vvEPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/600/AP-93/004b
December 1993
External Review Draft
Air Quality
Criteria for
Ozone and
Related
Photochemical
Oxidants
Review
Draft
(Do Not
Cite or
Quote)
Volume II of
Notice
This document is a preliminary draft. It has not been formally
released by EPA and should not at this stage be construed to
represent Agency policy. It is being circulated for comment on its
technical accuracy and policy implications.
-------
DRAFT-DO NOT QUOTE OR CITE EPA/600/AP-93/0041
December 1993
External Review Oral
Air Quality Criteria for Ozone
and Related Photochemical Oxidants
Volume II of III
NOTICE
This document is a preliminary draft. It has not been formally
released by EPA and should not at this stage be construed to
represent Agency policy. It is being circulated for comment on
its technical accuracy and policy implications.
U.S. Environmental Protection y
•fgion 5, Library (PL-12J)
«..vVesi Jfckson Boulevard, 12i
Chicago, JL 60604-3590
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Printed on Recycled Paper
-------
DISCLAIMER
This document is an external draft for review purposes only and does not constitute
U.S. Environmental Protection Agency policy. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
December 19P3 H-ii DRAFT-DO NOT QUOTE OR CITE
-------
PREFACE
In 1971, the U.S. Environmental Protection Agency (EPA) promulgated National
Ambient Air Quality Standards (NAAQS) to protect the public health and welfare from
adverse effects of photochemical oxidants. In 1979, the chemical designation of the
standards was changed from photochemical oxidants to ozone (O3). This document,
therefore, focuses primarily on the scientific air quality criteria for O3 and, to a lesser extent,
for other photochemical oxidants like hydrogen peroxide and the peroxyacyl nitrates.
The EPA promulgates the NAAQS on the basis of scientific information contained in
air quality criteria documents. The previous O3 criteria document, Air Quality Criteria for
Ozone and Other Photochemical Oxidants, was released in August 1986 and a supplement,
Summary of Selected New Information on Effects of Ozone on Health and Vegetation, was
released in January 1992. These documents were used as the basis for a March 1993
decision by EPA that revision of the existing 1-h NAAQS for O3 was not appropriate at that
time. That decision, however, did not take into account some of the newer scientific data
that became available after completion of the 1986 criteria document. The purpose of this
revised air quality criteria document for O3 and related photochemical oxidants is to critically
evaluate and assess the latest scientific data associated with exposure to the concentrations of
these pollutants found in ambient air. Emphasis is placed on the presentation of health and
environmental effects data; however, other scientific data are presented and evaluated in
order to provide a better understanding of the nature, sources, distribution, measurement,
and concentrations of O3 and related photochemical oxidants and their precursors in the
environment. Although the document is not intended to be an exhaustive literature review, it
is intended to cover all pertinent literature available through the end of 1993.
This document was prepared and peer reviewed by experts from various state and
Federal governmental offices, academia, and private industry for use by EPA to support
decision making regarding potential risks to public health and welfare. The Environmental
Criteria and Assessment Office of EPA's Office of Health and Environmental Assessment
acknowledges with appreciation the contributions provided by these authors and reviewers as
well as the diligence of its staff and contractors in the preparation of this document at the
request of the Office of Air Quality Planning and Standards.
December 1993 H-iii DRAFT-DO NOT QUOTE OR CITE
-------
Air Quality Criteria for Ozone
and Other Photochemical Oxidants
TABLE OF CONTENTS
Volume I
1. EXECUTIVE SUMMARY 1-1
2. INTRODUCTION 2-1
3. TROPOSPHERIC OZONE AND ITS PRECURSORS 3-1
4. ENVIRONMENTAL CONCENTRATIONS, PATTERNS, AND
EXPOSURE ESTIMATES 4-1
Volume n
5. ENVIRONMENTAL EFFECTS OF OZONE AND
RELATED PHOTOCHEMICAL OXIDANTS 5-1
APPENDIX 5A: COLLOQUIAL AND LATIN NAMES 5A-1
Volume HI
6. TOXICOLOGICAL EFFECTS OF OZONE AND RELATED
PHOTOCHEMICAL OXIDANTS 6-1
7. HUMAN HEALTH EFFECTS OF OZONE AND RELATED
PHOTOCHEMICAL OXIDANTS 7-1
8. EXTRAPOLATION OF ANIMAL TOXICOLOGICAL DATA
TO HUMANS 8-1
9. INTEGRATIVE SUMMARY OF OZONE HEALTH EFFECTS 9-1
APPENDIX A: GLOSSARY OF TERMS AND SYMBOLS A-l
December 1993 H-v DRAFT-DO NOT QUOTE OR CITE
-------
TABLE OF CONTENTS
Page
LIST OF TABLES Il-xi
LIST OF FIGURES II-xv
AUTHORS, CONTRIBUTORS, AND REVIEWERS H-xix
U.S. ENVIRONMENTAL PROTECTION AGENCY PROJECT TEAM
FOR DEVELOPMENT OF AIR QUALITY CRITERIA FOR OZONE
AND RELATED PHOTOCHEMICAL OXIDANTS H-xxiii
5. ENVIRONMENTAL EFFECTS OF OZONE AND RELATED
PHOTOCHEMICAL OXIDANTS 5-1
5.1 INTRODUCTION 5-1
5.2 METHODOLOGIES USED IN VEGETATION RESEARCH ... 5-6
5.2.1 Fumigation Systems 5-6
5.2.1.1 Methodologies Discussed in the Air Quality
Criteria for Ozone and Other Photochemical
Oxidants (1986) 5-7
5.2.1.2 Methodologies Referenced Since the Air
Quality Criteria for Ozone and Other
Photochemical Oxidants (1986) 5-9
5.2.1.3 Flux Measurement 5-10
5.2.2 Experimental Design and Data Analysis 5-16
5.2.3 Mechanistic Process Models 5-19
5.3 SPECIES RESPONSE/MODE OF ACTION 5-20
5.3.1 Introduction 5-20
5.3.2 Ozone Uptake 5-22
5.3.2.1 Ozone Uptake by Plant Canopies 5-22
5.3.2.2 Ozone Absorption by Leaves 5-25
5.3.3 Resistance Mechanisms 5-29
5.3.3.1 Stomatal Limitation 5-29
5.3.3.2 Detoxification 5-30
5.3.4 Physiological Effects of Ozone 5-31
5.3.4.1 Carbohydrate Production and Allocation 5-35
5.3.4.2 Compensation 5-37
5.3.5 Role of Age and Size Influencing Response to Ozone . . 5-38
5.3.5.1 Summary 5-41
5.4 FACTORS THAT MODIFY PLANT RESPONSE 5-43
5.4.1 Modification of Functional and Growth Responses .... 5-43
5.4.2 Genetics 5-44
5.4.3 Environmental Biological Factors 5-57
5.4.3.1 Oxidant-Plant-Insect Interactions 5-59
5.4.3.2 Oxidant-Plant-Pathogen Interactions 5-63
5.4.3.3 Oxidant-Plant-Symbiont Interactions 5-68
December 1993 H-vii DRAFT-DO NOT QUOTE OR CITE
-------
TABLE OF CONTENTS (cont'd)
Page
5.4.3.4 Oxidant-Plant-Plant Interactions-
Competition 5-69
5.4.4 Physical Factors 5-71
5.4.4.1 Light 5-71
5.4.4.2 Temperature 5-73
5.4.4.3 Humidity and Surface Wetness .......... 5-75
5.4.4.4 Drought and Salinity .,....„„,.„„ 5-76
5.4.5 Nutritional Factors 5-82
5.4.6 Interactions with Other Pollutants 5-84
5.4.6.1 Oxidant Mixtures 5-85
5.4.6.2 Sulfur Dioxide 5-85
5.4.6.3 Nitrogen Dioxide 5-88
5.4.6.4 Hydrogen Fluoride and Other Gaseous
Pollutants 5-91
5.4.6.5 Acid Deposition 5-92
5.4.6.6 Heavy Metals 5-96
5.4.6.7 Mixtures of Ozone with Two or More
Pollutants 5-97
5.4.7 Interactions with Agricultural Chemicals 5-97
5.4.8 Factors Associated with Global Climate Change 5-99
5.4.9 Summary 5-103
5.5 EFFECTS-BASED AIR QUALITY EXPOSURE INDICES .... 5-106
5.5.1 Introduction 5-106
5.5.1.1 Biological Support for Identifying Relevant
Exposure Indices 5-106
5.5.1.2 Historical Perspective on Developing Exposure
Indices . 5-108
5.5.2 Developing Exposure Indices 5-114
5.5.2.1 Experimental Design and Statistical Analysis .. 5-114
5.5.2.2 Studies with Two or More Different Patterns
of Exposure 5-119
5.5.2.3 Combinations of Years, Sites, or Species:
Comparisons of Yield Losses with Different
Exposure Durations 5-123
5.5.2.4 Comparisons of Measures of Exposure Based
on Reanalysis of Single-Year, Single-Species
Studies 5-136
5.5.3 Summary 5-151
5.6 EXPOSURE-RESPONSE OF PLANT SPECIES 5-154
5.6.1 Introduction 5-154
5.6.2 Summary of Conclusions from the Previous Criteria
Documents 5-155
5.6.3 Information in the Published Literature Since 1986 .... 5-161
December 1993 H-viii DRAFT-DO NOT QUOTE OR CITE
-------
TABLE OF CONTENTS (cont'd)
Page
5.6.3.1 Effects of Ozone on Short-Lived (Less Than
One-Year) Species 5-164
5.6.4 Effects of Ozone on Long-Lived Plants 5-184
5.6.4.1 Perennial Agricultural Crops 5-185
5.6.4.2 Effects of Ozone on Deciduous Shrubs
and Trees 5-187
5.6.4.3 Effects of Ozone on Evergreen Trees 5-200
5.6.5 Assessments Using Ethylene Diurea as a Protectant . . . 5-208
5.6.6 Summary 5-213
5.7 EFFECTS OF OZONE ON NATURAL ECOSYSTEMS 5-215
5.7.1 Introduction 5-215
5.7.2 Ecosystem Characteristics 5-216
5.7.2.1 Expected Sequence of Events 5-216
5.7.3 Ecosystem Response to Stress 5-219
5.7.3.1 Forest Ecosystems 5-219
5.7.3.2 The San Bernardino Forest Ecosystem—
Before 1986 5-223
5.7.3.3 The San Bernardino Forest Ecosystem—
Since 1986 5-229
5.7.3.4 The Sierra Nevada Mountains 5-231
5.7.3.5 The Appalachian Mountains—Before 1986 . . . 5-234
5.7.3.6 The Appalachian Mountains—Since 1986 .... 5-235
5.7.3.7 Foliage and Soil-Mediated Effects—Combined
Stress 5-237
5.7.3.8 Mycorrhizae-Plant Interactions 5-242
5.7.3.9 Rhizosphere and Soil Processes 5-246
5.7.4 Summary 5-248
5.8 EFFECTS OF OZONE ON AGRICULTURE, FORESTRY, AND
ECOSYSTEMS: ECONOMICS 5-250
5.8.1 Introduction 5-250
5.8.2 Agriculture 5-251
5.8.2.1 Review of Key Studies from the 1986
Document 5-252
5.8.2.2 A Review of Post-1986 Assessments 5-256
5.8.2.3 Limitations and Future Research Issues 5-258
5.8.3 Forests (Tree Species) 5-259
5.8.4 Valuing Ecosystem Service Flows 5-262
5.8.4.1 Background 5-262
5.8.4.2 The Economic Perspective 5-263
5.8.4.3 Nonmarket Valuation: Implications for
Ecosystem Service Flows 5-265
5.8.4.4 Challenges in Linking Valuation Techniques
to Ecosystem Service Flows 5-269
5.8.4.5 The Research Agenda 5-271
December 1993 H-ix DRAFT-DO NOT QUOTE OR CITE
-------
TABLE OF CONTENTS (cont'd)
Page
5.8.4.6 Valuing Ecosystem Service Flows:
Summary 5-273
5.8.5 Summary 5-274
5.9 INTEGRATIVE SUMMARY AND CONCLUSIONS FOR
VEGETATION AND ECOSYSTEM EFFECTS 5-275
5.9.1 Introduction 5-275
5.9.2 Species Response and Ecosystem Response 5-276
5.9.3 How Does Ozone Affect Plants? 5-278
5.9.4 Factors That Modify Plant Response to Ozone 5-280
5.9.4.1 Genetics 5-280
5.9.4.2 Environmental Factors 5-281
5.9.5 Exposure Dynamics 5-283
5.9.6 What Measure of Exposure Characterizes Species
Effects? 5-284
5.9.7 What Is the Estimated Crop Yield or Biomass Change
with Ozone Exposure? 5-286
5.9.8 Ozone Concentration Across the United States 5-287
5.9.9 What Are the Exposure Effects on Other Species—Trees
and Ornamentals? 5-290
5.9.10 Spatial Characterization of Ozone Effects 5-292
5.9.11 What Is the Effect of Ozone on Ecosystems? 5-305
5.9.12 Economic Assessments 5-310
5.10 EFFECTS OF OZONE ON MATERIALS 5-311
5.10.1 Introduction 5-311
5.10.2 Mechanisms of Ozone Attack and Antiozonant
Protection 5-311
5.10.2.1 Elastomers 5-311
5.10.2.2 Textile Fibers and Dyes 5-314
5.10.2.3 Paint 5-316
5.10.3 Exposure-Response Data 5-316
5.10.3.1 Elastomer Cracking 5-317
5.10.3.2 Dye Fading 5-325
5.10.3.3 Fiber Damage 5-334
5.10.3.4 Paint Damage 5-338
5.10.3.5 Cultural Properties Damage 5-341
5.10.4 Economics 5-344
5.10.4.1 Introduction 5-344
5.10.4.2 Methods of Cost Classification and
Estimation 5-345
5.10.4.3 Aggregate Cost Estimates 5-346
5.10.5 Summary and Conclusions 5-348
REFERENCES 5-352
APPENDIX 5A: COLLOQUIAL AND LATIN NAMES 5A-1
December 1993 n-x DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF TABLES
Number
5-1 Examples of Intraspecific Variation of Foliar Symptoms in
Ozone Response .................. 5-45
5-2 Examples of Intraspecific Variation in Growth Responses
Following Ozone Exposures 5-47
5-3 Mortality of Three Ozone Sensitivity Classes of Eastern
White Pine (Pinus strobus L.) Trees During the Period
from 1971 to 1986 5-53
5-4 Examples of Ozone Effects on Pollen Germination and Tube
Elongation 5-56
5-5 Ozone Effects on Insect Pests 5-60
5-6 Ozone-Plant-Pathogen Interactions 5-64
5-7 Field Studies of Ozone-Drought Stress Interactions in Crop
Species 5-77
5-8 Ozone-Soil Nutrient Interactions 5-83
5-9 Some Statistical Models of Combined Ozone and Sulfur Dioxide
Responses 5-89
5-10 References to Reports of Interaction or No Interaction Between
Ozone and Acid Rain or Acid Fog 5-93
5-11 A Summary of Studies Reporting the Effects of Ozone for Two or
More Exposure Patterns on the Growth, Productivity, or Yield
of Plants 5-120
5-12 A Summary of Studies Reporting the Effects of Ozone on the
Growth, Productivity, or Yield of Plants for Two or More
Replicate Studies Having Equal Total Exposures and Either
Varying Durations or Similar Durations 5-125
5-13 Summary of Ozone Exposures That Are Closest to Those
Predicted for 20% Yield Reduction per SUM06 Exposure
Response Models Used by Lee et al. (1991) in Selected
NCLAN Experiments ............................ 5-138
December 1993 rj-xi DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF TABLES (cont'd)
Number Page
5-14 Summary of Percentiles for Ozone Monitoring Sites in
1989 (April Through October) with a Maximum
Three-Month SUM06 Value Less Than 24.4 ppm-Hour but
with Second Hourly Maximum Concentration Greater Than
or Equal to 0.125 ppm 5-141
5-15 Summary of Percentiles for Ozone Monitoring Sites in
1989 (April Through October) with a Maximum
Three-Month SUM06 Value Greater Than or Equal to
24.4 ppm-Hour but with Second Hourly Maximum
Concentration Less Than 0.125 ppm 5-142
5-16 Estimates of the Parameters for Fitting the Weibull
Model Using the Seven-Hour Seasonal Mean Ozone
Concentrations 5-159
5-17 Summary of Ozone Exposure Indices Calculated for
Three- or Five-Month Growing Seasons from 1982 to 1991 5-163
5-18 Comparison of Exposure-Response Curves Calculated
Using the Three-Month, 24-Hour SUM06 Values for 54 NCLAN
Cases 5-165
5-19 Comparison of Exposure-Response Curves Calculated
Using the 24-Hour W126 Values for 54 NCLAN Cases 5-168
5-20 The Exposure Levels (Using Various Indices) Estimated
to Cause at Least 10% Crop Loss in 50 and 75% of
Experimental Cases 5-172
5-21 SUM06 Levels Associated with 10 and 20% Yield Loss for
50 and 75% of the NCLAN Crop Studies 5-173
5-22 A Summary of Studies Reporting the Effects of Ozone
on the Growth, Productivity, or Yield of Annual
Plants Published Since U.S. Environmental
Protection Agency (1986) 5-176
5-23 A Summary of Studies Reporting the Effects of Ozone
on the Growth, Productivity, or Yield of Perennial
Crop Plants Published Since U.S. Environmental
Protection Agency (1986) 5-188
December 1993 H-xii DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF TABLES (cont'd)
Number Page
5-24 A Summary of Studies Reporting the Effects of Ozone on the
Growth or Productivity of Deciduous Shrubs and Trees
Published Since U.S. Environmental Protection Agency
(1986) 5-192
5-25 Exposure-Response Equations That Relate Total Biomass
