* ceara*ce.  Exp. Lung
                     Health 4:  521-532


                                      «
  Rev. Reir.  is    s    s"* asthmatic -dole-cents.
                                           2 S5S3-

-------
                                28

 Kopp,  R.J.; Vaughan,  W.J.;  Hazilla,  M.  (1984)   Agricultural
 sector benefits analysis for ozone:   methods evaluation and
 demonstration.  Research Triangle  Park,  NC:  U.S. Environmental
 Protection Agency,  OAQPS; EPA report no.  EPA-450/5-84-003
 Available from:  NTIS,  Springfield,  VA; PB85-119477/XAB.

 Koren,  H.  S.;  Devlin,  R.  B.; Graham,  D. E.; Mann,  R.; Horstman,
,D.  H.  (1988a)  Cellular and  biochemical  changes in  the lower
 airways of subjects exposed to ozone.   Research Triangle Park
 NC:  U.  S.  Environmental Protection Agency,  Health  Effects
 Research Laboratory;  report no. EPA/600/D-88/031.  Available from-
 NTIS;  Springfield,  VA;  PB88-170048.

 Koren,  H.  S.;  Devlin,  R.  B.; Graham,  D. E.; Mann,  R.; Horstman,
 D.  H.;  Kozumbo, W.  J.;  Becker,  S.; McDonnell,  W. F.  (1988b)
 Cellular and biochemical  changes  in  the lower  airways of subjects
 exposed to ozone. In:  Sorg,  C., ed.  The alveolar macrophage.
 Local  Immunity  4: 36-49.

 Koren;  H.  S.; Devlin,  R.  B.; Graham,  D.  E.; Mann,  R.; McDonnel,
.W.  F.  (1989) The inflammatory response  in human lungs exposed to
 ambient levels  of ozone.  In:   Schneider, T.;  Lee, S. D.;
 Wolters, G.  J.  R.;  Grant, L.  D.,  eds.   Atmospheric ozone research
 and  its policy  implications:   Proceedings of:   The Third U.S.-
 Dutch  international symposium; May 1988; Nijmegen, The
 Netherlands. Amsterdam, The  Netherlands:  Elsevier Science
 Publishers,  pp.  745-753.   (Studies in environmental science  35).
                                <                               '
 Kreit,  J.  W.; Gross, K. B.;  Moore, T. B.; Lorenzen, T.  J.;
 D'Arcy,  J.;  Eschenbacher, W.  L. (1988)  Ozone-induced  changes  in
 pulmonary  function  and  bronchial  responsiveness in asthmatics.
 J. Appl. Physiol. 66:   217-222.

 Kress,  L.  W.; Skelly, J.  M.  (1982a) Response of several eastern
 forest  tree  species to  chronic doses of ozone  and  nitrogen
 dioxide. Plant  Dis. 66: 1149-1152.

 Kress,  L.  W.; Skelly, J.  M.;  Hinkelmann, K. H.  (1982b)  Growth
 impact  of  O3, N02, and/or SO2 on Pinus taeda. Environ. Monit.
Assess.  1: 229-239.

Kress,  L.  W.; Skelly, J.  M.;  Hinkelmann, K. H.  (1982c)  Growth
 impact  of  O3, NO2, and/or SO2 on Platanus occidentalis. Agric.
Environ. 7:  265-274.

Kress,  L.  W.; Skelly, I.  M.;  Hinkelmann, K. H.  (1982d)  Relative
sensitivity  of  18 full-sib families of Pinus toeda to 0 3  . Can.
J. For. Res. 12: 203-209.

Kress,  L. W.; Miller, J.  E.;  Smith, H. J. (I985a)  Impact of ozone
on winter wheat yield. Environ. Exp.  Bot. 25:  211-228.

-------
                                   29
                                                  «««• « «•" com

 needed source reductions: Part  ?  An o -n  T
 mathematical model,  j. Air Pollut  coJ^  A    lni?Ury
                         AJ.IT i'oj.iut:.  Control  Assoc. 33: 198-207.

ESS? Slnt^n^:/^" *°*™» «°P «*«*ion.  ,.  Mr

-------
                                 30
 Last,  J. A.; Greenberg, D. B.; Castleman, W. L. (1979) Ozone-
 induced alterations in collagen metabolism of rat lungs. Toxicol.
 Appl.  Pharmacol. 51: 247-258.                      •

 Last,  J. A.; Greenberg, D. B. (1980)  Ozone-induced alterations in
 collagen metabolism of rat lungs. II. Long-term exposure.
 Toxicol. Appl.  Pharmacol.  55: 108-114.

 Last,  J. A.; Hesterberg,  T.  W.;  Reiser,  K.  M.;  Cross, C. E. ;
 ^?J!i'tTA C"* ?U2n'  C"' Steffev'  E-  *•; Grandy,  J.; Henrickson, R.
 (1981)  Ozone-induced alterations in collagen metabolism of monkey
 lungs:  use of biopsy-obtained lung  tissue.   Toxicol.  Apol
 Pharmacol.  60:  579-585.

 Last,  J. A.; Gerriets, J.  E.; Hyde,  D. M.  (1983)  Synergistic
 effects on rat  lungs of mixture  of  oxidant  air  pollutants (ozone
 or  nitrogen dioxide)and respirable  aerosols.  Am.  Rev. Resnir
 Dis. 128:  539-544.                                        P

 Last,  J. A.; Hyde,  D.  M.;  Chang,  D.  P. Y.  (1984a)  A mechanism of
 synergistic lung damage by ozone and  a respirable  aerosol  EXD
 Lung Res.  7:  223-235.                                         F*

 Last, J.  A.;  Reiser,  K. M.; Tyler, W.  S.; Rucker,  R.  B.  (1984b)
 Long-term consequences of  exposure to ozone;  I.  lung  collagen
 content.  Toxicol. Appl. Pharmacol. 72:  111-118.

 Laurence, J.  A.;  Wood,  F.  A.  (1978) Effects  of  ozone  on  infection
 of  soybean  by Pseudomonas  glycinea. Phytopathology 68: 441-445.

 Lauritzen,  S. K. ; Adams, W, C,. (1985)  Ozone  inhalation effects
 consequent  to continuous exercise in  females: comparison to
 males.  J. Appl.  Physiol. 59:  1601-1606.

 Lebowitz, M.  D.;  O'Rourke, M. K.; Dodge, R.;  Holberg,  C.  J. ;
 Gorman,  G.;  Hoshaw, R.  W.; Pinnas, J.  L.; Barbee,  R.  A.;  Sneller,
M. R.  (1982)  The  adverse health effects  of biological aerosols,
other aerosols,  and indoor microclimate  on asthmatics and
nonasthmatics.   Environ. Int. 8: 375-380.