(Foliage, Stem, and Root) to 24-Hour SUM06 Exposures
Adjusted to 92 Days 5-196
5-26 SUM06 Levels Associated with 10 and 20% Total Biomass
Loss for 50 and 75% of the Seedling Studies 5-198
5-27 A Summary of Studies Reporting the Effects of Ozone on the
Growth or Productivity of Evergreen Trees Published Since
U.S. Environmental Protection Agency (1986) 5-201
5-28 Effects of Ethylenediurea on Ozone Responses 5-210
5-29 Properties of Ecological Systems Susceptible to Ozone at
Four Levels of Biological Organization 5-221
5-30 San Bernardino Forest—Status 1972 5-224
5-31 Ecosystem Response to Pollutant Stress 5-228
5-32 Interactions of Ozone and Forest Tree Ectomycorrhizae
Interactions 5-243
5-33 Recent Studies of the Economic Effects of Ozone and
Other Pollutants on Agriculture 5-254
5-34 Studies of the Economic Effects of Ozone and Other
Pollutants on Forests 5-261
5-35 Laboratory and Field Studies on Effects of Ozone
on Elastomers 5-318
5-36 Protection of Tested Rubber Materials 5-321
5-37 Effect of Ozone and Humidity on Interply Adhesion 5-324
December 1993 H-xiii DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF TABLES (cont'd)
Number Page
5-38 Laboratory and Field Studies of the Effects of Ozone
on Dye Fading 5-326
5-39 Laboratory and Field Studies of the Effects of Ozone
on Fibers 5-335
5-40 Laboratory and Field Studies of the Effects of Ozone
on Architectural/Industrial Paints and Coatings 5-339
5-41 Laboratory Studies of the Effects of Ozone on Artists'
Pigments and Dyes . 5-342
5-42 Summary of Damage Costs to Materials by Oxidants 5-347
December 1993 H-xiv DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF FIGURES
Number Page
5-1 Leaf absorption and possible functional changes that may occur
within the plant 5-3
5-2 Uptake of ozone from the atmosphere 5-23
5-3 Movement of gases into and out of leaves is controlled primarily
by the stomata—small openings in the leaf surface whose aperture
is controlled by two guard cells 5-26
5-4 Simulation of the effects of diurnal variation in stomatal aperture
and in ozone concentration on ozone uptake 5-28
5-5 Effects of ozone absorption into a leaf 5-32
5-6 Effect of ozone on plant function and growth 5-34
5-7 The average injury index for visible foliar injury after exposure of
one-year-old seedlings to 50 pphm ozone for 7.5 hours 5-49
5-8 Frequency distribution showing the variability in ozone response
(midpoint of whole-plant biomass) within one half-sib family
of loblolly pine (P. taeda L.) exposed to increasing levels
of ozone under chronic-level field conditions over several
growing seasons (Adams et al., 1988) 5-50
5-9 Distribution pattern showing the number of ozone concentrations
within specified ranges for the 1983 winter wheat proportional
addition experiment for the 1.4 x ambient air; 1.8 x
ambient air treatments; and for San Bernardino, California
in 1987 5-118
5-10 Comparison of the Weibull exposure-response functions and its
predicted relative yield loss curves using M-7 and daytime
SUM06 for replicate years of National Crop Loss Assessment
Network Program's data for cotton (var. Acala SJ-2), wheat
(var. Vona), kidney bean (var. California light red), and
potato (var. Norchip), respectively 5-129
December 1993 H-xv DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF FIGURES (cont'd)
Number Page
5-11 Predicted relative yield losses (lint weight) for Acala SJ-2
cotton for four sites and multiple years (1981, 1982, 1988,
1989) relative to 0.10 ppm for M7, 0.035 ppm for 2ndHDM,
0 ppm-hour for SIGMOID, and 0 ppm-hour for SUM06, which
correspond to typical levels in the CF chambers 5-131
5-12 Relative effect of ozone on growth and yield of spring wheat
cultivars (var. Star and Turbo) from two growing seasons 5-132
5-13 Weibull exposure-response curves for the relative effect of
ozone on grain yield of spring wheat for three years,
individually and combined 5-133
5-14 Quadratic exposure-response curves for the relative effect of
ozone on grain yield of spring wheat in 1989 and 1990 using
four different exposure indices 5-135
5-15 Percent reduction in net photosynthesis of pines (including
one point for red spruce) and agricultural crops in relation
to total ozone exposure for several ranges of peak
concentrations 5-144
5-16 Percent reduction in net photosynthesis and biomass growth of
coniferous species in relation to total exposure and estimated
total ozone uptake 5-145
5-17 Percent reduction in net photosynthesis and biomass growth of
hardwood species in relation to total exposure and
estimated total ozone uptake 5-146
5-18 Percent reduction in net photosynthesis and biomass growth of
agricultural crops in relation to total exposure and
estimated total ozone uptake 5-147
5-19 Percent reduction in biomass growth of tree seedlings in
relation to total exposure 5-148
5-20 Reduction in volume production of loblolly pine seedlings
(family 91) in relation to four exposure indices 5-149
December 1993 n-xvi DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF FIGURES (cont'd)
Number Page
5-21 A comparison between the resulting cumulative frequencies
for the exposure parameters sum of all hourly average
concentrations and the sigmoidally weighted integrated
exposure index, W126 5-150
5-22 Box-plot distribution of biomass loss predictions from Weibull
and linear exposure-response models that relate biomass and
ozone exposure as characterized by the 24-hour SUM06 statistic
using data from 31 crop studies from the National Crop
Loss Assessment Network program and 26 tree seedling
studies conducted at the Environmental Research Laboratory
in Corvallis, Oregon; Smoky Mountain National Park, Tennessee;
Michigan; Ohio; and Alabama 5-175
5-23 Effects of ozone on plant function and growth 5-220
5-24 Effects of environmental stress on forest trees are presented
on a hierarchial scale for the leaf, branch, tree, and strand
levels of organization 5-222
5-25 Total basal area for each species as percent of the total basal
area for all species in 1974 and 1988 on plots with severe
to moderate damage, plots with slight damage, and plots with
very slight damage or no visible symptoms 5-230
5-26 Impact of a reduced supply of carbon to the shoot, or
water and nitrogen to the roots, on subsequent
allocation of carbon 5-241
5-27 Regions of the United States for analysis of trends of ozone
concentration 5-288
5-28 Ten-year trends in three-month SUM06 values by region at rural
monitoring sites 5-288
5-29 Ten-year trends in five-month 24-hour SUM06 values by region at
rural monitoring sites 5-289
5-30 Ten-year trends in second highest daily maximum values by
region at rural monitoring sites 5-289
5-31 1988 ozone monitoring site locations and calculated three-month
SUM06 at each site 5-294
December 1993 pj-xvii DRAFT-DO NOT QUOTE OR CITE
-------
LIST OF FIGURES (cont'd)
Number Page
5-32 Estimated ozone exposure across the eastern half of the
United States for 1988 and 1989 5-295
5-33 Second highest daily maximum for 1989 and 1988 estimated
across crop growing regions in the eastern half of the country . . . 5-296
5-34 Estimated relative yield loss for all crops from NCLAN
database across crop growing regions of the eastern half of the
United States in 1989 and 1988 5-297
5-35 Estimated relative yield loss in 1988 for soybean and
wheat from NCLAN database across crop growing regions
of the eastern half of the United States in 1989 and 1988 5-298
5-36 Variation in biomass reduction with year-to-year exposure
variation 5-299
5-37 Estimated biomass reduction for black cherry and red maple
with 1988 ozone exposure 5-300
5-38 Estimated relative biomass reduction for loblolly pine
and eastern white pine with 1988 ozone exposure 5-301
5-39 Box-plots of annual area-weighted yield reduction for the four
major agronomic crop species and all crops from NCLAN with
estimated 1988 and 1989 ozone exposure 5-303
5-40 Box-plots of annual area-weighted biomass reduction for
the eight tree species with estimated 1988 and 1989 ozone
exposure 5-304
5-41 Postulated mechanism for damage to elastomers by ozone 5-313
5-42 Reaction of anthraquinone dyes with ozone and with nitrogen
oxides 5-315
5-43 Relative decrease in stress with time as a function of ozone
concentration for polyisoprene vulcanizate 5-321
5-44 Relaxation of rubber compounds in ozone is affected by the
combination of rubber formulation and type of ozone protection . . 5-323
December 1993 n-xviii DRAFT-DO NOT QUOTE OR CITE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHAPTER 5. ENVIRONMENTAL EFFECTS OF OZONE
AND RELATED PHOTOCHEMICAL OXIDANTS
Principal Authors
Dr. Richard M. Adams—Department of Agriculture and Resource Economics, Oregon State
University, Corvallis, OR 97331
Dr. Christian Andersen—Environmental Research Laboratory, U.S. Environmental Protection
Agency, 200 SW 35th Street, Corvallis, OR 97333
Dr. J.H.B. Garner—Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Beverly A. Hale—Department of Horticultural Science, University of Guelph, Guelph,
Ontario, NIG 2W1, Canada
Dr. William E. Hogsett—Environmental Research Laboratory, U.S. Environmental Protection
Agency, 200 SW 35th Street, Corvallis, OR 97333
Dr. David F. Karnosky—School of Forestry and Wood Products, Michigan Technological
University, Houghton, MI 49931
Dr. John Laurence—Boyce Thompson Institute for Plant Research at Cornell University,
Tower Road, Ithaca, NY 14853
Dr. E. Henry Lee—ManTech Environmental Technology, Inc., 1600 W. Western Boulevard,
Corvallis, OR 97333
Dr. Allen S. Lefohn—A.S.L. & Associates, 111 Last Chance Gulch, Suite 4A,
Helena, MT 59601
Dr. Paul Miller—Pacific Southwest Forest and Range Experiment Station, USDA-Forest
Service Fire Lab, 4955 Canyon Crest Dr., Riverside, CA 92507
Mr. Doug Murray—TRC Environmental Corporation, 5 Waterside Crossing,
Windsor, CT 06095
Dr. Victor Runeckles—Department of Plant Science, University of British Columbia,
Vancouver, British Columbia, V6T 1Z4, Canada
December 1993 n-xix DRAFT-DO NOT QUOTE OR CITE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (cont'd)
Dr. James A. Weber—Environmental Research Laboratory, U.S. Environmental Protection
Agency, 200 SW 35th Street, Corvallis, OR 97333
Dr. Ruth D. Yanai—Boyce Thompson Institute for Plant Research at Cornell University,
Tower Road, Ithaca, NY 14853
Reviewers
Ms. Vicki Atwell—U.S. Environmental Protection Agency, OAQPS MD-R,
Research Triangle Park, NC 27711
Dr. Glen R. Cass—Environmental Engineering Science Department, Mail Code 138-78,
California Institute of Technology, Pasadena, CA 91125
Dr. Arthur Chapelka—Auburn University School of Forestry, Auburn, AL 36849-4201
Dr. Robert Goldstein—Electric Power Research Institute, 3412 Hillview Ave.,
Palo Alto, CA 94303
Dr. Marcia Gumpertz—Department of Statistics, North Carolina State University,
Raleigh, NC 27695
Dr. Allen Heagle—U.S. Department of Agriculture, ARS, 1505 Varsity Drive,
Raleigh, NC 27606
Dr. Robert Heath—Dept. of Botany and Plant Science, University of California,
Riverside, CA 92521
Dr. Thomas M. Hinckley—School of Forest Resources, University of Washington,
Seattle, WA 98195
Dr. Robert Kohut—Boyce Thompson Institute, Cornell University, Tower Road,
Ithaca, NY 14853-1801
Dr. Virginia Lesser—Department of Statistics, Oregon State University, Corvallia, OR 97333
Dr. Fred Lipfert—23 Carll Ct., Northport, NY 11768
Dr. Patrick McCool—SAPRC, University of California, Riverside, CA 92521
Dr. Delbert McCune—Boyce Thompson Institute, Cornell University, Tower Road,
Ithaca, NY 14853-1801
December 1993 H-xx DRAFT-DO NOT QUOTE OR CITE
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS (cont'd)
Dr. Robert Musselman—U.S. Department of Agriculture, Forestry Service, Rocky Mountain
Experiment Station, 240 West Prospect Road, Fort Collins, CO 80526
Dr. Eva Pell—Pennsylvania State University, Department of Plant Pathology, 321 Buckhout
Laboratory, University Park, PA 16802
Dr. Phillip Rundel—Laboratory of Biomedical and Environmental Science, University of
California, LA, 900 Veterans Ave., Los Angeles, CA 90024
Dr. Jayson Shogren—School of Forestry and Environmental Studies, Yale University,
New Haven, CT 06511
Dr. James Shortle—Department of Agricultural Economics, Pennsylvania State University,
Armsby Building, 208C, University Park, PA 16802-5502
Dr. John Skelly—Pennsylvania State University, 108 Buckhout Laboratory,
University Park, PA 16802
Dr. Boyd Strain—Department of Botany, 136 Biology Science Building, Duke University,
Durham, NC 27708
Dr. George E. Taylor, Jr.—Biological Sciences Center, Desert Research Institute,
7010 Dandini Blvd., Reno, NV 89512
Dr. Patrick Temple—Statewide Air Pollution Research Center, University of California,
Riverside, CA 92521-0312
Dr. David Tingey—U.S. Environmental Protection Agency, Environmental Research
Laboratory, 200 SW 35th Street, Corvallis, OR 97333
Dr. Michael Unsworth—Department of Atmospheric Sciences, Oregon State University,
Corvallis, OR 97333
Dr. David Weinstein—Boyce Thompson Institute, Cornell University, Tower Road,
Ithaca, NY 14853-1801
Dr. John Yocom, 12 Fox Den Road, West Simsbury, CT 06092
December 1993 D-xxi DRAFT-DO NOT QUOTE OR CITE
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
PROJECT TEAM FOR DEVELOPMENT OF AIR QUALITY CRITERIA
FOR OZONE AND RELATED PHOTOCHEMICAL OXIDANTS
Scientific Staff
Mr. James A. Raub—Health Scientist, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. A. Paul Altshuller—Physical Scientist, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Mr. William G. Ewald—Health Scientist, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. J.H.B. Garner—Ecologist, Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Dr. Judith A. Graham—Associate Director, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Ellie R. Speh—Secretary, Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Ms. Beverly E. Tilton—Physical Scientist, Environmental Criteria and Assessment Office
(MD-52), U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
Technical Support Staff
Mr. Douglas B. Fennell—Technical Information Specialist, Environmental Criteria and
Assessment Office (MD-52), U.S. Environmental Protection Agency, Research Triangle Park,
NC 27711
Mr. Allen G. Hoyt—Technical Editor and Graphic Artist, Environmental Criteria and
Assessment Office (MD-52), U.S. Environmental Protection Agency, Research Triangle Park,
NC 27711
Ms. Diane H. Ray—Technical Information Manager (Public Comments), Environmental
Criteria and Assessment Office (MD-52), U.S. Environmental Protection Agency, Research
Triangle Park, NC 27711
Mr. Richard N. Wilson—Clerk, Environmental Criteria and Assessment Office (MD-52),
U.S. Environmental Protection Agency, Research Triangle Park, NC 27711
December 1993 H-xxiii DRAFT-DO NOT QUOTE OR CITE
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
PROJECT TEAM FOR DEVELOPMENT OF AIR QUALITY CRITERIA
FOR OZONE AND RELATED PHOTOCHEMICAL OXIDANTS
(cont'd)
Document Production Staff
Ms. Marianne Barrier—Graphic Artist, ManTech Environmental, P.O. Box 12313,
Research Triangle Park, NC 27709
Mr. John R. Barton—Document Production Coordinator, ManTech Environmental
Technology, Inc., P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Lynette D. Cradle—Lead Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Joria R. Followill—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. Wendy B. Lloyd—Word Processor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Mr. Peter J. Winz—Technical Editor, Mantech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Technical Reference Staff
Mr. John A. Bennett—Bibliographic Editor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
Ms. S. Blythe Hatcher—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC 27709
Ms. Susan L. McDonald—Bibliographic Editor, Information Organizers, Inc.,
P.O. Box 14391, Research Triangle Park, NC 27709
Ms. Carol J. Rankin—Bibliographic Editor, Information Organizers, Inc., P.O. Box 14391,
Research Triangle Park, NC 27709
Ms. Deborah L. Staves—Bibliographic Editor, Information Organizers, Inc.,
P.O. Box 14391, Research Triangle Park, NC 27709
Ms. Patricia R. Tierney—Bibliographic Editor, ManTech Environmental Technology, Inc.,
P.O. Box 12313, Research Triangle Park, NC 27709
December 1993 U-xxiv DRAFT-DO NOT QUOTE OR CITE
-------
i 5. ENVIRONMENTAL EFFECTS OF OZONE AND
2 RELATED PHOTOCHEMICAL OXIDANTS
3
4
5 5.1 INTRODUCTION
6 Analyses of photochemical oxidants in the ambient air have revealed the presence of a
7 number of phytotoxic compounds, including O3, peroxyacyl nitrates, and NO2. Ozone, the
8 most prevalent photochemical oxidant, has been studied the most and its effects are better
9 understood than those of other photochemically derived oxidants. Ozone affects vegetation
10 throughout the United States, impairing crops, native vegetation, and ecosystems more than
11 any other air pollutant (Heck et al., 1980). The phytotoxicity of nitrogen oxides has been
12 assessed in Air Quality Criteria for Oxides of Nitrogen (U.S. Environmental Protection
13 Agency, 1993) and will not be discussed here. On the basis of concentration, the peroxyacyl
14 nitrates are more toxic than O3, with PAN being about ten times more phytotoxic than
15 O3 (Darley et al., 1963; Taylor and MacLean, 1970; Pell, 1976). Although more phytotoxic
16 than O3, the peroxyacyl nitrates generally occur at significantly lower ambient
17 concentrations, however, and phytotoxic concentrations are therefore less widely distributed
18 than those of O3. Ambient concentrations of O3 and PAN, as well as their concentration
19 ratios, are discussed in detail in Chapter 4.
20 The effects of photochemical oxidants were first observed as foliar injury on vegetation
21 growing in localized areas in Los Angeles County, California (Middleton et al., 1950).
22 In these early reports, foliar injury was described as glazing, silvering, and bronzing of the
23 lower leaf surface of leafy vegetables and as transverse bands of injury on monocotyledonous
24 species. Subsequent studies showed that these symptoms of photochemical oxidant injury
25 were caused by peroxyacetyl nitrate (Taylor et al., 1960). The characteristic O3 stipple on
26 grape leaves reported in the late 1950s was the first observation of O3 injury to vegetation in
27 the field (Richards et al., 1958). Subsequent studies with tobacco and other crops confirmed
28 that O3 was injuring vegetation at sites near urban centers (Heggestad and Middleton, 1959;
29 Daines et al., 1960). It is now recognized that vegetation at rural sites may be injured by
30 O3 transported long distances from urban centers (Edinger et al., 1972; Heck et al., 1969;
December 1993 5-1 DRAFT-DO NOT QUOTE OR CITE
-------
1 Heck and Heagle, 1970; Kelleher and Feder, 1978; Miller et al., 1972; Skelly et al., 1977;
2 Skelly, 1980; Garner et al., 1989; see also Chapters 3 and 4).
3 Plant stress from O3 exposures occur when the atmospheric concentrations exceed the
4 limits of plant tolerance, not because the gas is unique. Not all plants are sensitive. The
5 effects of O3 on terrestrial vegetation begin with the responses of individual plants (see
6 Figure 5-1). When there are many sensitive individuals within a population, populations are
7 affected. Changes within sensitive populations, or stands, ultimately can affect community
8 and ecosystem structure and function. The occurrence and magnitude of the effects depends
9 on the pollutant concentration, duration of the exposure, length of time between exposures,
10 genetic composition (i.e., sensitivity) of the plants, and various environmental and biological
11 factors influencing plant response.
12 Foliar injury is usually the first visible sign of O3 exposure and indicates impairment of
13 physiological processes within the leaves. As illustrated in Figure 5-1, in order for O3 to
14 affect an individual plant, sufficient amounts of O3 derived from atmospheric and various
15 canopy processes must be able to reach the leaves of the plant. To cause injury, ozone must:
16 (1) enter the plant through the leaf stomata; (2) dissolve in the aqueous layer lining the cell
17 walls within the air spaces; and (3) O3 or its decomposition products diffuse through the
18 membrane into the cell, where it can react with cellular components and affect metabolic
19 processes, unless the plant is able to detoxify or metabolize O3 or its metabolites (Tingey and
20 Taylor, 1982). Cellular injury has been shown to subsequently manifest itself in a number of
21 ways. These include: (1) visible foliar injury; (2) premature needle senescence; (3) reduced
22 photosynthesis; (4) reduced carbohydrate production and allocation; (5) reduced plant vigor;
23 and (6) reduced growth or reproduction or both (McLaughlin et al., 1982; U.S.
24 Environmental Protection Agency, 1986). The impact of pollutant exposure is determined by
25 the developmental stage, the genetic composition, the complex interactions among natural and
26 pollutant stresses, their temporal and spatial variation, and finally, the action these have on
27 the biochemical and physiological processes within the plants. Most plants undergo some
28 form of stress during the various stages of their life cycle; however, the multiple stresses
29 they encounter during their lifetimes do not usually all occur at one time (Osmond et al.,
30 1987).
December 1993 5-2 DRAFT-DO NOT QUOTE OR CITE
-------
Atmospheric Processes
Canopy Processes
Leaf Processes/
Ozone Uptake
Leaf Processes/
Mode of Action
Plant Response
Ecosystem Response
[Reduced Photosynthesis
[Reduced Carbohydrate Production)
' '
[Reduced Carbohydrate Allocation]
r
Compensation
Reduced Growth ]
V
(Reduced Reproduction)
*
Increased Susceptibility to
Biotic and Abiotic Stresses
T
Decrease in Mycorrhizae
Formation
f Individual Response 1
y
[ Population Response]
[Community Response]
[Ecosystem Response]
Figure 5-1. Leaf absorption and possible functional changes that may occur within the
plant. Ecosystem response begins at the level of the individual and is
propagated upward.
December 1993
5-3
DRAFT-DO NOT QUOTE OR CITE
-------
1 Ozone can affect all aspects of plant growth (Figure 5-1). Plants accumulate, store,
2 and use carbon compounds to build their structure and maintain physiological processes
3 (Waring and Schlesinger, 1985). Within the leaf, carbon dioxide absorbed from the
4 atmosphere is converted to carbohydrates during the process of photosynthesis. The water
5 and minerals necessary for growth are absorbed by plants from the soil. Growth and seed
6 formation depend not only on the rate of photosynthesis and uptake of water and nutrients,
7 but also on the subsequent metabolic processes and the allocation of the carbohydrates
8 produced during photosynthesis. Most plants require a balance of resources (i.e, energy,
9 water, and mineral nutrients) to maintain optimal growth, but these are seldom available in
10 natural environments (Chapin et al., 1987). Plants compensate for injury and/or stresses by
11 allocating their available resources to the point of injury or stress (Mclaughlin et al., 1982;
12 Miller et al., 1982; Tingey et al., 1976). Altering the allocation of carbohydrates has been
13 shown to decrease plant vigor, increase susceptibility to insect pests and fungal pathogens,
14 interfere with mycorrhizal formation and reduce plant growth and reproduction (McLaughlin
15 et al., 1982; Miller et al., 1982; U.S.Environmental Protection Agency, 1986; Garner et al.,
16 1989).
17 Most of our knowledge concerning the effects of O3 on vegetation comes from studies
18 of the responses of important agricultural crop plants and some selected forest tree species,
19 mostly as seedlings or saplings. Crop plants, because of the importance of their yield to
20 human food demand, usually have been selected for their productivity. In addition, crop
21 plants usually are fertilized, weeded, frequently irrigated and grown in monocultures.
22 In other words, competition for water, nutrients, space and light is greatly diminished when
23 compared to plants growing in natural ecosystems.
24 The number of crop species and cultivars for which information regarding O3 effects
25 exists encompass a mere fraction of the total of those cultivated as crops or found growing in
26 natural communities. It is not possible to predict the sensitivity of the species and cultivars
27 that have not been investigated, except in very general terms, because of the wide range of
28 sensitivities to O3 known to exist among the species that have been studied and even among
29 cultivars of individual crops. A number of attempts have been made to develop a general
30 framework of response covering a range of species using the fragmented knowledge
31 available. All of them have their shortcomings.
December 1993 5-4 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ecosystem responses are hierarchical and range from those that are characteristic of
2 individual organisms through populations to communities, and ultimately, to ecosystems
3 (U.S. Environmental Protection Agency, 1978, 1986). Organisms, not ecosystems, respond
4 to O3 exposure (Sigal and Suter, 1987). The only well documented study of ecosystem
5 change is that of the San Bernardino Forest ecosystem in southern California where the
6 impact of O3 on the keystone species, ponderosa and Jeffrey pine resulted in the degradation
7 of the forest (Miller, 1982, 1984; U.S. Environmental Protection Agency, 1978, 1986).
8 Studies within the forests of the eastern United States, until recently, have dealt with only a
9 few plant species, chiefly eastern white pine, none of which have been critical in maintaining
10 ecosystem structure and function (McLaughlin et al., 1982; Skelly, 1980; Skelly et al., 1984;
11 U.S. Environmental Protection Agency, 1978, 1986). The absence of long term studies
12 dealing with the impacts of O3 on various ecosystem components and how these influence
13 ecosystem structure and functions makes determination of the impact of O3 on ecosystems in
14 general difficult.
15 The sequential organization of this chapter closely follows the discussion dealing with
16 the responses of plants to O3 presented above. First, the methodologies (Section 5.2) that
17 have been used to obtain the information on which the rest of the chapter is based are
18 described. The next section (5.3) details the known biochemical and physiological responses
19 occurring within leaf cells after O3 entry into the leaves and how these responses affect plant
20 vigor, growth, productivity and reproduction. Factors within and external to plants influence
21 their response to O3 and other stresses. How these factors, observed during experimentation
22 or in the field, can modify functional and growth responses of plants exposed to O3 is
23 presented in the next section (Section 5.4).
24 A complicating factor in assessing plant response to O3 exposure is determining the
25 amount of O3 entering the leaves. Entry of O3 into plant leaves is not a simple uptake
26 process, but involves gas exchange; the entry of O3 through the stomata and carbon dioxide
27 (CO2) from the atmosphere at the same time oxygen and water vapor are exiting. For many
28 years, attempts have been made to develop mathematical equations that quantify the
29 relationship between pollutant exposure and agricultural crop yield. The problem is the need
30 for combined incorporation of the age and the genetic composition of the plant, along with
31 parameters for pollutant concentration, duration, frequency of exposure and the respite time
December 1993 5,5 DRAFT-DO NOT QUOTE OR CITE
-------
1 between exposures into an exposure statistic or index which may be used to predict crop
2 yield loss. Section 5.5 discusses the strong points and shortcoming of the various indices
3 that have been developed to aid in predicting the effects of O3 on crops.
4 Some O3 exposures (concentration and duration) result in foliar injury to the plant.
5 Other exposures result in growth reduction and decrease in productivity. Section 5.6,
6 exposure-response of plant species, presents the data from many different experiments and
7 assesses the potential impact of different concentrations and exposure durations on growth of
8 a variety of plants ranging from cultivated annual crops to woody perennials such as trees
9 and shrubs. Of particular interest is the yield of crops. The majority of the data concerning
10 the response of ecosystems deals with the response of individual organisms, populations or
11 forest stands. Therefore, in section (5.7) the knowledge concerning ecosystem response to
12 perturbations obtained from a variety of studies, and the California San Bernardino Forest
13 study in particular, is used to present a view of what might be expected from continued
14 O3 exposure.
15 The costs to the nation of O3 exposure of crops and ecosystems is discussed in
16 Section 5.8. The scientific names of the plants cited in this chapter are presented in a table
17 in Appendix A. Section 5.9 discusses the effects of O3 on nonbiological materials, followed
18 by an integrative summary and conclusions section (5.10) drawing together and interpreting
19 key information from the earlier chapter sections concerning vegetation effects.