Lebowitz, M.  D.;  Holberg, C.  J.; Dodge,  R. R. (1983)  Respiratory
effects  on populations  from low level  exposures to ozone.
Presented at: 34th annual meeting of the Air  Pollution Control
Association; June; Atlanta, GA. Pittsburgh, PA: Air Pollution
Control Association; paper no. 83-12.5.

Lebowitz, M. D.  (1984) The effects of  environmental tobacco smoke
exposure and gas  stoves on daily peak  flow rates in asthmatic and
non-asthmatic families. Eur. J. Respir. Dis.  65(suppl. 133):  90-
97«

-------
                                 31
 Lebowitz, M. D.; Holberg, C. j. ; Boyer, B.; Hayes

 production and for estimating agricultural losses
                                               "
Lee, E. H . ; Tingey, D.  T. ,  and Hogsett, W.  E.  (1988c)
Interrelationship of  Experimental  Exposure  and Ambient  Air
Quality Data for Comparison of Ozone  Exposure
Lefohn  A. s.; Tingey, D. T. (1984) The co-occurrence of


-------
                                 32
 N  P' • Fif;L*U n T ^ V Spier' C- E'; Ble«sey, R. L. ; Jones,
 M. P., Fischer, D. A.; Hackney, J. D. (1978) Health effects of
 ozone exposure in asthmatics. Am. Rev. Respir. Dis. 117:  835-843.
         voETwM'vP*; Bachmayer' E- *•; Spier, C. E.; Mazur,
         ™1'*-1-; Hackney, a. D. (1980) Short-term respiratory
      OV^H  P°llute<* .fir:  a laboratory study of volunteers in a
     -oxidant community.  Am. Rev. Respir. Dis. 121: 243-252.

       W; S*; Medwav'  D- A-; Anzar, u. T. ; Valencia, L. M. ; Spier
      ; Tsao, F.  S-0. ; Fischer, D. A.; Hackney, J. D. (1982)
 Persistence of adaptation to ozone in volunteers exposed
 repeatedly over six weeks. An. Rev.  Respir. Dis, 1251 491-495.

 Linn,  W.  S.; Avol, E. L. ; Hackney, J. D. (1983a)  Effects of
 ambient oxidant pollutants on humans: a movable environmental
 chamber study  In: Lee, S. D. ; Mustafa,  M.  G. ; Mehlman, M.  A.,
 eds.   International symposium on the biomedical effects of ozone
 and related photochemical oxidants;  March 1982; Pinehurst  NC
 Princeton,  NJ:  Princeton Scientific  Publishers, Inc.;  pp. ' 125-
 137..  (Advances  in modern toxicology: v. 5) .

 Linn,  W.  S.; Shamoo,  D. A.;  Venet, T. G. ;  Spier,  C. E. ; Valencia,
 L.  M.;  Anzar,  U.  T. ;  Hackney,  J.  D.  (1983b)  Response to ozone in
 volunteers  with  chronic obstructive  pulmonary  disease.  Arch
 Environ.  Health  38: 278-283.

 Linn,  W  s.; Avol,  E.  L. ;  Shamoo,  D.  A.; Spier,  C.  E. ;  Valencia,
 L.  M.;  Venet, T.  G. ; Fischer,  D.  A.;  Hackney, J.  D.  (1986) A dose
 response  study of healthy heavily exercising mean exposed to
 ozone  at  concentrations near  the ambient air quality standard
 Toxicol.  Ind. Health  2:  99-112.

 Linn,  W.  S.;  Avol,  E.  L. ;  Shamoo,  D.  A.;  Peng,  R.-C.; Valencia,
 L.  M.;  Little, D.  E. ;  Hackney, J.  D.  (1988) Repeated laboratory
 ozone  exposures  of volunteer Los  Angeles residents:  apparent
 seasonal  variation in  response.   Toxicol. .Ind.  Health 4:  505-520.

 Lioy, P.  j.; Vollmuth,  T.  A.;  Lippmann,  M.  (1935) Persistence of
 peak flow decrement in children  following ozone exposures
 exceeding the National  Ambient Air Quality Standard. J. Air
 Pollut. Control Assoc.  35: 1068-1071.

 Lioy, P.  J.; Dyba, R.  V.  (1989) The dynamics of human exposure to
 tropospheric ozone.  In:   Schneider,  T. ; Lee, S.  D. ; Wolters,  G.
 J. R. ; Grant, L. D. , eds.  Atmospheric ozone research and  its
 policy implications:   Proceedings  of:  The Third  U.S. -Dutch
 international symposium; May 1988; Nigmegen, The  Netherlands.
Amsterdam, The Netherlands:  Elsevier Science Publishers; pp.
 711-721.  (Studies in  environmental science 35) .
dose-

-------
                              33
                                                       , D.;
 me.,.  »
                                         p-          to
                            '     "6
 ser.es:  43,.  Available from:



 SiS              - -

                                            by
                        wed^ Si       .
stress. Environ. Exp.Bot. 20: 99-iJS?      chronic air pollution



-------
                                34
 Mauderly,  J.  L.  (1984)  Respiratory function responses of animals
 and man to oxidant gases and with pulmonary emphysema. J.
 Toxicol.  Environ.  Health 13:  345-361.

 McAllen,  S. J.;  chiu,  S.  P.;  Phalen, R.  F.; Rasmussen, R. E.
 (1981)  Effect of in vivo ozone exposure  on in vitro pulmonary
 alveolar macrophage mobility.  J.  Toxicol.  Environ.  Health 7: 373-
 3 o J» •

 McCarthy,  E.  F.;  Stankunas,  A.  R.;  Yocum,  J.  E.,;  Rae,  D.  (1983)
 Damage  cost models for  pollutant  effects on material.   East
 Hartford,  CT:  TRC Environmental  Consultants,  Inc.;  TRC project
 no.  1574-J80.  U.S.  Environmental Protection Agency; EPA report
 no.  600/3-84-012.

 McClellan, R.  O.  (1989)  Letter from Chairman of Clean Air
 Scientific Advisory Committee  to  the EPA Administrator concerning
 "closure"  on  the Ozone  Criteria Document Supplement  and the Ozone
 Staff Paper,  dated May  1,  1989.

 McClenahen, J. R.  (1979)  Effects  of ethylene diurea  and ozone on
 the growth of  the  tree  seedlings.  Plant  Dis.  Rep. 63:  320-323.