20
21
22 5.2 METHODOLOGIES USED IN VEGETATION RESEARCH
23 5.2.1 Fumigation Systems
24 The methodologies used in vegetation research have become more sophisticated over
25 the years as new technology has developed. New exposure systems have been devised with
26 pollutant dispensing systems that make it possible to more nearly duplicate the exposures
27 plants receive in the field. These systems and their good points and short-comings are
28 discussed below.
29 Air pollution exposure-response studies usually require exposure chambers or other
30 apparatus for maintaining controlled pollutant exposures. Ozone exposure systems are
31 designed to maintain a modified gaseous atmosphere around a plant for a period of exposure,
December 1993 5-6 DRAFT-DO NOT QUOTE OR CITE
-------
1 for the purpose of monitoring plant responses to that modified gaseous atmosphere.
2 Whichever system is used, they all share some features in common, namely: general plant
3 growth conditions (light, temperature, humidity, carbon dioxide, soil water) must be met,
4 and differential concentrations of O3 generated either artificially or naturally must be
5 supplied to the fumigation system. Exposure systems have been established in controlled
6 environments, greenhouses and the field, many of these were described in the earlier criteria
7 document, Air Quality Criteria for Ozone and Other Photochemical Oxidants (U.S.
8 Environmental Protection Agency, 1986). Exposure systems may range from cuvettes which
9 enclose leaves or branches (Bingham and Coyne, 1977; Legge et al., 1979) to a series of
10 tubes with calibrated orifices spatially distributed over a field to emit gaseous pollutants to a
11 plant canopy (Lee et al., 1978). Each type of system was designed for specific objectives
12 and operates most efficiently under the conditions for which it was intended. Each has
13 advantages and limitations and must be evaluated in terms of the objectives it was designed
14 to meet.
15
16 5.2.1.1 Methodologies Discussed in the Air Quality Criteria for Ozone and Other
17 Photochemical Oxidants (1986)
18 Controlled Environment Exposure Systems
19 Controlled environment fumigation systems are those in which light sources, and
20 control of temperature and relative humidity are artificial. Light quality and quantity are
21 likely to be lower than in ambient environments, usually resulting in lower photosynthetically
22 active radiation (PAR). Temperature and relative humidity will likely be more consistent in
23 a controlled environment than in ambient air. Controlled environment exposure systems are
24 typified by the widely-used continuous stirred tank reactor (CSTR), a system originally
25 designed for mass balance studies of O3 flux to vegetation. The main benefit of controlled
26 environment chambers is that the environmental conditions during exposure can be very well
27 characterized, controlled and replicated over time. They work very well for evaluating
28 O3 effects on physiological processes, which are themselves sensitive to changes in other
29 environmental parameters. The major limitation of controlled environment exposure systems
30 is the extrapolation of the data to field situations.
31 Greenhouse system designs are similar to those found in controlled environments,
32 except that light, temperature and relative humidity conditions fluctuate with those occurring
December 1993 5-7 DRAFT-DO NOT QUOTE OR CITE
-------
1 in the greenhouse. Thus, they are more closely related to field studies than are controlled
2 environments, but plant culture and environmental conditions are still quite different from
3 those of field exposure chambers, making direct extrapolation difficult. These studies are,
4 however, more applicable to phytotoxicity of O3 to greenhouse grown ornamental and
5 floriculture crops (U.S. Environmental Protection Agency, 1986). Some greenhouse
6 exposure systems use activated charcoal filtration to remove pollutants from the incoming air
7 prior to the addition of experimental O3, and either vent directly to the outside, or use
8 charcoal filtration of the outgoing air to prevent contamination of the greenhouse air supply.
9 Other greenhouse exposure systems filter neither incoming nor outgoing air.
10
11 Field Exposure Systems
12 Fumigation of plants with O3 in the field is most frequently carried out using open-top
13 chambers. The most widely utilized design (U.S. Environmental Protection Agency, 1986)
14 consists of a cylindrical aluminum frame, covered with transparent film. The bottom half of
15 the transparent covering is doubled layered, the inside panel of which is perforated.
16 Charcoal and particulate filtered air, non-filtered air or O3 supplemented air is blown into the
17 bottom layer, forced through the perforations into the plant canopy, and then escapes through
18 the top of the chamber. The positive pressure maintained by the forced movement of air up
19 through the chamber minimizes influx of ambient air into the chamber through the open-top.
20 The open-top chamber exposure system was employed in the National Crop Loss Assessment
21 Network (NCLAN) from 1980 to 1988 and a description and discussion of the chambers is
22 provided in the 1986 criteria document Section 6.2.4 (U.S. Environmental Protection
23 Agency, 1986). The design of these chambers has been modified with frusta to reduce such
24 incursions by ambient air, making the chambers more viable under windy conditions.
25 Canopies (moveable) have been added so that rain exclusion studies can be carried out.
26 Finally, they have been modified in shape or increased in size so that species such as mature
27 trees and grapevines can be enclosed. Crop loss studies conducted in open-top chambers
28 more closely simulate field losses of crops due to O3 than controlled environments or
29 greenhouse fumigation chambers. However, there are some important differences in
30 environmental conditions between open-top chambers and the field. Specifically, the frusta
31 and walls of the chamber block some incoming light, as well as trap some long-wave
December 1993 5-8 DRAFT-DO NOT QUOTE OR CITE
-------
1 radiation within the chamber, so that PAR levels are lower and temperatures are higher in
2 the chambers than in the field. The walls also cause a rain shadow necessitating irrigation.
3 The rate at which air moves across the foliage contributes significantly to canopy resistance
4 to O3 uptake, and is similar between chambered and non-chambered plants. This suggests
5 that O3 uptake by the foliage is similar between open-top chambers and the field (U.S.
6 Environmental Protection Agency, 1987). To promote normal dew formation, many studies
7 turn off the blower fans during the night.
8 Limited use (for O3 studies) has been made of chamberless field exposure systems,
9 which rely on ambient wind conditions to move O3 across an open field canopy. The O3 is
10 emitted from vertical pipes, which are spaced in a circle around the experimental plot of
11 plants. The amount of O3 emitted from each vertical pipe, as well as the number and
12 compass direction of emitting pipes, depends on the wind direction and speed, this whole
13 process usually being computer controlled.
14
15 5.2.1.2 Methodologies Referenced Since the Air Quality Criteria for Ozone and Other
16 Photochemical Oxidants (1986)
17 Branch and Leaf Chambers
18 Most of the developments in exposure systems since 1986 have been modifications of
19 existing systems. The tremendous interest in evaluation of mature tree response to O3 has
20 prompted the development of large branch chambers for estimating O^ flux to trees. These
21 branch chambers share many of the design characteristics of a CSTR. The chamber walls
22 are transparent film spread over a supporting frame, there is a fan to reduce boundary layer
23 resistance across the foliar surface, and an air inlet and outlet so that differential O3, carbon
24 dioxide (photosynthesis) and water vapor (leaf diffusive resistance) measurements can be
25 taken (Ennis et al., 1990; Houpis et al., 1991; Teskey et al., 1991). The advantages of this
26 system include the ease with which the Teflon bag can be replaced, uniform light
27 transmission can be maintained, the branch chamber can be moved from plant-to-plant, can
28 be used in situ, and can be modified for different sized branches. One of the disadvantages
29 of the branch chamber, and indeed of any such cuvette which isolates one part of the plant
30 under different environmental conditions than the rest of the plant, is whether the isolation
31 leads to a response different from that which would have been observed if the branch was
32 under the same environmental conditions as the rest of the plant. In addition, total tree
December 1993 5.9 DRAFT-DO NOT QUOTE OR CITE
-------
1 growth cannot be estimated using branch chambers, as only part of the plant is treated with
2 O3.
3
4 5.2.1.3 Flux Measurement
5 Estimation of O3 flux to foliage can be made directly, by measuring the difference in
6 O3 concentration between air going into a leaf chamber and the same air stream exiting the
7 chamber after passing over the leaf. It can also be inferred from measurements of leaf
8 diffusive resistance during exposure of a leaf to O3. The former method requires a chamber
9 or cuvette fumigation system with uptake of O3 which is quite small or extremely
10 nonvariable relative to the amount being taken up by the leaf. Otherwise, it is difficult to
11 detect O3 flux to a leaf with good precision. Such cuvettes can be adapted from those
12 commercially available for portable photosynthesis meters (Graham and Ormrod, 1989) or
13 constructed from a novel design, such as that developed by Fuentes and Gillespie (1992) to
14 estimate the effect of leaf surface wetness on O3 uptake of maple leaves. The criteria for
15 flux cuvette design include good light transmissibility, ease of leaf manipulation, minimal
16 reaction of chamber wall surface with O3 and good air mixing within the chamber. Good
17 mixing of air is necessary to avoid a gradient in pollutant concentration, and to maintain a
18 boundary layer resistance which is greater than stomatal resistance. Maintenance of leaf
19 temperature close to that of the surrounding air, so that transpiration rates are not abnormally
20 high, is another benefit of good air mixing. The physical design of the Fuentes and Gillespie
21 chamber was simple, consisting of two glass hemispheres that were clamped together,
22 separated by a Viton O-ring, over the petiole of the leaf under investigation. Inlet and outlet
23 air attachments were on opposite sides of the cuvette. Other cuvette designs have been used
24 to estimate leaf gas exchange responses to O3; their principals of operation are similar, but
25 there are differences in materials and design (Amiro et al., 1984; Freer-Smith and Dobson,
26 1989; Laisk et al., 1989; Moladau et al., 1991; Skarby et al., 1987).
27 Compared to the CSTR, which has been used for mass balance measurement of gas flux
28 by whole plants during fumigation (LeSueur-Brymer and Ormrod, 1984), cuvette systems
29 usually determine flux to one leaf at a time. This results in a more precise understanding of
30 the interaction among leaf age, leaf diffusive resistance, leaf illumination and O3 flux.
31 However, these data are not particularly well adapted to estimating flux of O3 to a large
December 1993 5-10 DRAFT-DO NOT QUOTE OR CITE
-------
1 vegetated surface. Finally, regardless of the methodology used to determine O3 flux to
2 foliage, there exist only very sketchy mechanistic-process models which would link O3 fluxes
3 to decreases in growth and productivity of plants. At this time, these data are mainly useful
4 for developing a relationship between internal O3 dose and plant response, and estimating the
5 strength of vegetation as sinks for O3 flux on a large scale.
6
7 Pollutant Dispensing Systems
8 Although exposure chambers have changed little in design in the last several years, the
9 profile characteristics and method of dispensing pollutant profiles have. Whereas early
10 studies utilized static or square-wave exposures, usually controlled by hand-set flowmeters,
11 many more recent system expose plants with so-called dynamic exposures, during which the
12 O3 concentration gradually reaches a maximum, thus simulating diurnal variation in
13 O3 concentration (Hogsett et al., 1985a). These profiles may be achieved by mass flow
14 controllers that are themselves computer controlled. Proportional add systems such as that
15 used in NCLAN usually achieve ambient type profiles using rotameters instead of mass flow
16 controllers. The O3 concentration in each of the chambers is logged at pre-set intervals, so
17 that the integrated exposure for the entire fumigation period can be calculated. Deviations
18 from the planned O3 episode can occur, due to failure in dispensing or monitoring
19 equipment, as well as incursions of air through the tops of the chambers. The length of the
20 interval between determinations of O3 concentration in the chambers can be an important
21 contribution to the control of O3 profile. In general, longer intervals lead to less well-
22 controlled and well-characterized O3 exposure profiles (Lefohn et al., 1993). These
23 deviations from the expected profiles can be mathematically quantified, and monitored among
24 treatments and replications (Hale-Marie et al., 1991).
25
26 Open Air Field Fumigation Systems
27 Open air field fumigation systems have the potential to most closely estimate filed
28 losses due to O3, as the plants are grown and exposed under ambient field environmental
29 conditions. However, of all the fumigation systems, this is the least controllable and
30 repeatable. It has been used in the past to expose plants to "static" concentrations (desired
31 concentration is the same throughout the exposure period) of such pollutants as sulphur
December 1993 5. j j DRAFT-DO NOT QUOTE OR CITE
-------
1 dioxide or hydrogen fluoride (Hogsett et al., 1987). The Zonal Air Pollution System (ZAPS)
2 has been vastly modified and improved upon to enable fumigation of plants with the diurnally
3 varying pattern of concentration typical of ambient O3 fumigation (Runeckles et al., 1991).
4 The system represents a significant advancement over earlier open-air field fumigation
5 systems, in, that 12 discrete seasonal treatments which simulate ambient patterns are achieved
6 rather than the usual two or three. Ozone was supplied to plots, which were laid out in
7 groups of four, through a manifold suspended over the plant canopy. The wind speed and
8 direction determined the actual seasonal O3 exposures, although the O3 was released in
9 concentrations proportional to that observed at the time in the ambient environment. While
10 the twelve treatments are not repeatable over time, a regression relationship between
11 pollutant exposure and plant response can be established for each growing season.
12 The Liphook study in England of long-term responses of Picea sitchensis (Bong.) Carr,
13 Picea abies (L.) Karst., and Pinus sylvestris L. to sulphur dioxide and O3 in combination
14 consisted of seven growth plots, 50 m in diameter, five of which were surrounded by
15 64 vertical pipes from which pollutant gasses were emitted (McLeod et al., 1992). The
16 64 pipes were divided into four quadrants of 16 adjacent pipes, and each quadrant had diluted
17 pollutant gases supplied to it from a computer controlled mass flow controller. The emitting
18 quadrant(s) as well as the rate at which the gases were supplied to the quadrant depended on
19 wind speed and direction. The gases were emitted from the vertical pipes into the plant
20 canopy at two heights, 0.5 and 2.5 m above a reference height, which was approximately
21 two-thirds of tree height. This pattern of gas dispersion resulted in a uniform horizontal
22 distribution of hourly mean gas concentration across each central 25-m diameter experimental
23 plot. Measured over a winter wheat canopy, SO2 concentration differed by less than 1 nl.l"
24 over a 5-h period of measurement; measurement of consecutive 2-min mean values at five
25 locations across the plots demonstrated high uniformity (McLeod et al., 1985). This
26 exposure system, like all open-air exposure systems, clearly simulates field plant growth
27 conditions far better than open- or closed-top chambers. The usefulness of the data is limited
28 however, by the low number of treatments and lack of replication of those treatments. And,
29 with five "enclosures", and two nonenclosed ambient plots, this is by far the largest of the
30 very few of these systems which are in operation.
31
December 1993 5-12 DRAFT-DO NOT QUOTE OR CITE
-------
1 Field Chamber Exposure Systems
2 Open-top field chambers are used in most field studies of plant response to gaseous
3 pollutants. They were first designed for studies on annual herbaceous crop plants (Mandl
4 et al., 1973) but enlarged versions have been used successfully in tree seedling and sapling
5 studies also (Adams et al., 1990; Chappelka et al., 1990; Qui et al., 1992; Kress et al.,
6 1992; Hogsett et al., 1989; Andersen et al., 1990; Karnosky et al., 1993; Wang et al., 1991;
7 Temple et al., 1991). Because the results from these studies using tree species are
8 extrapolated to predict the effects of O3 on forests, these studies require good exposure
9 control in order to replicate ambient O3 profiles characteristic of many low-elevation, rural
10 areas of eastern North America. This condition could have been met using an open field
11 exposure system. Open-top chambers which are large enough for mature trees to have been
12 developed but are expensive (Grandl et al., 1989).