 McCool, P. M.; Menge, J.  A.; Taylor, O.  C.  (1979) Effects of
 ozone and  HC1  gas  on the  development of  the mycorrhizal fungus
 Glomus  faciculatus and  growth  o of  'Troyer'  citrange.  J.  Am.  Soc.
 Hortic. Sci.  104:  151-154.

 McCool, P. M.; Musselman,  R. C.;  Teso, R.  R.;  Oshima,  R.  J.
 (1986)  Determining crop yield  losses from  air  pollutants.  Calif.
 Agric.  July-August:  9-10.

 McCool, P. M.; Musselman,  R. C.; Teso, R.  R.  (1987)  Air pollutant
 yield->loss assessment for  four  vegetable crops. Ag.  Ecosys.  and
 Environ. 20: 11-21.

 McCurdy, T. (1984)   Miscellaneous analyses of  alternative O3
 NAAQS formulations and  their impact on O3  non-attainment  and  cost
 analyses procedures.  Research  Triangle Park,  N.C.:  U.S.
 Environmental Protection Agency.

McCurdy, T. (1987)   Descriptive statistical analyses of ozone air
 quality indicators  in rural/remote  and metropolitan  areas.
Research Triangle  Park, N.C.: U.S.  Environmental Protection
Agency.

McCurdy, T.; Atherton, R.A.  (1988) Variability of ozone air
quality indicators in selected metropolitan statistical areas.
Research Triangle  Park, N.C.:  U.S. Environmental Protection
Agency.

-------
                   35
' I' n"*  ?°fstmann' D-  H. ; Hazucha,  M. J. ;  Seal, E
                                      t-j=r-
                                         -•
                            -




-------
                                 36

 McLaughlin, S. B.; McConathy, R. K.; Duvick, D,,; Mann  L  K
 (1982)  Effects of chronic air pollution stress on photosynthesis
 carbon allocation, and growth of white pine trees. For. Sci. 28:
 60—70.

 McLaughlin, S.B.  (1985)   Effects of air pollution on forest:  A
 critical review.   J.  Air Pollut. Assoc. 35: 512-534.

 Meiners, J. p.;  Heggestad,  H. E. (1979) Evaluation of snap bean
 cultivars for resistance to ambient oxidants in field plots and
 to ozone in chambers.  Plant Dis. Rep.  63:  273-277.

 Menser,  H.  A.; Street,  0.  E.  (1962)  Effects of air pollution
 nitrogen levels,  supplemental irrigation,  and plant spacing on
 weather  fleck and leaf losses of Maryland  tobacco. Tobacco 155:
 192—19 6.

 Menser,  H.  A.; Hodges,  G.  H.  (1972)  Oxidant injury to shade
 tobacco  cultivars developed in Connecticut for weather fleck
 resistance.  Agron.  J.  64:  189-192.

 Merz, T.; Bender,  M. A.; Kerr,  H. D.;  Kulle,  T.  J. (1975)
 Observations of aberrations in chromosomes of lymphocytes  from
 human subjects exposed  to  ozone at  a concentration of 0.5  ppm for
 6  and 10 hours. Mutat. Res.  31:  299-302.    -

 Miller,  P.  R.; Parmeter, J. R.,  Jr.; Taylor,  O.  C.;  Cardiff,  E.
 A.  (1963) Ozone injury to the foliage  of Pinus ponderosa
 Phytopathology 53:  1072-1076.

 Miller,  P.  R.; Parmeter, J. R.,  Jr.; Flick,  B.  H.; Martinez,  C.
 W.  (1969) Ozone dosage response of ponderosa  pine  seedlings.  J.
 Air Pollut.  Control Assoc.  19:  435^-438.

 Miller,  P. R.  (1973) Oxidant"induced community change in a mixed
 conifer  forest. In: Naegle, J.  A., ed.  Air pollution damage to
 vegetation.  Washington, DC: American Chemical  Society;  pp.  101-
 117. (Advances in Chemistry series no.  122).

Miller,  P. M.; Tomlinson, H. ; Taylor,  G. S.  (1976) Reducing
 severity of  ozone damage to tobacco and beans  by combining
benomyl  or carboxin with contactnematicides. Plant Dis. Rep.  60-
433-436.

Miller,  P. R.; Elderman, M. J.,  eds. (1977) Photochemical  oxidant
air pollutant effects on a mixed conifer forest ecosystem:  a
progress report, 1976. Corvallis, OR: U.S.   Environmental
Protection Agency; EPA report no. EPA-600/3-77-104.   Available
from:  NTIS,  Springfield, VA;  PB-274531.

-------
                                37
 lung to lower respiratory tract 2™???  of ozone in the human
 Toxicol. Appl. PharmacolT 7^11-??        a°d t0 exerci^-



                                          S^i-SS?  ^

                     8*^^
                                       2'  *'
 Toxicol. Lett. 2:    .-mcuced respiratory infections.
 bten manaabratoryaniS^fin'  D'  L"  (19?8b)
 ozone. Environ. Res? "7° 84^101       re9lon^ deposition of









 EPA report no.  EPA-eSo/D-SV/oJo    Effects Research Laboratory;
EPA report no.  EPA-600/D-87/1?!    Efe^s Research Laboratory;
     N.C.: u.S.
                                  of bonnet monkeys. Exp. Lung




                                                  National air

-------
                                38

 Monitoring and Data Analysis Division (MDAD,  1987)   Metropolitan
 statistical areas (MSA)  regulatory analysis air quality data
 base,  1983-1985."  Durham,  N.C.:  U.S. EPA.

 Mooi,  J.  (1980)  Influence of ozone on the growth of two ponlar
 cultivars.  Plant Dis.  64: 772-773.                  .

 Moore,  P.  F.;  Schwartz,  L.  W.  (1981)  Morphological  effects of
 prolonged  exposure to  ozone and sulfuric acid aerosol on the rat
 lung.  Exp.  Mol.  Pathol.  35:  108-123.

 Moore,  G.  S.;  Calabrese,  E.  J.; Schulz,  E.  (1931) Effect of in
 vivo ozone exposure to Dorset  sheep,  an  animal model with low"
 levels of  erythrocyte  glucose-6-phosphate dehydrogenase activity.
 Bull.  Environ.  Contam. Toxicol. 26: 273-280.

 Moorman, W.  J.;  Chmiel,  J.  J.;  Stara,  J.  F.;  Lewis,  T.  R.  (1973)
 Comparative decomposition of ozone  in the nasopharynx of beagles:
 acute  vs.  chronic exposure.  Arch. Environ.  Health 26:  153-155.

 Mueller, W.  J.;  Stickney, P. B. (1970)   A survey and economic
 assessment  of  the effects of air pollution  on elastomers.
 Columbus,  OH:   Battelle  Memorial Institute; NAPCA contract no.
 CPA-22-69-146.

 Mumford, R.  A.;  Lipke, H.;  Laufer,  D.  A.; Feder, W.  A.  (1972)
 Ozone  induced  changes  in  corn pollen.  Environ.  Sci.  Technol.  6-
 427-430.

 Musselman,  R.  C.;  Huerta, A. J.; McCool,  P. M.;  Oshima,  R.  J.
 (1986) Response  of beans  to  simulated  ambient and uniform  ozone
 distributions with- equal  peak concentrations.  J. Amer.  Soc.
 Horticult.  Sci.  ill: 470-473.

 Musselman,  R. C.;  McCool, P. M.; Younglove, T.  (1988)  Selecting
 ozone exposure statistics for determining crop  yield loss  from
 air pollutants.  Environ.  Pollut. 53:  63-78.

Musselman,  R. C.;  Oshima, R. J.; Gallavan, R.  E. (1983)
 Significance of  pollutant concentration distribution in  the
 response of  'red kidney1  beans to ozone. J. Am.  Soc. Hortic.  Sci.
 108: 347-351.

Mustafa, M. G.;  Tierney, D.  F.  (1978)  Biochemical and metabolic
changes in  the lung with  oxygen, ozone, and nitrogen dioxide
toxicity. Am. Rev. Respir. Dis. 118: 1061-1090.

National Research  Council.  (1977)  Toxicology.  In: Ozone  and other
photochemical oxidants. Washington, DC: National Academy of
Sciences, Committee on Medical and Biologic Effects  of
Environmental Pollutants; pp. 323-387.

-------
                                39
quality and N-fi
                         of
NC.
  Engl.  j.  Med.  291:  755-758.
 518-520.
                             NC
  Springfield, VA; PB-26423
of -one^
Soc. Hortic. Sci.103:  348-350
                                                   °n yield'
                                          from:  NTIS,
                                     Y°Ung  ci?arette  smokers. N
                                                         --
                                           fading.  Text. cheln.
                                               ecosystem.
                                   beet.  J.  Environ.  Qual.  2:
                                                       a:
                                             in  parsley,  j.  Am.

-------
                                 40
   97E                                 '  P. K-; Baldwin, D. W.
 (1975) Effect of ozone on the yield and plant biomass of a
 commercial variety of tomato. J. Environ. Qual. 4: 463-464.
Oshima, R  J. (1978) The impact of sulfur dioxide on vegetation:
a sulfur dioxide-ozone response model. Sacramento, CA: California
Air Resources Board; final report; agreement no. A6-162-30.

Overton, J. H. ;  Graham, R. c. ; Miller, F. J. (1987) A model of
the regional uptake of gaseous pollutants in the lung. II  The
sensitivity of ozone uptake in laboratory animal lungs to
aat
       o         ventilatory parameters.  Toxicol.  Appl.  Pharmacol.
 88 '.  418—432 .

 P'an,  A.  Y.  S.;  Beland,  J. ;  Jegier,  Z.  (1972)  Ozone-induced
 arterial  lesions.  Arch.  Environ.  Health 24:  229-232.

 Parry, M.  A..  J. ; Day,  W.  (1988) A comparison of  indices that
 describe  the  relationship between exposure to ozone and reduction
 in the yield  of  agricultural crops [Comments on  article by A S
 Lefohn, J.A.  Laurence  and R.J. Kohut] .  Atmos.  Environ.  22:  (in*
 press) .

 Patton, R. L.  (1981) Effects of ozone and sulfur dioxide on
 height and stem  specific  gravity  of  Populus  hybrids. Broomall
 PA: Northeastern Forest Experiment Station;  FSRP-NE-471
 Available  from: NTIS,  Springfield, VA;  PB81-17 1274.

 Paul,  R, A.;  Johnson,  T.;  Pope, A.;  Ferdo, A.; and  Biller,  W.  F.
 (1986) National estimates of exposure to ozone under alternative
 national standards. Durham,  N.C.:  PEI Associates, Inc.

 Pell,  E. J.;  Weissberger, W.  C. ;  Speroni, J. J.  (1980)  Impact of
 ozone  on quantity and  quality of  greenhouse-grown potato plants
 Environ. Sci. Technol. 14: 568-571.

 Peterson,  D.C.; Sueker, J.K.; Wallsten, T.S.  (1987).  Risks of
 Forest Tree Response Due  to  Ambient Ozone (Draft Final  Report) .
 Energy and Resource Consultants,  Inc. for U.S., Environmental
 Protection Agency.  Research  Triangle Park, NC,.

 Peterson, D.C. and D. Violette.   1986.  Dimensioning Uncertainty
 in Estimates  of Regional  Fish Population Damage Caused  by
Acidification in Adirondak Ponded Waters, U.S. EPA, 200-12-86-
 022.

Peterson, D.C. and J.K. Sueker.  1987.  "Risks of forest response
due to ambient ozone," APCA  87-36.3.

-------
                               41
  Environ. Health 6
                    978lO

P. 11.                   '


environmental science 35)
                                         (Studies in
Pollution centre   sSociation;  paper°no' S?^
                                                   2SS-
                                                   '  PA= Air
                                                29: 207-214.

-------
                                 42

 Raub,  J. A.; Miller, F. J.; Graham, J. A. (1983) Effects of low-
 level  ozone exposure on pulmonary function in adult and neonatal
 rats.  In: Lee,S.  D.; Mustafa,  M. G.; Mehlman, M. A., eds.
 International symposium on the biomedical effects of ozone and
 related photochemical oxidants; March 1982;  Pinehurst  NC
 Princeton,  NJ:  Princeton Scientific Publishers,  Inc.; pp."363-
 367.   (Advances in modern environmental toxicology: v. 5).

 Raven,  P. B.,  Drinkwater,  B.  L.; Horvath, S.  M.; Ruhling, R. o.;
 Gliner,  J.A.;  Sutton,  J.  C.;  Bolduan,  N.  W.  (1974a) Age, smoking
 habits,  heat stress,  and their interactive effects with carbon
 monoxide and peroxyacetyl nitrate on man's aerobic power   Int
 J.  Biometeorol.  18: 222-232.

 Raven,  P. B.,  Drinkwater,  B.  L.; Ruhling, R.  o.; Bolduan, N •
 Taguchi,  S.; Gliner,  J.  A.; Horvath,  S. M.  (1974b)  Effect of
 carbon  monoxide and peroxyacetyl nitrate  on man's maximal aerobic
 capacity. J. Appl.  Physiol.  36: 288-293.