13 Microclimatic modification by open-top chambers as well as O3 exposure schedules
14 which are disconnected from typical O3 episode meteorology have been addressed in a
15 seasonal study of tree response to O3 in the United Kingdom (Wiltshire et al., 1992). This
16 study is using open top chambers with roll-up sides; except for fumigation days, the plants
17 are maintained in ambient climatic conditions. The exposure episodes number between
18 27 and 30 throughout the growing season, and occur on days with ambient meteorology
19 associated with naturally occurring O3 episodes (i.e., high incident radiation and temperature
20 with little air movement) (Wiltshire et al., 1992). The maintenance of near-ambient
21 meteorological conditions during both growth and exposure periods ensures that this study
22 will represent field grown plant responses to O3 better than traditional open-top chamber
23 studies.
24 Several designs of field fumigation chambers have been developed to overcome some of
25 the disadvantages of the open-top chambers, namely small plot size and incursion of ambient
26 air. Closed top chambers were first developed in the 1950's; generally, smaller in dimension
27 to the open-top design, have been more recently constructed in California to assess crop loss
28 to O3. Closed top chambers were chosen because the authors wished to characterize the
29 pollutant dose to the plants very precisely; pollutant gradients within the chamber were
30 minimal (Musselman et al., 1986). The chambers were octagonal in shape and covered with
31 Teflon® film; the soil was completely replaced with standard greenhouse mix. Temperatures
December 1993 5_13 DRAFT-DO NOT QUOTE OR CITE
-------
1 in the chamber were higher (2 to 4 °C at midday, 1 to 2 °C at night) than in the ambient air,
2 and light levels were reduced by 11 % (spectral quality of the light in the chambers was not
3 reported). The authors concluded that, although the chambers were not suitable for studies
4 requiring close approximation of field conditions, they were very useful when tight control
5 over soil moisture and pollutant concentration were needed.
6 Closed-top chambers were constructed and installed in the United Kingdom to study
7 responses of shrubs and large herbaceous species to long term, low (chronic) concentrations
8 of sulphur dioxide, nitrogen dioxide and O3 (Rafarel and Ashenden, 1991). They were a
9 smaller version of an earlier design, as the larger chambers required pure gas sources of
10 nitrogen dioxide and sulphur dioxide to be diluted into the ventilating air stream, which
11 resulted in highly variable exposure concentrations. The flow rate of the smaller chambers
12 meant that premixed gases were sufficient to maintain steady control of treatment
13 concentrations. Because the gases were discharged from the source at constant
14 concentrations, different treatments were achieved by placing one or more pollutant supply
15 tubes in the fumigation chambers. Good air circulation (and moderate ambient temperatures)
16 maintained the domes at near ambient conditions.
17
18 Ambient Gradients for Evaluation of Plant Response to Ozone
19 The exposure system which most simulates ambient conditions of O3 exposure,
20 temperature, humidity, soils, soil moisture, is the ambient gradient system. By this method,
21 plants are grown along a transect of known differential pollutant concentrations, usually
22 downwind of a major point source or urban area. The concentration of pollutant(s) is diluted
23 as distance from the source increases. A study of four cultivars of red clover (Trifolium
24 repens L) and spring barley (Hordeum vulgare L) was conducted along such a transect of
25 gradient sulphur dioxide, nitrogen dioxide, and O3 concentrations in the United Kingdom
26 (Ashmore et al., 1988). Ozone concentration was inferred from injury to Bel W3 and Bel B
27 cultivars of tobacco, but was found to have very little relationship to cultivar performance.
28 The authors cautioned that these results must be interpreted with an understanding that
29 differences among sites in other environmental parameters could contribute to the detection
30 (or failure to detect) of O-j effects on the crops. For ambient gradient studies to be
31 interpretable, good characteri/.ation of site parameters (rainfall, temperature, radiation, soil
December 1993 5-14 DRAFT-DO NOT QUOTE OR CITE
-------
1 type) is needed. Additionally, the modeler needs to know how these factors should be used
2 to adjust the apparent plant response. In order to know that, a good knowledge base of how
3 all of these factors modify plant response to O3 is needed.
4 At this time, although some information is available, the relationships are still
5 incompletely understood. Many investigators consider that ambient gradients are impossible
6 to find without major differences in environmental conditions which may affect plant
7 response to O3 and therefore confound interpretation of the results.
8
9 Comparison of Fumigation Systems
10 Each type of fumigation system is particularly well suited to certain types of studies of
11 plant response to O3. To gather data relevant to the mechanisms of plant response to O3,
12 chambers with a high degree of control over environmental conditions are needed. Many of
13 the physiological bases of plant response to O3 can be themselves highly influenced by small
14 differences in light, humidity, and temperature. For example, measurements of O3 effects on
15 photosynthetic rates in chambers with illumination levels well below saturation for
16 photosynthesis may be highly variable with very small changes in PAR, as the relationship
17 between PAR and photosynthetic rates for many species is steeply linear at low light levels.
18 Alternately, it is notoriously difficult to get reliable photosynthetic data from field chambers
19 under ambient light, as cloud cover and diurnal changes in the solar angle cause change in
20 instantaneous rates of photosynthesis. For high precision in estimating photosynthetic
21 responses to O3, controlled environment exposure chambers with saturating light sources are
22 needed; such systems are not yet in wide-spread use.
23 A comparison of plant growth and plant response to O3 fumigation in open-top
24 chambers, closed-top chambers, and air exclusion systems has been carried out (Olszyk
25 et al., 1986). The authors discovered that there was interaction between plant response to
26 O3 and type of exposure system for less than 10% of the growth parameters measured in
27 California, suggesting that plant response to O3 was the same regardless of fumigation
28 system. Plants from exclusion systems were shorter than in open-top chambers, and
29 generally weighed more. Of the three groups of plants, those in the control plots of the
30 exclusion system (i.e., receiving ambient O3 exposure) were most similar in size to plants
31 grown in field plots. While this and another study (Olszyk et al., 1992) indicate that
December 1993 5_15 DRAFT-DO NOT QUOTE OR CITE
-------
1 environmental modification caused by chambers will affect plant growth, there is no evidence
2 that there is a large effect of chambers on plant response to 63. It is assumed that, because
3 of the decreasing relative effects on plant environment caused by controlled environment,
4 greenhouse, closed cop field chambers, open-top field chambers, open air systems, and
5 ambient gradients, the system effects on plant response to O3 will decrease in the same
6 order.
7 Considerable concern has been raised about plant response to trace pollutants in
8 exposure chambers, specifically N2O5 and NO in chambers receiving O3 generated from dry
9 air, and NO2 in chambers receiving ambient air. These trace pollutants may have a direct
10 effect (positive or negative) on plant processes, or may change how plants respond to O3,
11 and without careful evaluation, these effects may go undistinguished from those of O3.
12 A comparison of alfalfa response to the same O3 exposure, generated electrostatically from
13 air or through nonfiltration of ambient air, indicated that the generated O3 treatment was
14 more phytotoxic than the ambient Q3 treatment, likely due to the co-generation of N2O5 and
15 NO along with O3 from dry air (Olszyk et al., 1990). The results of this study and the
16 comparison among field exposure systems, suggest that estimates of plant response to
17 O3 from open field air exclusion systems will be most representative of true field response,
18 as this system uses ambient O3 and no chambers. Next in accuracy will be open-top
19 chamber studies using filtered versus nonfiltered ambient O3. The drawback of these two
20 approaches is that plant responses to low ambient levels of O3 as might occur in many years
21 is quite subtle. To detect statistically significant differences between filtered and nonfiltered
22 chamber grown plants requires a high number of replications (Rawlings et al., 1988). Since
23 seasonal O3 concentrations vary greatly with geographic location and year, the collection of
24 data for yield loss estimates by either of these two methods will be much slower than if
25 O3-enriched chamber exposure systems are used.
26
27 5.2.2 Experimental Design and Data Analysis
28 The statistical design of an experiment is crucial in defining the types of analyses to
29 which the data can be subjected, and consequently the kinds of information about plant
30 response to O3. For these reasons, the information goals of a study must be clearly defined
31 in order to choose an appropriate experimental design. The various experimental designs and
December 1993 5-16 DRAFT-DO NOT QUOTE OR CITE
-------
1 approaches to data analysis have been well reviewed in the 1986 criteria document (U.S.
2 Environmental Protection Agency, 1986) and will not be repeated here. In summary, it is
3 clear that over the last 15 years, studies have more frequently used experimental designs
4 which generate data suitable for determination of regression-type dose-response relationships.
5 These exposure-response relationships generalize the mathematical relationship between the
6 plant parameter of interest and O3 exposure. Plant response to concentrations other than
7 those used in the experiment can be interpolated from these relationships and thresholds of
8 plant response can be determined (Ormrod et al., 1988). In the latter half of the NCLAN
9 program, the Weibull model was chosen to characterize yield response to O3 because of its
10 flexibility to describe a wide range of data patterns (Rawlings and Cure, 1985) and,
11 consequently to allow a common model to be fit when pooling data across years and sites
12 (Lesser etal., 1990).
13 Experimental designs for exposure-response relationships can easily be expanded so that
14 plant response to O3 and another factor at multiple levels can be determined. Because of the
15 need to contain each O3 treatment by a chamber, incomplete factorial designs are a more
16 efficient approach to multi-factor studies, leading to exposure-response surfaces (Allen et al.,
17 1987). Choosing the appropriate incomplete factorial design for a response surface study
18 requires forethought on whether all areas of the surface are of equal interest. For many
19 O3 plant response studies this is not so, as extremely high concentrations, although increasing
20 the precision with which plant response to lower concentrations is estimated, are not as likely
21 to occur in the ambient environment.
22 Because the U.S. Environmental Protection Agency (1986) decided to place greater
23 emphasis on damage (i.e., effects that reduce the intended human use of the plant) than on
24 injury, studies have more frequently used experimental designs which generate data suitable
25 for regression and treatment mean separation analyses to model and test the impact of O3 on
26 plant response. While the impact at current O3 levels is of primary interest and can be
27 studied using two O3 levels generated by charcoal-filtered (CF) and nonfiltered (NF)
28 treatments, the development of exposure-response models necessitates the use of additional
29 treatments at above ambient concentrations. The optimal number, range, and spacing of
30 treatment levels depends upon the anticipated exposure-response model but, in the case of the
31 Weibull and polynomial models, greater precision for estimation of relative yield loss at
December 1993 5_17 DRAFT-DO NOT QUOTE OR CITE
-------
1 ambient O3 concentration is obtained when the lowest treatment level is near zero and the
2 highest treatment level is well above ambient. For the Weibull model, the highest treatment
3 should correspond to a concentration where yield loss is at least 63% of the yield at zero
4 exposure (Basset and Rawlings, 1988; Rawlings et al., 1988).
5 When studying the impact of mixtures of pollutants on plant processes in chambers,
6 response surfaces can be generated from complete or incomplete factorial designs. These
7 designs have been shown to increase the precision and efficiency of estimating relative yield
8 loss at ambient concentrations. (Allen et al., 1987). The optimal design cannot be specified
9 a priori and necessitates the use of treatment levels from near zero to well above ambient for
10 each pollutant. However, response surface designs have not been widely used in pollutant
11 mixture studies to date. Nor have these designs been widely used to study the interaction
12 between pollutant exposure and quantitative environmental parameters such as light,
13 temperature or soil moisture. The interaction between O3 and phytotoxic concentrations of
14 other pollutants, in particular sulfur dioxide, has not been extensively studied because
15 instances of co-occurrence of O3 and other pollutants are not common in the United States.
16 An analysis of pollution monitoring data showed fewer than 10 periods of co-occurrence
17 between O3 and phytotoxie concentrations of sulfur dioxide during the growing season at the
18 sites where the two pollutants were monitored (Lefohn and Tingey, 1984; U.S.
19 Environmental Protection Agency, 1986).
20 Design and analysis of pollutant effects studies have used various characterizations of
21 exposure to determine optimum spacings of treatment levels and to relate exposure to
22 response. Most notably, the daytime mean concentration index (i.e., either M7 or M12) has
23 been adopted by the NCLAN program to determine the effects of O3 on plant response.
24 However, there has been considerable debate over the use of the mean index in
25 exposure-response modelling; the variety of ways to compute the characterizations of plant
26 exposure will be discussed elsewhere in this document (Section 5.5). When plant yield is
27 considered, plant response is affected by the concentration of exposure as well as by other
28 exposure-dynamic factors (e.g., duration, frequency, threshold, respite time) in combination
29 with physiological, biochemical, and environmental factors which may mask treatment effects
30 over the growing period. Research goals to understand the importance of exposure dynamic
31 factors have utilized experimental designs that apply two or more different patterns of
December 1993 5-18 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure that are equal on some scaling (e.g., total exposure). Experiments designed
2 specifically to address the importance of components of exposure may require the use of
3 exposure regimes that are not typical of the ambient environment.
4 In experiments in which treatments are quantitative, the traditional approach is to use
5 regression analysis which relates O3 exposure to plant response (U.S. Environmental
6 Protection Agency, 1986). The regression approach has been used to fit a common model to
7 combined data from replicate studies of the same species when it is reasonable to assume that
8 the primary cause of biological response is pollutant exposure and that differences in
9 environmental, edaphic, and/or agronomic conditions among sites do not significantly change
10 the shape of the regression relationships. When pooling data across sites and years,
11 additional terms for site and year effects are often included in the model as either fixed or
12 random components, depending upon the population of interest. Inferences over random
13 environments implies that the environments sampled by the experiments are representative of
14 the population of regions of interest under a variety of environmental conditions. In this
15 case, site and year effects are incorporated as random components in the model when fitting
16 a common model. The appropriate analysis is to use a mixed model analysis to fit an
17 exposure response model with variance components. This analysis has been recently used to
18 combine data from replicate studies of varying durations to test the importance of length of
19 exposure in influencing plant response; this approach is described elsewhere in the document
20 (Section 5.5).
21
22 5.2.3 Mechanistic Process Models
23 In addition to regression type models of plant response to O3, which are empirical and
24 statistical in nature, there are mechanistic-process models (Luxmoore, 1988; Kickert and
25 Krupa, 1991; Weinstein et al., 1991). The key difference between these two types of models
26 is how the changes in the dependent variable over time are handled. Empirical models treat
27 a time period (e.g., a growing season) as a single point, and report the response of the
28 dependent variable as a single point as well. Regression models may also over-simplify the
29 characteristics of an O3 exposure, in that the description of the O3 exposure is compressed
30 over time to a single number. The variety of ways to compute this single number will be
31 discussed elsewhere in this document (Section 5.5).
December 1993 5_ \ 9 DRAFT-DO NOT QUOTE OR CITE
-------
1 Mechanistic-process models on the other hand describe the rate of change of a variable
2 in response to the treatment (such as O3) with change in time (dy/dt). The latter type of
3 model has the potential to capture the interaction among plant age/stage of development,
4 variability of ambient exposure concentrations and plant response to Oj. For this reason,
5 mechanistic-process models have been rated much more highly than regression models for
6 their realism, scientific value and applicability to other locations (Kickert and Krupa, 1991).
7 However, compared to regression models, mechanistic-process models require more input
8 data and the input data are less accessible. The mechanistic-process models are more
9 complex than regression models, requiring more computer time and memory to develop.