 Rawlings, J. o.;  Lesser,  V. M.; Heagle, A. S.; Heck W.  W. (1988)
 Alternative  ozone dose metrics  to characterize ozone impact on
 crop yield loss.  J.  Environ. Qual.  17:  285-291.

 Rawlings, J. O.;  Cure,  W.  W.  (1985)  The Weibull  function as a
 doseresponse model  to  describe  ozone effects  on  crop yields. Crop
 Science  (in  press).

 Reich, P. B.  (1983)  Effects of  low  concentrations of 0,  on  net
 photosynthesis, dark respiration, and  chlorophyll contents  in
 aging hybrid poplar leaves. Plant Physiol. 73:: 291-296.

 Reich, P. B.; Amundson, R.  G.  (1986) Ambient  levels of  ozone
 reduce net photosynthesis  in tree and  crop species.  Science 230:
 566—570.

 Reich, P. B.; Amundson, R.  G.  (1984) Low  level 03 and/or  SO,
 exposure  causes a linear decline in  soybean yield.  Environ.
 Pollut.  (Series A)  34: 345-355.

 Reinert,  R.  A.; Nelson, P.  V.  (1979) Sensitivity and growth of
 twelve elatior begonia cultivars  to ozone. Hort  Science  14:  747-
 748.

Reinert,  R. A.; Gray, T. N..(1981) The response  of  radish to
 nitrogen  dioxide, sulfur dioxide, and ozone,  alone  and  in
 combination. J. Environ. Qual.  10: 240-243.

Reisenauer,  C. S.; Koenig, J. Q.; McManus, M. S.; Smith,  M. S.;
Kusic,  G.; Pierson, W. E.  (1988) Pulmonary responses  to ozone
exposures in healthy individuals aged 55 years or greater.  J.  Air
Pollut. Control Assoc. 38:  51-55.

-------
                              43
                                       2
 Appl. Pharmacol? sJT jJJ-L      9 C°llagen of m°^eys. Toxicol
 Richards, B.  L. ; Middleton, J. T. ; Hewitt  w
                                 "'
                  913-917.
Atnosphirio                      pc          ...




environmental science 35).                   (Studies in


                                  a sss


-------
                                 44
 !*"d?f!4 
-------
                               45
Shannon .  j .  G . :  Mu 1 r h i   r  T   / 1 n •? ^ \  ^
varieties anthisis  crip sci!  1" ^5^ daMge tO Wheat
Sherwin,  R.  P.; Richters, V. (1985)  Effect
                      S"~
      J- M'  (1980) Photochemical oxidant  impact
                                               on
                                               °


-------
                                 46

 Skelly,  J.  M.; Yaw-Shing Yang; Chevone, B. J.; Long, S. J. (1984)
 Ozone concentrations and their influence on forest species in the
 Blue Ridge  Mountains of Virginia.  In:  Davis, D. D.; Millen, A.A.;
 Dochinger,  L., eds.  Air pollution and the productivity of the
 forest:  proceedings  of the symposium;  October 1983; Washington
 D.C.   Arlington,  VA: Izaak Walton League of America; pp.143-159.

 Slade, R.;  Stead,  A. G.; Graham, J.  A.; Hatch, G.  E. (1985)
 Comparison  of  lung antioxidant levels  in humans and laboratory
 animals.  Am. Rev.  Respir.  Dis. 131:  742-746.

 Smith, G.;  Greenhalgh,  B.;  Brennan,  E.; Justin, J. (1987)  Soybean
 yield in  New Jersey  relative to ozone  pollution and antioxidant
 application. Plant Dis.  71:  121-125.

 Solic, J. J.;  Hazucha,  M.  J.;  Bromberg,  P.  A.  (1982)  The acute
 effects of  0.2 ppm ozone in  patients with chronic  obstructive
 pulmonary disease. Am.  Rev.  Respir.  Dis.  125:  664-669.

 Spektor,  D. M.; Lippmann,  M. ;  Lioy,  p.  J.;  Thurston,  G. D.;
 Citak, K.;  James,  D.  J.;  Bock,  N.; Speizer,  F.  E.; Hayes,  C.
 (1988a) Effects of ambient ozone on  respiratory function in
 active normal  children.  Am.  Rev. Respir.  Dis.  137: 313-320.

 Spektor,  S. M.; Lippmann, M.;  Thurston,  G.  D.;  Lioy,  P. J.;
 Stecko, J.; O'Connor, G.;  Garshick, E.;  Speizer, F.  E.; Hayes,  C.
 (1988b) Effects of ambient ozone on respiratory function in
 healthy adults exercising  outdoors.  Am.  Rev.  Respir. Dis.  138:
 821-828.

 Stark, R. w.;  Miller, P. R.; Cobb, D. L.,  Jr.;  Wood,  D. L.;
 Parmeter, J. R., Jr.(1968) Photochemical  oxidant injury and bark
 beetle (Coleoptera:  Scolyticlea)  infestation  of  ponderosa pine:  I.
 Incidence of bark  beetle infestation of ponderosa  pine: I.
 Incidence of bark  butter infestation in  injured trees.   Hilqardia
 39: 121-126.

 Stark, R. W.;  Cobb,  F. W., Jr.  (1969) Smog  injury,  root diseases
 and bark beetle damage in ponderosa pine.  Calif. Agric. 23: 13-
 JLO •

 Stephens, R. J.; Sloan, M. F.;  Evans, M.  J.; Freeman, G.  (1974)
 Early response of  lung to low  levels of ozone.  Am.  J. Pathol.  74-
 31-58.

 Stephens, R. J.; Sloan, M. F.;  Groth, D.  G.  (1976)  Effects  of
 long-term, low level exposure of NO2 or O3 on rat lungs. EHP
 Environ.  Health Perspect. 16: 178-179.

 Stephens, R. J.; Sloan, M. F.;  Groth, D.  G.; Negi,  D. S.; Lunan,
K. D. (1978)  Cytologic responses of postnatal  rat lungs to 0, or
N02 exposure.  Am.  J.  Pathol. 93: 183-200.

-------
                                  47
the biomedical effects of ozone and reStS"^1?11^ sYmP°sium °*
ovirJan+'c.. M,—i. ,««_  _T  «*«-»*'« ana related  photochemi r?ai
  the  biomedical' effec   of ozone SS'reSSTSh1?"^ 8Y»Po«iu»
  oxidants;  March 1982-  Pin^h,,?=J  SS r®1?ted Photochemical

  Scientific Publisher^   inc  
-------
                                 48
 Thompson, C. R. ; Kats, G. ; Cameron, J. w.  (1976) Effects of
 ambient photochemical air pollutants on growth, yield, and ear
 character of two sweet corn hybrids. J. Environ. Qua 1. 5? 4?0-
 4 Xb e

 Thurlbeck, W. M. (1984) The pathobiology .and epidemiology of
 human emphysema.  In:  Miller, F.J. and MenzelT D.B., eds.
 SS?SI2S 5 ?f . e^rapolation modeling of inhaled toxicants.
 Hemisphere Publishing Corporation; pp. 323-343.