10 The precision of the output regression models is greater than mechanistic-process models (for
11 interpolative examinations only), as is their ability to estimate response probabilities. The
12 authors conclude that the popularity of single-equation time-lumped models is related to the
13 fact that the studies of plant responses to O3 are more oriented to air quality standard setting
14 as an endpoint, rather than the physiological processes underlying plant responses. The
15 problems with process-based models are the necessity for some large assumptions (in place of
16 real data) and the lack of validation. Without that validation, using estimates from these
17 models is questionable. If these estimations are used, the uncertainties associated with them
18 must be identified and quantified.
19
20
21 5.3 SPECIES RESPONSE/MODE OF ACTION
22 5.3.1 Introduction
23 Plant adaptation to changing environmental factors or to stresses involves both short-
24 term physiological responses and long-term physiological, structural, and morphological
25 modifications. These changes help plants to minimize stress and maximize the use of
26 internal and external resources. A great deal of information is available on the physiology of
27 single leaves, however, relatively little is known about whole-plant systems and whether the
28 physiological mechanisms involved are initiated wholly within the leaf or are the result of
29 whole-plant interactions (Dickson and Isebrands, 1991).
30 The many regulatory systems contained in leaves change both as a function of leaf
31 development and in response to different environmental stresses. Leaves function as the
December 1993 5-20 DRAFT-DO NOT QUOTE OR CITE
-------
1 major regulators of anatomical and morphological development of the shoot and control the
2 allocation of carbohydrates to the whole plant (Dickson and Isebrands, 1991). This section
3 discusses the movement of O3 into plant leaves and what is known about their biochemical
4 and physiological responses.
5 Movement of O3 into plant leaves involves both a gas and a liquid phase. The
6 phytotoxic effects of air pollution on plants appear only when sufficient concentrations of the
7 gas diffuse into the leaf interior and pass into the liquid phase within the cells. Therefore, to
8 modify or degrade cellular function O3 must diffuse in the gas-phase from the atmosphere
9 surrounding the leaves through the stomata into the air spaces and enter into the cells after
10 becoming dissolved in the water coating the cell walls (U.S. Environmental Protection
11 Agency, 1986). The exact site or sites of action are not known. Biochemical pathways are
12 closely interrelated, and at the present time, we do not have sufficient knowledge of all the
13 control and regulatory mechanisms (Heath 1987). The previous criteria document
14 summarized quite well the overall processes controlling plant response to O3:
15 "The response of vascular plants to O3 may be viewed as the culmination of a
16 sequence of physical, biochemical, and physiological events. Ozone in the
17 ambient air does not impair processes or performance, only the O3 that
18 diffuses into the plant. An effect will occur only if a sufficient amount of
19 O3 reaches the sensitive cellular sites within the leaf. The O3 diffuses from
20 the ambient air into the leaf through the stomata, which can exert some control
21 on O3 uptake to the active sites within the leaf. Ozone injury will not occur if
22 (1) the rate of O3 uptake is sufficiently small that the plant is able to detoxify
23 or metabolize O3 or its metabolites; or (2) the plant is able to repair or
24 compensate for the O3 impacts (Tingey and Taylor, 1982). The uptake and
25 movement of O3 to the sensitive cellular sites are subject to various
26 physiological and biochemical controls" (U.S. Environmental Protection
27 Agency, 1986).
28
29 Responses to O3 exposure that have been measured include reduced net CO2 exchange
30 rate (photosynthesis minus respiration), increased leaf/needle senescence, increased
31 production of ethylene, and changes in allocation patterns. Overall understanding of the
32 response of plants to O3 has been refined since the last criteria document (U.S.
33 Environmental Protection Agency, 1986). Increased emphasis has been placed on the
34 response of the process of photosynthesis to O3, on identification of detoxification
35 mechanisms, and on changes in biomass (sugar, carbohydrate) allocation.
December 1993 5_2i DRAFT-DO NOT QUOTE OR CITE
-------
1 As indicated above, entry of O3 into leaves involves the gas-phase external to the plant
2 and the liquid—phase within the leaf cells. A precondition for O3 to affect plant function is
3 that it be absorbed into the tissues. We will divide Oj uptake into two components,
4 adsorption to surfaces and absorption into tissues. Adsorption will affect surface materials
5 (e.g., cuticles, and have little direct affect on physiological processes), whereas
6 O3 absorption can affect physiological function, if O3 is not first detoxified. In the following
7 section the processes that control movement of O3 into the plant canopy and then into the
8 leaf will be examined.
9
10 5.3.2 Ozone Uptake
11 Uptake of O3 in a plant canopy is a complex process involving adsorption of O3 to
12 surfaces (stems, leaves, soil) and absorption into tissues, primarily leaves (Figure 5-2).
13 Movement of O3 from the atmosphere to the leaf involves micrometeoroloical processes
14 (especially wind) and the architecture of the canopy (including the leaves). Within the
15 canopy O3 can be scavenged by chemicals in the atmosphere (Kotzias et al., 1990; Gab
16 et al., 1985; Becker et al., 1990; Yokouchi and Ambe, 1985; Bors et al., 1989; Hewitt
17 et al., 1990) however the products of these reactions may themselves be phytotoxic (Kotzias
18 et al., 1990; Gab et al., 1985; Becker et al., 1990; Hewitt et al., 1990). The extent to
19 which these scavenging processes affect O3 absorption by leaves is not well known.
20 Absorption of O3 by leaves is controlled in large part by canopy and leaf conductances. The
21 former is a shorthand for the complex of microclimate and canopy architecture that control
22 movement of O3 from the bulk air to the leaf. Leaf conductance is determined by leaf
23 boundary layer conductance and stomatal conductance. In this section we will examine
24 theoretical and empirical studies on O3 uptake at the canopy and leaf levels.
25
26 5.3.2.1 Ozone Uptake by Plant Canopies
27 Integration of O3 uptake at the stand level requires attention to several levels of
28 organization (Enders et al., 1992; Hosker and Lindberg, 1982), because uptake at this level
29 includes not only absorption by leaves, but adsorption by stems, the soil, and other structures
30 with which O3 can react. While the actual pathway, and therefore conductance, will vary
31 within the canopy depending on position and wind profile, an integrated average conductance
December 1993 5-22 DRAFT-DO NOT QUOTE OR CITE
-------
Atmospheric Processes
->/ Ozone in 1
I atmosphere]
I Canopy
I Boundary Layer
Leaf Processes/
Mode of Action
Plant Response
Ecosystem Response
Canopy
Conductance
Ozone in
Vegetation
Canopy
IStomatal
Conductance!
Figure 5-2. Uptake of O3 from the atmosphere. Ozone is moved from the atmosphere
above the canopy boundary layer into the canopy primarily by turbulent
flow of air. Canopy conductance, controlled by the complexity of the
canopy architecture and the wide distribution within the canopy, is a
measure of the ease with which gases move into the canopy. Within the
canopy, O3 can be adsorbed by surfaces as well as being absorbed into the
foliage. Foliage absorption is controlled by two conductances, leaf
boundary layer and stomatal, which together determine leaf conductance.
Solid black arrow denotes O3 flow; gray stipled arrows connect factors and
processes that affect O3 flow; open arrows connect major processes.
Shaded boxes at the left are those processes described in the figure. The
shaded rounded box is end of pathway on this figure.
1 is frequently used to describe canopy conductance (Monteith and Unsworth, 1990). For most
2 tree species canopy conductance tends toward high values while for crops it tends to be low.
3 Two general approaches have been used to estimate O3 uptake by a plant stand:
4 (1) measurement of gradients over the canopy using micrometeorological methods and
December 1993
5-23
DRAFT-DO NOT QUOTE OR CITE
-------
1 (2) simulation of canopy conductance. The results of the two methods are generally different
2 because the micrometeorological techniques include O3 uptake by all surfaces whereas
3 simulation only accounts for O3 absorbed by the surfaces simulated, primarily the foliage.
4 Two micrometeorological methods, Bowen ratio and eddy correlation, have been used
5 to calculate canopy O3 uptake. The Bowen ratio assumes a constant relationship (the Bowen
6 ratio) between heat and water vapor fluxes (i.e., sensible and latent heat), then calculates
7 O3 uptake assuming a constant relation between water vapor and O3 fluxes (Leuning et al.,
8 1979a). The eddy correlation technique requires more elaborate instrumentation for
9 measurement of variation in temperature, water vapor, and O3 concentration over time and
10 has stringent site requirements (Weseley et al., 1978).
11 Wesely et al. (1978), using eddy correlation, found a strong diurnal variation in the
12 deposition velocity (the inverse of canopy conductance) and Oj flux over a corn canopy.
13 They also found evidence that 20 to 50% of the flux was to the soil and to the surface of the
14 canopy. Ozone flux to a dead corn canopy also had a diurnal variation, but a lower
15 magnitude, probably reflecting the absence of uptake through the response. Single time
16 measures of deposition velocity, or canopy resistance, have been taken in a Gulf Coast pine
17 forest (54 s m" ; Lenschow et al., 1982) and in a New Jersey pine forest (120 and
18 300 s cm •. Greenhut. 1983. Ozone uptake in a maple forest varied diurnally in a pattern
19 explainable by variation in leaf conductance and O3 concentration (Fuentes et al., 1992).
20 Ozone flux below the tree canopy at 10 in was about 10% of the flux above the canopy at
21 33 m. Measurements in specially constructed chambers showed that O3 uptake, as well as
22 photosynthesis, could occur when the foliage was wet (Fuentes and Gillespie, 1992). The
23 fact that wet leaves could take up significant CO2 is evidence that the response were not
24 blocked by the water on the leaf surface. This result is counter to assumptions made in
25 earlier work (Baldocchi et al., 1987) in which water on the surface of the leaf was presumed
26 to interfere with O3 uptake.
27 Simulation of canopy conductance requires scaling uptake from individual leaves to
28 individual trees to that of a stand using a combination of canopy models (one for each
29 species) and a stand model to handle interactions among individuals. Several assumptions
30 are required for this approach: (1) the primary sink for O3 is the foliage, (2) variation in
31 stomatal conductance can be simulated through the canopy using either direct measurements
December 1993 5-24 DRAFT-DO NOT QUOTE OR CITE
-------
1 or models, and (3) canopy/plant models adequately simulate response when competition is
2 occurring.
3 Leuning et al. (1979a,b) used a simple model to estimate canopy uptake in corn and
4 tobacco. Comparison of the results of these simulations with estimates using the Bowen ratio
5 technique indicated that about 50% of the O3 absorbed by the stands entered the leaves.
6 Baldocchi et al. (1987) presented a model for canopy uptake of O3 that incorporated stomatal
7 function, some aspects of canopy architecture, and soil uptake. The results of the simulation
8 of O3 uptake by a corn canopy correlated well with estimations using the Bowen ratio, but
9 tended to overestimate the magnitude. These authors point out that results of model
10 simulation are quite sensitive to the assumptions used. As part of a series of simulations,
11 Reich et al. (1990) explored the effects of different O3 exposures (daily average
12 O3 concentrations of 0.035, 0.05, 0.065, and 0.080 ppm) on canopy carbon gain in a mixed
13 oak-maple forest. Depending on the response function and O3 exposure used, reductions in
14 carbon gain were between 5 and 60%. An important result of these simulations is that the
15 effect of O3 was strongest in the upper layer of the canopy, where most of the photosynthesis
16 occurred. While all these simulations provide some interesting insights into how O3 uptake
17 (and response) varies with time and exposure, data for validating the models is still needed.
18 In summary, O3 uptake (absorption to surfaces and absorption by tissues) by plant
19 canopies has been measured only a few times. The results are consistent with the hypothesis
20 that stomatal conductance plants a major role in the process. Modelling of O3 absorption by
21 leaves provides a means of assessing our understanding of the processes controlling
22 O3 absorption. Combining direct measurements over canopies with modelling will provide a
23 means for assessing the dynamics of O3 uptake in a canopy.
24
25 5.3.2.2 Ozone Absorption by Leaves
26 The importance of stomatal conductance for the regulation of O3 uptake by a canopy
27 has been hypothesized for some time (Heck et al., 1966; Rich et al., 1970). Absorption of
28 O3 by leaves is primarily controlled by stomatal conductance which varies as a function of
29 stomatal aperture (Figure 5-3). Kerstiens and Lendzian (1989) found that the permeability to
30 O3 of cuticles from several species was about 0.00001 that of open response. Movement of
31 guard cells, which control stomatal opening, are affected by a variety of environmental
December 1993 5-25 DRAFT-DO NOT QUOTE OR CITE
-------
Light
Humidity
Figure 5-3. Movement of gases into and out of leaves is controlled primarily by the
stomata—small openings in the. leaf surface whose aperture is controlled by
two guard cells. Guard cells respond to a number of external and internal
factors, including light, humidity, CO2, water stress. In general, the
stomata open in response to light and increasing temperature and close in
response to decreasing humidity, increased CO2, and increasing water
stress. They may also close in response to air pollutants, such as O3.
1 and internal factors, including light, humidity, CO2 concentration, and water status of the
2 plant (Zeiger et al., 1987; Kearns and Assmann, 1993). Air pollutants, including O3, may
3 also affect stomatal function (U.S. Environmental Protection Agency, 1986).
4 For O3 to be absorbed by a leaf, O3 must be present in the atmosphere surrounding the
5 leaf, and the stomata must be open. Any factor that affects stomatal opening affects
6 O3 absorption (Figure 5-3). Under drought conditions when stomatal conductance is reduced
7 the relative effect of O3 is less when compared with well-watered controls (Tingey and
December 1993
5-26
DRAFT-DO NOT QUOTE OR CITE
-------
1 Hogsett, 1985; Flagler et al., 1987; Temple et al., 1993, also see Section 5.4). Low
2 humidity has been shown to modify plant response to O3 (McLaughlin and Taylor, 1981),
3 presumably due to reduced O3 absorption (Wieser and Havranek, 1993).
4 To calculate O3 absorption, some estimate of the internal O3 concentration must be
5 made. In earlier work, a finite O3 concentration was assumed to exist in the intercellular air
6 space of the leaf (Benett et al., 1973; Tingey and Taylor, 1982; Lange et al., 1989).
7 Estimating this concentration is difficult because the rate of Oj absorption into the leaf must
8 be known. Recently Laisk et al. (1989) presented evidence that this concentration is near
9 zero, a result that is consistent with the highly reactive nature of O3. Further studies on
10 other species must be made to test the hypothesis that internal O3 concentration is negligible
11 in leaves.
12 The other component of absorption, O3 concentration outside the leaf, may vary greatly
13 with time of day and season (Chapter 4). Data on the effect of variations in O3 profile—
14 from constant concentrations to equal daily peaks to variable (episodic) peaks—suggest that
15 those profiles that have periodic high concentrations have a greater effect than those with low
16 peaks even though the exposure is equivalent (Hogsett et al., 1985a; Musselman et al., 1986;
17 see Section 5.6). Taylor and Hanson (1992) show how variations in conductance can affect
18 O3 absorption and conclude that conductances in and near the leaf surface have a major
19 influence on absorption of O3. Figure 5-4 shows a simulation of the effect of diurnal
20 variation hi stomatal conductance and O3 concentration on the O3 absorbed into the leaf.
21 Amiro and Gillespie (1985) found that cumulative O3 absorption correlated with visible
22 injury in soybean. Weber et al. (1993) found that rate of uptake may play an important role
23 in the response of ponderosa pine. The roles of cumulative uptake versus uptake rate have
24 not been clarified and need further study.
25 Absorption of O3 by leaves depends on variations in both stomatal conductance and
26 O3 concentration. The highly reactive nature of O3 makes measuring its absorption difficult;
27 therefore, models of stomatal conductance are used along with O3 concentrations to estimate
28 O3 absorption. The relative importance of absorption rate versus cumulative absorption is
29 not known at present.
30
31
December 1993 5-27 DRAFT-DO NOT QUOTE OR CITE
-------
Ozone Profiles
10 15
Time (h)
20
-14:00 peak —12:00 peak-—16:00 peak —constant
Variation in Conductance
Simulation
10 15 20
Time (h)
B
*
10
Simulated Absorption
I
0)
§
&
6
*
^
0
10 15
Time (h)
20
- 14:00 peak —12:00 peak - -16:00 peak —constant
Figure 5-4. Simulation of the effects of diurnal variation in stomatal aperture and in
O3 concentration on O3 uptake. A. Simulated conductance. B. Diurnal
O3 concentrations. C. O3 Uptake.
December 1993
5-28
DRAFT-DO NOT QUOTE OR CITE
-------
1 5.3.3 Resistance Mechanisms
2 Resistance mechanisms can be divided into two types: (1) exclusion from sensitive
3 tissue and (2) detoxification near or in sensitive tissue. For leaves the former involve
4 response and cuticles, and the latter various potential chemical and biochemical reactions that
5 chemically reduce O3 in a controlled manner. While these systems potentially provide
6 protection against O3 damage to tissue physiology, they come at some cost, either in the
7 reduction in photosynthesis in the case of stomatal closure or carbohydrate used to produce
8 detoxification systems.
9 Injury to leaf/needle cuticles does not appear to have a major affect on leaf function; at
10 least the data are not consistent. Barnes et al. (1988) found that O3 exposure could damage
11 leaf cuticles. However Lutz et al. (1990) found no consistent changes in Norway spruce.
12
13 5.3.3.1 Stomatal Limitation
14 As noted above, response can be affected by a wide variety of environmental factors.
15 Some early research showed a decrease in leaf conductance (Figure 5-3) with O$ exposure
16 implying a direct effect of O3 on stomatal conductance (U.S. Environmental Protection
17 Agency, 1986). In studies at high O3 concentrations (>300 ppb) stomatal response was
18 rapid (Modlau et al., 1990). In other studies reduction in conductance in response to
19 O3 required hours to days of exposure (Dann and Pell, 1989; Weber et al., 1993). Several
20 studies have shown that discrimination against 13C in C3 plants decreases with
21 O3 fumigation (Okano et al., 1985; Martin et al., 1988; Greitner and Winner 1988; Saurer
22 et al., 1991; Matyssek et al., 1992). These data are consistent with an increased restriction
23 to diffusion of CO2 into the leaf (Farquhar et al., 1989). However, Matyssek et al. (1992)
24 and Saurer et al. (1991) found that internal CO2 increased with O3 exposure and water use
25 efficiency decreased, both the opposite of expectation, which indicates that photosynthesis
26 decreased relatively more than conductance. While response limit O3 uptake, whether this
27 limitation results from direct effects of O3 on guard cell function, indirectly through effects
28 on internal CO2 concentration, or a combination has not be clearly demonstrated.
29
30
December 1993 5_29 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.3.3.2 Detoxification
2 When O3 enters a cell, several highly reactive compounds can be produced (e.g.,
3 superoxide, free radicals, and peroxides) (Heath, 1987). The effects of these compounds
4 depends on their reactivity, mobility, and half-life. For detoxification to occur oxidant and
5 antioxidant must occur proximately. In addition, the rate of production of antioxidant must
6 be a significant portion of the rate of oxidant entry into the system of effective detoxification
7 to occur. Two general kinds of detoxification systems have been reported in plants:
8 (1) those that utilize reductants (e.g., ascorbate) to reduce O3, and (2) those that utilize
9 enzymes (superoxide dismutase). In either case excess oxidizing power is dissipated in a
10 controlled manner, effectively protecting the tissue from damage. These systems probably
11 developed to protect cells from photooxidation which can occur, for example, at low
12 temperatures (Powles, 1984).
13 Several antioxidants have been reported, the most studied being ascorbate and
14 glutathione. Much of this work has occurred since the 1986 criteria document (U.S.
15 Environmental Protection Agency, 1986). Alscher and Amthor (1988) reviewed the
16 literature in this area. In the chloroplast, the process requires NADPH and may be a cause
17 for the transient reduction in photosynthesis observed in some studies (Alscher and Amthor,
18 1988).
19 Evidence for the participation of antioxidants in protecting cells from O3 injury is
20 primarily indirect (i.e., changes in levels of antioxidants or of associated enzymes). In red
21 spruce, glutathione levels increased in year-old needles in response to O3, but not in current
22 year needles (Hausladen et al., 1990; Madamanchi et al., 1991). Dohmen et al. (1990)
23 found increased concentrations of reduced glutathione in Norway spruce in response to
24 long-term O3 fumigation. In a poplar hybrid, total glutathione increased with O3 fumigation;
25 however, the ratio of reduced to oxidized forms declined, indicating oxidation of glutathione
26 was possibly stimulated by O3 (Sen Gupta et al., 1991). Mehlhorn et al. (1986) found that
27 both glutathione and ascorbic acid increased with O3 fumigation in white fir (Abies alba)
28 and Norway spruce. The potential for ascorbic acid to protect cells from O3 damage was
29 explored by Chameides (1989). He concluded that such protection was possible if ascorbic
30 acid occurred in the apoplast at sufficient concentrations and sufficient production rates;
31 however, experimental data are needed to test the hypothesis.
December 1993 5-30 DRAFT-DO NOT QUOTE OR CITE
-------
1 The response of enzymes involved in detoxification is not clear. Activities of enzymes
2 involved in antioxidant production increased in response to O3 in one study (Price et al.,
3 1990); however, in several others no effect was found (Madamanchi et al., 1992; Pitcher
4 et al., 1991; Anderson et al., 1992; Nast et al., 1993). Activity of superoxide dismutase, an
5 enzyme that can reduce one of the products of O3 interaction with the cytoplasm, can be
6 increased by O3 fumigation (Alscher and Amthor, 1988; Sen Gupta et al., 1991). There are
7 both cytosolic and chloroplastic forms of this enzyme, but the role the different forms plant
8 in detoxification of O3 is not clear. Teppermann and Dunsmuir (1990) and Pitcher et al.
9 (1991) found that increased production of SOD had no effect on resistance to 63 in tobacco.
10 The extent to which these detoxification systems can protect tissue from O3 damage is
11 unknown. However, "If plants have detoxification mechanisms which are kinetically limited,
12 the rate of O3 uptake may be important, so that even an integrated absorbed dose may be
13 insufficient to account for observed responses" (Cape and Unsworth, 1988). Potential rates
14 of detoxification for given tissues are needed to estimate the importance of these systems to
15 overall O3 response. In addition, the sites in which the detoxification systems occur need to
16 be identified.
17
18 5.3.4 Physiological Effects of Ozone
19 The initial reactions of O3 with cellular constituents are not known. The high reactivity
20 and non-specificity of O3 reactions coupled with the absence of a useful isotopic tag for
21 O3 make studies of the initial reactions difficult at best. The data on changes in biochemical
22 function resulting from O3 exposure probably represent effects one or more steps beyond the
23 initial reactions. Nonetheless, we do have data which indicate the wide range of cellular
24 processes that can be affected by O3.
25 Ozone that has not been neutralized by one of the detoxification systems (Figure 5-5)
26 acts first at the biochemical level to impair the functioning of various cellular processes
27 (Tingey and Taylor, 1982; U.S. Environmental Protection Agency, 1986). The result of
28 these impairments are reflected in integrated changes in enzyme activities, membrane
29 function, and energy utilization (Queiroz, 1988). Several related papers have shown that the
30 activity of the primary carboxylating enzyme (RuBP-carboxylase) is reduced by O3 exposures
31 in the range of those measured at some sites (Dann and Pell, 1989; Enyedi et al., 1992; Pell
December 1993 5.31 DRAFT-DO NOT QUOTE OR CITE
-------
Atmospheric Processes
Ecosystem Response
Reduced
Photosynthesis
Leaf Processes/
Ozone Uptake
Physiological Response to Ozone
Increased
Foliage
Senescence
Repair
Replacement
Reduced Canopy
Carbohydrate
Production
Respiratory
Losses
Figure 5-5. Effects of O3 absorption into a leaf. Once inside the leaf Oj can have a
number of effects all of which affect carbohydrate production and
utilization. Reduced photosynthesis, increased leaf senescence, production
of detoxification systems, and increased respiration (both maintenance and
growth) reduce the amount of carbohydrate available for allocation.
Compensation through production of new leaves, for instance, can counter
some or all of these effects depending on the O3 exposure, physiological
state of the plant, and the species. Integration of these processes leads to
changes in the amount of carbohydrate available for allocation from the
canopy. Solid black arrows denote O3 flow; gray stipled arrows show the
cascade of effects of O3 absorption on leaf function.
1 et al., 1992; Landry and Pell, 1993). Membrane injury has been found in some experiments
2 using acute levels of O3 (Heath, 1987). Chronic exposure can lead to changes in lipid
3 composition and changes in cold resistance (Brown et al., 1987; Davison et al., 1987;
4 DeHayes et al., 1991; Lucas et al., 1988; Wolfenden and Wellburn, 1991). Recently, Floyd
December 1993
5-32
DRAFT-DO NOT QUOTE OR CITE
-------
1 et al. (1989) have been shown that O3 can affect nuclear DNA through the formation of
2 aducts.
3 Changes in the in vivo concentrations of various growth regulators in response to
4 O3 exposure could have important consequences for plant function. However, the effects of
5 O3 on levels and activities of growth regulators have not been studied extensively. Ozone
6 has been shown to stimulate ethylene production and inhibitors of ethylene production have
7 been found to reduce the effects of O3 in short-term experiments (Pell and Puente, 1986;
8 Rodecap and Tingey, 1986; Taylor et al., 1988; Mehlhorn et al., 1991; Telewski, 1992;
9 Langebartels et al., 1991; Mehlhorn and Wellburn, 1987; Kargiolaki et al., 1991; Reddy
10 et al., 1993). Ethylene is produced during ripening of fruit, during periods of stress, and
11 during senescence (Abeles et al., 1992). Increased levels of ethylene in the leaves could play
12 a role in the early senescence of foliage. In some cases there is a correlation between
13 ethylene production and O3 sensitivity; however, the relationship is complex and makes use
14 of ethylene production as an index of sensitivity problematic (Pell, 1988).
15 Abscisic acid plays an important role in stomatal function (Davies et al., 1980).
16 Atkinson et al. (1991) found that response from O3 fumigated leaves were less sensitive to
17 ABA than control leaves. These data could explain the observation that stomatal function is
18 unpaired by long-term O3 exposure. Kobriger et al. (1984) found no effect of O3 on
19 whole-leaf content of abscisic acid, but changes in compartmentation could not be ruled out.
20 Physiological effects of O3 uptake are manifest in two primary ways: (1) reduced net
21 photosynthesis and (2) increased senescence (Figure 5-5). Both decreased photosynthesis and
22 increased leaf senescence result in the loss of capacity for plants to form carbohydrates,
23 thereby having a major impact on the growth of the plant (Figure 5-6).
24 Ozone-induced reduction in net photosynthesis has been known for some time (U.S.
25 Environmental Protection Agency, 1986). Changes in stomatal conductance, in
26 photosynthetic capacity, carbohydrate allocation and respiration have been documented. The
27 relationship between O3 exposure and photosynthesis is not well known. Photosynthesis
28 provides plants with the energy and structural building blocks necessary for their existence.
29 The photosynthetic capacity of a plant is an important aspect of plant response to stresses in
30 natural environments and is strongly associated with leaf nitrogen content and with water
December 1993 5.33 DRAFT-DO NOT QUOTE OR CITE
-------
Atmospheric Processes
Canopy Processes
Leaf Processes/
Ozone Uptake
Leaf Processes/
Mode of Action
Ecosystem Response
Carbohydrate
Allocation
i
\
f Compensation
Figure 5-6. Effect of O3 on plant function and growth. Reduction in carbohydrate
allocation affects the pool of carbohydrates available for growth. Changes
in relative growth rate of various organs as a function of O3 exposure
suggest that allocation patterns of carbohydrate is affected. Solid black
arrows denote where O3 absorption affects the allocation processes of the
plant; gray stipled arrows show the cascade to plant growth.
1 movement. Both resources are essential if the process is to occur and involves the allocation
2 of carbohydrates from the leaves to the roots for nitrogen acquisition and water uptake. Leaf
3 photosynthetic capacity is also age dependent. As the plant grows, the canopy structure
4 changes altering the amount and angle of light hitting a leaf. Allocation of carbohydrates and
5 nutrients to new loaves is especially important in stimulating growth production (Pearcy
6 et al., 1987). Reductions in photosynthesis are likely to be accompanied by as shift in
7 growth pattern which favors shoots and by an increase or decrease in life span (Winner and
8 Atkinson, 1986). Therefore, alteration of the processes of photosynthesis and carbohydrate
December 1993
5-34
DRAFT-DO NOT QUOTE OR CITE
-------
1 allocation affects plant response to stresses such as O3. Reduction in photosynthesis, reduced
2 carbohydrate formation and allocation to leaf repair or to new leaf formation decreases the
3 availability of carbohydrates, alters the normal allocation pattern and, therefore, all aspects
4 of plant growth and reproduction (Figure 5-6). The effects of a reduction in photosynthesis
5 on growth and reproduction was discussed in the previous criteria document (U.S.
6 Environmental Protection Agency, 1986).
7 Carbohydrate production by a single plant is controlled not only by photosynthetic
8 capacity of the foliage but also by the amount and distribution of that foliage. Stow et al.
9 (1992) and Kress et al. (1992) found that O3 exposure affected needle retention in loblolly
10 pine. Similar data have been reported for slash pine (Byres et al., 1992). Keller (1988) and
11 Matyssek et al. (1993a,b) reported increased senescence with increased 63 exposure in
12 aspen, as did Wiltshire et al. (1993) in apple. Replacement of injured leaf tissue has been
13 reported for some species when they are exposed to low O3 concentrations (Held et al.,
14 1992; Temple et al., 1993). Temple et al. (1993) also found increased photosynthetic
15 capacity of new needles in O3 treatments compared to controls.
16 Few direct effects of O3 have been found outside leaves. Kargiolaki et al. (1991) found
17 that intumescences appear on stems of poplar after 72 days of O3 fumigation (70 to 80 ppb).
18 Ozone probably enters the stem through the lenticles that occur on the surface of the stem
19 and allow direct exchange of gases between the stem and the air. The consequence of this
20 response to O3 is not clear; however, it may be related to the reduction in phloem transport
21 rate observed in loblolly pine (Spence et al., 1990).
22
23 5.3.4.1 Carbohydrate Production and Allocation
24 The importance of photosynthesis and carbohydrate allocation in plant growth and
25 reproduction has been pointed out previously. The patterns of carbohydrate allocation
26 directly affect growth rate. Plants require a balance of resources to maintain optimal growth,
27 however, in natural environments optimal conditions seldom occur. Therefore, plants must
28 compensate for differences in resource availability as well as for environmental stresses.