 Tice, R. R.; Bender, M. A.; Ivett, J.  L. ; Drew, R. T. (1978)
 Cytogenetic effects of inhaled ozone.  Mutat. Res. 58: 293-304.

 Tingey,  D. T. ;  Wilhour, R.  G. ; Standley,  E. (1976) The effect of
 chronic  ozone exposures on the metabolite content of ponderosa
 pine seedlings.  For. Sci.  22:  234-241.                ponaerosa

 Tingey,  D. T.;  Reinert, R.  A.; Carter,  H.  B. C1972)  Soybean
 cultivars: acute -foliar response to ozone.  Crop Sci.  12:  268-270.
 7™ «??'«?'  ?"'  ?hUtt: f*  L:;  GumPertz'  M-  !••;  Hogsett, W. E.
 (1982)  Plant water status influences ozone sensitivity of bean
 plants.  Agric.  Environ.  7:  243-254.

 Tingey,  D.  T. ;  Taylor, G.  E. ,  Jr.  (1982) Variation in plant
 response to ozone:  a conceptual  model of physiological events.
 In:  Unsworth, M.  H. ; Ormrod,  D.  P.,  eds. Effects of gaseous air
 pollution in agriculture  and  horticulture.  London,  United
 Kingdom  Butterworth Scientific; pp.  113-138.

 Tingey,  p.  T. ;  Reinert, R.  A.  (1975)  The effect  of ozone and
 sulfur dioxide  singly and  in  combination on plant growth
 Environ.  Pollut.  9:  117-125.

 Tingey,  D.  T.;  Hogsett, W.  E. ; Lee,  E.  H.  (1989)  Analysis of crop
 loss for alternative ozone  exposure  indices. In:  Atmospheric
 Ozone Research  and  its Policy  Implications.  T. Schneider  G
 Wolters,  s.  Lee and  L. Grant  (eds.).  Elsevier  Science Publishers
 B.V., Amsterdam,  The Netherlands.   (Studies in environmental
 science  35). pp.  219-227.

 Tingey,  D.  T.;  HecJc,  W. W. ; Reinert,  R. A.  (1971)  Effect of low
 concentrations  of ozone and sulfur dioxide  on  foliage,  growth and
 yield of radish. J.  Am.Soc. Hortic. Sci. 96: 369-371.

 Tingey, D.  T. (1974)  Ozone-induced alterations in the metabolite
pools and enzyme activities of plants.  In:  Dugger, M. ,  ed.  Air
pollution effects on  plant growth.  Washington, DC: American
Chemical Society; pp. 40-57.

-------
                                 49

 Toivonen, P. M. A.; Hofstra, G.; Wukasch, R. T. (1982) Assessment
 of yield losses in white bean due to ozone using the antioxidant
 EDU.  Can. J. Plant Path. 4: 381-386.

 Tonneijck,  A. E. G. (1984)  Effects of peroxyacetyl nitrate (PAN)
 and ozone on some plant species. In: Proceedings of the OECD
 workshop on ozone; Gothenburg, Sweden.

 Treshow, M. (1980) Pollution effects on plant distribution.
 Environ. Conserv.  7:  279-286.

 Treshow, M.; Stewart,  D. (1973) Ozone sensitivity of plants in
 natural communities.  Biol.  Conserv.  5: 205-214.

 Tyler,  W. S.; Tyler,  N.  K.; Last,  J.A.; Gillespie,  M.J.; Barstow,
 T.J.  (1988) Comparison of daily and  seasonal exposures of young
 monkeys to  ozone.  Toxicology 50:  131-144.

 Tyson,  C. A.; Lunan,  K.  D.; Stephens,  R.  J.  (1982)  Age-related
 differences in GSH-shuttle  enzymes in NO 2-  or O3-exposed rat
 lungs.  Arch.  Environ.  Health 37:  167-176.

 U.S.  Court  of Appeals  for the  District of Columbia,  American
 Petroleum Institute vs.  EPA. No.  79-1104  and 79-1201,  September
 1981.

 U.S.  Court  of Appeals  for the  District of Columbia  Circuit, Lead
 Industries  Association.  Inc. vs.  EPAf  "No. 78-2201 and  78-2220.
 June  27,  1980.

 U.S.  Department of Health and  Human  Services (1981)  Current
 estimates from the National Health Interview Survey: United
 States,  1979.  Hyattsville,  MD:  Public  Health Service,  Office  of
 Health  Research, Statistics and Technology,  National Center for
 Health  Statistics;  DHHS  publication  no.  (PHS)  81-1564.  (Vital and
 health  statistics:  series 10,  no. 136).

 U.S.  Environmental  Protection  Agency (1978)  Air quality  criteria
 for ozone and  photochemical oxidants.  Research  Triangle  Park,  NC:
 U.S.  Environmental  Protection  Agency,  Environmental Criteria  and*
 Assessment  Office;  EPA-600/8-78-004.   Available from:  NTIS
 Springfield, VA; PB80-124753.

 U.S. Environmental  Protection  Agency (1980)  Air quality  data—
 1979 annual statistics including summaries with reference to
 standards. Research Triangle Park, NC: Office of Air Quality
Planning  and Standards;  EPA report no. EPA-450/4-80-014.
Available from: NTIS,  Springfield, VA; PB81-123093.

-------
                                50

 U.S.  Environmental Protection Agency (1981)  Air quality data—
 1980  annual statistics including  summaries with reference to
 standards.  Research Triangle Park,  NC:  Monitoring and Data
 Analysis Division; EPA report no.  EPA-450/4-81--027.   Available
 from:   NTIS,  Springfield,  VA; PB82-106501.

 U.S.  Environmental Protection Agency (1982a)  Air quality criteria
 for oxides  of nitrogen.  Research  Triangle Park,  NC:  U.S.
 Environmental Protection Agency,  Environmental Criteria and
 Assessment  Office; EPA report no.  EPA-600/8-82--026F.   Available
 from:  NTIS,  Springfield, VA;  PB83-163337.

 U.S.  Environmental Protection Agency (1982b)   Air quality
 criteria for  particulate matter and sulfur  oxides.   Research
 Triangle Park, NC:  U.S.  Environmental Protection Agency; EPA
 report no.  EPA-600/8-82-029a,b,c,d.

 U.S.  Environmental Protection Agency (1986)   Air quality criteria
 for ozone and other photochemical  oxidants.   Draft Final.
 Environmental Criteria and Assessment Office;  EPA Report No.  EPA-
 600/8-84/020a to  020e.   Available  from: NTIS,  Springfield,  VA-  PB
 87-142949.

 U.S.  Environmental Protection Agency (1988) Summary  of Selected
 New Information on Effects of Ozone  on Health and Vegetation:
 Draft  Supplement  to Air  Quality Criteria  for  Ozone and Other
 Photochemical Oxidants.  Research Triangle Park,  NC:  Office of
 Health and  Environmental Assessment, Environmental Criteria and
 Assessment .Office;  EPA/600/8-88/105A. Available  from  NTIS
 Springfield,  VA.

 U.S. Senate  (1979)   93rd Congress, 2nd Session Committee on
 Environment and Public Works  (Comm.  Print 1974).  A  legislative
 history  of the clean air act  amendments of 1970,  Vol.  l, p.  410.

 Upham, J. B.; Haynie,  F. H.;  Spence, J. W.  (1976)  Fading of
 selected  drapery  fabrics by air pollutants.   J.  Air Pollut.
 Control Assoc. 26:  790-792.

 van Bree, L.; Rombout  , P. J. A.; Rietjens, I. M. C.  M.; Dormans,
 J. A.  M.  A.; Marra,  M.  (1989) Pathobiochemical effects  in rat
 lung related  to episodic ozone exposure.  In:  Schneider, T.;
 lee, S. D.; Wolters, G. J. R.; Grant, L. D.,  eds.  Atmospheric
 ozone research and  its policy implications:  Proceedings of:  The
Third U.S.-Dutch  international symposium; May  1988; Nijmegen, The
Netherlands.  Amsterdam, The Netherlands:  Elsevier Science
Publishers; pp. 723-732.   (Studies in environmental science  35).

Veith, A. G.; Evans, R. L.  (1980)  Effect of  atmospheric
pressure  on ozone cracking of rubber.  Polym.  Testing 1: 27-38.

-------
51


                                                   on
                       adhesion
      function and
                 Assessment of
 Toxiool.  Appl.  Phar.        «9H137
                    '   -   "
          Sprague-Dawley  Rats
Wilhour, R. G.; Neely  G  E   (1Q77)

^ternatLnl/confe^encrS^photochemi J?^^*?lade8' °*"*  «*•
its control: proceedings, vol. "-. ^-_.   ^ox_ ant Pollution  and
Park, NC.  U.S. Environmental*
report no.  EPA-600/3-77-ooib
Springfield, VA; PB-264233
            :  NTIS,
                   635-645; EPA

-------
                                 52

 Witz  G.; Amoruso, M. A.; Goldstein, B. D. (1983) Effect of ozone
 on alveolar macrophage function: membrane dynamic properties. In-
 Lee, S.  D.; Mustafa,M. G.; Mehlman, M. A., eds. International
 symposium on the biomedical effects of ozone and related
 photochemical oxidants; March 1982; Pinehurst, NC.  Princeton
 NJ:  Princeton Scientific Publishers, Inc,; pp. 263-272. (AdvaAces
 in modern enviromental toxicology:  v.  5).

 Woodwell, G.  M.  (1970) Effects of pollution on the structure and
 physiology of ecosystems. Science (Washington, DC)- 168: 429-433.

 Woodwell, G.  M.  (1974) Success,  succession,  and Adam Smith.
 BioScience 24:  81-87.

 World Health  Organization (1977)  Manual of the international
 statistical classification of diseases,  injuries, and causes of
 death. Geneva,  Switzerland:  World Health  Organization.

 Wright,  J.  L.;  Lawson, L.  M.;  Pare,  P.  D.;  Wiggs, B.'J.;  Kennedy,
 S.;  Hogg,  J.C.  (1983)  Morphology of peripheral airways  in current
 smokers  and ex-smokers.  Am.  Rev.  Respir.  Dis.  127:  474-477.

 Wright,  E.  S.; Kehrer,  J.  P.;  White, D. M.;  Smiler,  K.  L.  (1988a)
 Effects  of  chronic exposure  to ozone on collagen in rat lunq
 Toxicol.  Appl. Pharm.  92:  445-452.

 Wukasch,  R. T.; Hofstra,  G.' (I977a)  Ozone  and Botrvtis
 interactions  in onion leaf dieback:  open-top  chamber studies
 Phytopathology 67:  1080-1084.

 Wukasch,  R. T.; Hofstra,  G.  (1977b)  Ozone and  Botrvtis  spp.
 interaction in onion  leaf  dieback:  field  studies.  J.  Am.  Soc
 Hortic.  Sci.  102:  543-546.

 Yang, Y.S.; Skelly, J. M.; Chevone,  B. I.; Birch,  J.  B.  (1983)
 Effects  of  long-term  ozone exposure  on photosynthesis and dark
 respiration of eastern white pine. Environ. Sci.  Technol. 17-
 371-373.

 Yokoyama, E.; Frank, R.  (1972) Respiratory uptake of  ozone  in
 dogs. Arch. Environ. Health 25: 132-138.

 Zelac, R. E.; Cromroy, H.  L.;  Bolch, W. E., Jr.;  Dunavant,  B. G.;
 Bevis, H. A.  (1971b) Inhaled ozone as a mutageri.  II.  Effect on
 the frequency of chromosome aberrations observed  in irradiated
 Chinese hamsters.  Environ. Res. 4: 325-342.

 Zelac, R. E.; Cromroy, H.  L.;  Bolch, W. E., Jr.;  Dunavant,  B. G.;
Bevis, H. A.  (I97la) Inhaled ozone as a mutagen.  I. Chromosome
aberrations induced in Chinese hamster lymphocytes. Environ.  Res.
 4: 262-282.

-------
                          53
-376.                      .  .   nviron. Pathol. Toxicol

-------

-------
                                      TECHNICAL REPORT DATA
                               .Please read Instructions on the reverse be/ore completing)
   1. REPORT NO.
         EPA-450/a-92-OQl
  [<*. TITLE AND SUBTITLE

     Review of  the National Ambient Air Quality  Standards
     tor Ozone-Assessment of  Scientific and Technical
     Information:   QAQPS Staff  Paper
  7. AUTHORIS)                       —	
      McKee, D. J.;  Johnson, P. M.;  McCurdy, T  M  •
      Richmond, H.  M.                              ''
  9. PERFORMING ORGANIZATION NAME AND ADDRESS	'	
     Office of Air  and  Radiation
     Office of Air  Quality Planning  and  Standards
     u.  S.  Environmental Protection  Agency
     Research Triangle  Park,  NC 27711
  12. SPONSORING AGENCY NAME AND ADDRESS '
  15. SUPPLEMENTARY NOTES
                                                            5. REPORT DATE
                                                               June 1989
                                                            6. PERFORMING ORGANIZATION CODE
                                                           8. PERFORMING ORGANIZATION REPORT NO.
                                                            10. PROGRAM ELEMENT NO?
                                                           11. CONTRACT/GRANT NO.
                                                               3. TYPE OF REPORT AND PERIOD COVERED
                                                                  Final
                                                               4. SPONSORING AGENCY CODE
                                                   scientific and  technical information

 ambient air quality  standardTfor^onrasUwen IT^ °5 ****!***  ^^  nati°nal
 ozone.   This staff paper provides staff  eon ?      secondary (welfare)  standards for






oxidants;  (2)  The current  fora of the ozone      con"ntrations  of photochemical
standard  is  attained when  expected number"of''davs/vfar8!!?^  bVeta*ned  (i>e-'
concentrations  above the level of the standard  HT     maximum hourly
1-hour standards  should be retained  althh   ls.e^ual or greater than one;  (3)  Th
-future development of alte^e fl^f b^th^p^im^rfa^"1^ ^^ b' S^
        range of  consideration for orimarv c;fanHa^^   u  TJ v  n
        r    \   j  f              *•***•  t'j.-Lujdry scanaard should be 0
        ^ppm; and  for welfare  atanrfarrf .^ range should be 0>Q6
                                                                                        the
                                                                                         The
17.
                  DESCRIPTORS
                                 KEY WORDS AND DOCUMENT ANALYSIS
          Ozone
          Photochemcal  Oxidants
          Air Pollution
          Health Effects
          Welfare Effects
18. DISTRIBUTION STATEMENT


          Release  to Public


   Form 2220-1 (R.v. 4-77)   PREVIOUS EDITION is OBSOLETE
                                               b.lDENTIFIERS/OPEN ENDED TERMS
                                                Air Quality  Standards
                                            19. SECURITY CLASS (Tins Report,
                                              Unclassified
                                                                           c. COSATI Field/Group
                                            20 SECURITY CLASS (This pagei
21 NO. OF PAGES
                                                                       22 PRICE

-------
                                                        INSTRUCTIONS

  1    REPORT NUMBER
      Insert the 1 PA report number as it appears on the cover ot the publication.

  2.   LEAVE BLANK

  3.   RECIPIENTS ACCESSION NUMBER
      Reserved for use by each report recipient.

  4.   TITLE AND SUBTITLE
      Title snould indicate clearly and briefly the subject coverage ol the report, and be displayed promiucniK  .VI siilniile. i! used. ,n smaller
      type or otherwise subordinate U to mam  title.  When a report is prepared in more than one volume, rcpr.it il:c prmi.irv  Mile, add volume
      number and include subtitle for the specific title.

  5.   REPORT DATE
      Each report shall carry a date indicating at least month and year. Indicate Hie lusis on which ii «.is selected /<• v . date <>/ mi/c, dale of
      approval, date of preparation, etc.).

  6.   PERFORMING ORGANIZATION CODE
      Leave blank.

  7.   AUTHOR(S)
      Give name(s)  in conventional order (John R. Doc. J  Robert Doc. etc I  List author's allilialion il it fillers irr.in the pertofinint! organi-
      zation.

  8.   PERFORMING ORGANIZATION REPORT NUMBER
      Insert if performing organization wishes to assign (his number

  9.   PERFORMING ORGANIZATION NAME AND ADDRESS
      Give name, street, city, state, and ZIP code.  List no more than two levels ol an orujiii/juunjl lurearUiv

  10.  PROGRAM ELEMENT  NUMBER
      Use the program element number under which the report *as ore-pared.  Subordmalc numbers cn.i> be indiuud  m p.iieniltcscs

  11.  CONTRACT/GRANT NUMBER
      Insert contract or giant  number under which report was prepaied.

  12.  SPONSORING AGENCY NAME AND ADDRESS
      Include ZIP code.

  13.  TYPE OF  REPORT AND PERIOD COVERED
      Indicate interim final, etc., and if applicable, dates covered.

  14.  SPONSORING AGtNCY CODE
      Insert appropriate code.

  15.  SUPPLEMENTARY NOTES                                                          ,,,„,,.
      Enter information not included elsewhere but useful, such as  Prepared  m cooperation with. Ir.inslation <•!. 1 rc-suitci! .it COHIUCIU.C »i
      To be published in, Supersedes, Supplements, etc.


      Include a brief (200 words or lessl factual summarv  ot the mo>t Mgnihcant mtormaimn contained in ilu upori- II  ilie icpoi i loiiums j
      significant bibliography or  literature survey, mention it here.

  17   KEY WORDS AND DOCUMENT  ANALYSIS
       (a) DESCRIPTORS • Select from the Thesaurus of Engineering and Suentilic lerms ihc proper julluiri/cU terms Hui identity the nuiur
      concept of the research  and are sufficiently specitic and precise to be used js index entries tor cataloging

      (b) 1DENTITIERS AND OPEN-ENDED  TERMS - Use identities for project names, code names, equipment  designators, etc  L se open-
      ended terms written in  descriptor form for those subjects for which no descriptor exists

       (ciCOSATl 1 1ILDGROLP- 1 icld and  group assignments are to be taken trom the 1965 ( OSA I I Sulnect ( JUT.TS I ist. Since the ma-
      jority of documents are multidisciplinary in nature, the Primary  I itld/Gioup assignments w,ll be MXXIIK d,s,,plmc  .-tea ol  mman
       endeavor, or type of physical object. The applications) will be cross-relerenced with sex umlary I leUI '< -r.u.p -issiuiiincnts that « ill K.iinw.
       the primary postine(s)

  18.  DISTRIBUTION STATEMENT                                                                        ,.             .i.u.iiu',,
       Denote releasabtlin lo  the public  or umiiation lor reasons other than sccuritv tor example Relejs,. \ ninmKii.    n, .m\ jvni.ini.nv
       the public, with address and price.

  19. & 20. SECURITY CLASSIFICATION
       DO NOT  submit classified reports to tnc National Technical Information service


       Insert the  total number ot pages, including this one and unnumbered pages, hut exclude distribution list, n any

  22   PRICE
       Insert the price set by the  National  technical  Intormation S,r\^e ur the Government I'rmtmg Ullu e n known
EPA Form 2220-1 iRev.  4-77) (Reverse.

-------