29 They do this by changing the way they allocated carbohydrates (Chapin et al., 1987). Each
30 response to stress affects the availability of carbohydrates for allocation from the leaves
31 (Figure 5-5). The carbohydrate pool is affected both by a reduction in the carbohydrate
December 1993 5.35 DRAFT-DO NOT QUOTE OR CITE
-------
1 produced and by a shift of carbohydrate to repair and replacement processes. The effect is
2 particularly noticeable in the roots where O3 exposure significantly reduces available
3 carbohydrate (Andersen et al., 1991; Andersen and Rygiewicz, 1991). Effects on leaf and
4 needle carbohydrate content have ranged from a reduction (Barnes et al., 1990; Miller et al.,
5 1989), to no effect (Alscher et al., 1989), to an increase (Luethy-Krause and Landolt, 1990).
6 Cooley and Manning (1987) reviewed the literature on carbohydrate partitioning and noted
7 that "storage organs...are most affected by O3-induced partitioning changes when
8 O3 concentrations are in the range commonly observed in polluted ambient air."
9 The above discussion supports the information in the previous criteria document (U.S.
10 Environmental Protection Agency, 1986), which pointed out that roots were usually more
11 affected by O3 exposures than the shoots. Studies by Miller et al. (1969), Tingey et al.
12 (1976a), McLaughlin et al. (1982) and Price and Treshow, (1972) were cited in support of
13 the view. Miller et al., noted reduction in photosynthesis was accompanied by decreases in
14 sugar and polysaccharide fraction in injured needles of ponderosa pine seedlings and altered
15 allocation of carbohydrates. Exposure were for 30 days, 9 h/day to concentrations of 0.15,
16 0.30 or 0.40 ppm. These exposures reduced photosynthesis by 10, 70 and 85%,
17 respectively. The observations of Tingey et al. (1976) indicated that O3 exposures
18 differentially affected metabolic pools in the roots and tops of ponderosa pine seedlings
19 grown in field. Further, this study indicated that the amounts of soluble sugars, starches,
20 and phenols tended to increase in the tops and decrease in plant roots of ponderosa pine
21 seedlings exposed to 0.10 ppm O3 for 6 h/day for 20 weeks. The sugars and starched stored
22 in the tree roots were significantly less than those in the roots of controls. In another study
23 cited in the 1986 document, McLaughlin et al. (1982) also observed the reduced availability
24 of carbohydrate for allocation to the roots and stated that the result was reduced vigor and
25 enhanced susceptibility of trees to root diseases. Loss of vigor was due to a sequence of
26 events, including premature senescence and loss of older needles, lower gross photosynthetic
27 productivity, and reduced photosynthate (carbohydrates) available for growth and
28 maintenance, that were associated with exposure to O3. Carbon-14 transport patterns also
29 indicated changes in carbon allocation. Older needles were found to be the source of
30 photosynthate for new needle growth in the spring and were storage sinks in the fall.
December 1993 5-36 DRAFT-DO NOT QUOTE OR CITE
-------
1 Retention of l4C-pholosynthale by foliage and branches of sensitive trees indicated that
2 allocation to the trunks and roots was reduced.
3 Lost carbohydrate production has effects throughout the plant (Figure 5-6). The roots
4 and associated mycorrhizal fungi are especially susceptible to reduced carbohydrate
5 availability and quite frequently show the greatest decline in growth (Adams and O'Neill
6 1991; Edwards and Kelly, 1992; McQuattie and Schier, 1992; Meier et al., 1990; Taylor and
7 Davies, 1990). However, in some cases, increased mycorrhizal formation has been reported
8 (Goriseen et al., 1991; Reich et al., 1985). It might be expected that reduced allocation to
9 roots would affect shoot growth through increased susceptibility to water stress, reduced
10 nutrient availability (Flagler et al., 1987), and reduced production of growth factors (Davies
11 and Zhang, 1991; Letham and Palni, 1983). Effects on production and retention of leaves
12 and needles were described above. Effects on stem growth have been found in tree species
13 (Hogsett et al., 1985b; Mudano et al., 1992; Pathak et al., 1986; Matyssek et al., 1992;
14 Matyssek et al., 1993b). Changes in canopy density, root/shoot ratio, and stem growth will
15 affect the functioning of the plant and may make plants more susceptible to environmental
16 stresses, such as drought and nutrient limitation, that are characteristic of many ecosystems.
17
18 5.3.4.2 Compensation
19 Compensatory responses occur as plants attempt to minimize the effects of stress.
20 Responses include adjustments to changes in physiological processes (e.g., photosynthetic
21 capacity and foliage production) that tend to counteract the effects of O3 absorption by the
22 leaves. Pell et al. (in press) have reviewed the extensive literature produced in the ROPIS
23 experiment. A wide range of compensatory responses have been identified, especially
24 reallocation of resources leading to increased relative growth in the shoot compared to the
25 root (see above). Compensation can take the form of production of new tissue (e.g., leaves)
26 to replace injured tissue and/or biochemical shifts, including increased photosynthetic
27 capacity in new foliage.
28 Changes in respiratory rate have been attributed to such repair processes (U.S.
29 Environmental Protection Agency, 1986). Recent studies have found stimulation of dark
30 respiration in Norway spruce (Barnes et al., 1990; Wallin et al., 1990) and bean (Amthor,
31 1988; Amthor and Gumming, 1988; Moldau et al., 1991). Repair of membranes (Sutton and
December 1993 5-37 DRAFT-DO NOT QUOTE OR CITE
-------
1 Ting, 1977; Chevrier et al., 1988, 1990) and of injured enzymes are two probable reasons
2 for increased respiration. Ozone has been shown to increase the ATP/ADP ratio, which is
3 consistent with increased respiratory activity (Weidmann et al., 1990; Hampp et al., 1990).
4 As in the case of detoxification, the importance of repair processes in the overall
5 carbohydrate budget of the plant and of their influence of apparent threshold is unknown.
6 Recovery of photosynthetic capacity after O3 exposure has been noted in some studies.
7 Early work indicated that recovery of photosynthetic capacity could occur after exposure of
8 high concentrations (>0.25 ppm) of O3 (e.g., Botkin, 1971, 1972). Dann and Pell (1989)
9 found that photosynthetic rate, but not Rubisco activity, recovered within a few days in
10 potato after exposure to 0.2 ppm O3. In ponderosa pine, photosynthetic rates in O3 treated
11 needles recovered to that of controls within 40-50 days (Weber et al., 1993). To what
12 extent this recovery can offset losses in carbohydrate gain is not known nor is the
13 mechanism.
14 Replacement of injured foliage (see Section 5.3.4) is another method to counteract the
15 effects of O$ exposure. The extent to which increased leaf/needle production and increased
16 photosynthetic capacity in the new foliage compensates for O3 injury is not known.
17 The importance of various compensation mechanisms is not sufficiently well-known to
18 allow an estimate of the degree to which they might mitigate the effect of O3. The fact that
19 increases in photosynthesis and in leaf production have been measured indicates that these
20 processes, at least, may be important.
21
22 5.3.5 Role of Age and Size Influencing Response to Ozone
23 Plant age, physiological state of growth and the frequency to which plants are exposed
24 play an important role in plant response to O3. In annual species, effects of O3 exposure can
25 result in a reduction in the amount and size of seed production through changes in allocation
26 that occur over the years. In perennial species, reduced growth occurs due to a reduction in
27 the amount of carbohydrates available for storage. In both cases, carbohydrates will have
28 been used to help the plants minimize the stress of O3 exposure. It should also be noted that
29 in absence of O3 exposure, in hardwoods leaf photosynthetic capacity reaches an early peak
30 and then gradually declines until senescence. Conifers, on the other hand retain leaves
31 longer than one year (Pye, 1988).
December 1993 5-38 DRAFT-DO NOT QUOTE OR CITE
-------
1 Carry over effects of O3 have been documented in growth of tree seedlings (Hogsett
2 et al., 1989; Sasek et al., 1991; Temple et al., 1993) and in regrowth of roots (Andersen
3 et al., 1991). Accumulation of these effects will affect survival and ability to reproduce.
4 Data on cumulative effects of multiple years of O3 exposure are extremely limited (Kress
5 et al., 1992; Hogset et al., 1989). Controlled exposures have been for the most part 2 to
6 3 years.
7 A tacit assumption in much of the research on Oj effects on trees is that seedling
8 response to O3 is a good predictor of large tree response. This assumption has been
9 necessitated by the difficulty in exposing large trees for long times to Oj. Pye (1988)
10 reviewed the problems of extrapolation from seedling/sapling experiments to large trees and
11 noted several areas of difference between seedling/saplings and large trees. The problems
12 include not only size but are those equivalent to a single tree growing in a field and a tree
13 growing in a stand. There exists, between seedlings and large trees, a difference the ratio of
14 photosynthetic to respiratory tissue; as a tree ages, the photosynthetically inactive tissue
15 grows. Cambial tissues associated with stem, branch, and coarse roots also increase.
16 Another difference is the existence of microclimatic and morphologic gradients across a
17 canopy and altered water and nutrient regimes. Deep forest canopies generate a
18 microclimatic gradient that is paralleled by a morphological gradient from sun leaves to
19 shade leaves. Associated with this gradient is net radiation, which can vary by an order of
20 magnitude within the canopy. Windspeed and air temperature decrease from above to below
21 the crown, while air humidity and CO2 increase. Ozone concentrations could also vary with
22 forest crown depth, however, measurements have not been made. Cregg et al. (1989) also
23 argue that these differences in scale can affect growth responses seen. Some studies have
24 indicated that seedlings may be more sensitive (i.e., greater visible injury) than large trees
25 (Kozlowski, Kramer, and Pallardy, 1992); however, Samuelson and Edwards (in press) have
26 found that leaves on large red oak trees are more sensitive than those on seedlings. It is
27 likely that a variety of factors determine sensitivity to O3, including stomatal function and
28 presence of detoxification systems, so that in some cases seedlings will be more sensitive and
29 in others large trees will. While each of the four differences between small and large trees
30 mentioned above can be supported on theoretical grounds, little direct information is
31 available to evaluate the importance of these differences, especially with respect of O3.
December 1993 5.39 DRAFT-DO NOT QUOTE OR CITE
-------
1 The microclimate of the canopy of mature trees is quite different from that of seedlings.
2 Radiation intensity through the multilayer canopy can vary by an order of magnitude or more
3 can be expected (Jones, 1992). In addition, gradients of other important microclimatic
4 variables (temperature, humidity, wind speed) exist within the canopy. These will all affect
5 stomatal conductance and some (e.g., wind speed), will affect canopy conductance.
6 The effect of size on transport processes and the subsequent response to O3 is
7 unknown. The simple fact of greater distance over which transport must occur will affect the
8 timing of response of organs distant from the primary site of O3 impact, the foliage. Studies
9 using methods that integrate functions over the whole tree could provide useful information.
10 For example, combinations of porometer measurements on foliage and whole plant water use
11 measured (Schulze et al., 1985) on individuals of different sizes could provide very useful
12 information on the coupling of leaf level processes to whole canopy and whole plant
13 response. Greater evaporative demand in large than small trees as the result of greater leaf
14 area and different microclimate could lead to transient water stress and stomatal closure,
15 because of insufficient water transport capacity.
16 As a tree grows from a seedling to a large tree the ratio between photosynthesis and
17 respiration declines as a greater portion of the plant tissue is non-photosynthetic. It is
18 reasonable to assume that such a change could result in less resource being available for
19 detoxification and repair as the plant grows. How this change affects the ability of a plant to
20 survive O3 (or any other stress) is not known. Recently, Samuelson and Edwards (in press)
21 present data on northern red oak that show O3 decreased photosynthetic capacity more on
22 lower leaves within the canopy of large trees than on leaves near the top of the canopy—
23 a result apparently counter to the model results of Reich et al. (1990). Seedling
24 photosynthesis was not affected by the same O3 exposure. A more interesting result of this
25 work is the reduction in total canopy biomass found in large trees exposed to O3. It is not
26 possible to directly assess the relative importance of reduced photosynthesis versus loss of
27 canopy from these data, but they do show that differences may exist between large trees and
28 seedlings in their response to O3. These differences may be due to changes in carbon
29 budgets, stomatal characteristics, microclimate, and flushing patterns that develop as
30 seedlings become trees. The ability of northern red oak seedlings to produce three flushes
December 1993 5-40 DRAFT-DO NOT QUOTE OR CITE
-------
1 and thus replace injured foliage may be an important defense mechanism in the seedling
2 stage. The generality and physiological basis of these findings need further investigation.
3 In evergreen perennial plants, foliage must be maintained from one year to the next,
4 frequently through periods unfavorable to growth. In evergreen species that retain a few to
5 several years of leaves increased susceptibility to stress (e.g., frost) could further reduce
6 potential canopy photosynthesis in subsequent years (Brown et al., 1987; Davison et al.,
7 1987; DeHayes et al., 1991; Lucas et al., 1988). Fincher (1993) found that O3 decreased
8 frost tolerance in red spruce in both seedlings and trees. The consequences of this change to
9 seedlings and large trees is in need of further work.
10 The effect of O3 on storage of carbohydrates in large compared to small trees is not
11 known. Changes in storage could affect the ability of the plant to withstand other stresses
12 and/or to produce adequate growth during each growing season.
13 Dendrochronology (tree-ring analysis) provides the opportunity to do retrospective
14 studies over the leaf of large trees. Reduction in annual radial growth has been found in the
15 southern Sierra Nevada for Jeffrey pine but not for ponderosa pine (Peterson et al., 1987;
16 1989, 1991; Peterson and Arbaugh, 1988). One difficult with using tree-ring data to
17 estimate O3-related effects is that it is not always possible to separate reductions due to
18 O3 from other effects (e.g., drought).
19 Development of reliable methods for scaling from small to large trees are crucial to the
20 prediction of the long-term effects of O3 on forest function. Measurement of the response of
21 different size trees to O3 could provide useful data on the relative responses of small and
22 large trees. However, problems exist in giving similar exposures to trees of widely different
23 sizes. The most direct method is to fumigate trees over a significant portion of their life
24 span. Time is the primary obstacle to these studies because they would require decades to
25 complete. Whatever methods are used must be based on a good understanding of the
26 physiological changes that occur as trees grow.
27
28 5.3.5.1 Summary
29 In the previous criteria document it was concluded that the "Critical effects, including
30 reduction in photosynthesis and a shift in the assimilation of photosynthate, will lead to
31 reduced biomass, growth, and yield" (U.S. Environmental Protection Agency, 1986).
December 1993 5-41 DRAFT-DO NOT QUOTE OR CITE
-------
1 In addition, changes in carbohydrate allocation patterns and affects on foliage were noted as
2 important. Since that report additional information has been developed, especially on the
3 effects of O3 on photosynthetic capacity. However, at present there is still no clear
4 understanding of the initial biochemical changes resulting within the leaf cells after the entry
5 of O3 and how these changes interact to produce the observed responses. Much of the
6 earlier research used very high (0.25 ppm or greater) O3 concentrations which produced what
7 could be characterized as acute responses. More recent research has used lower
8 concentrations, usually including near ambient O3 level, so that the observed responses may
9 be more relevant to field conditions. One characteristic of these more recent data is that a
10 longer exposure (days to weeks instead of hours) is needed to show a response.
11 As a result of the research since the last criteria document, we have a better
12 understanding of the reduction of photosynthesis as a result of O3 exposure, especially its
13 affects on the central carboxylating enzyme (ribulose-6-P-carboxylase/oxygenase). The rate
14 of senescence of leaves has been shown to increase as a function of increasing O3 exposure.
15 At near-ambient exposures, leaf production has been shown to increase in some species,
16 thereby off-setting the increased loss to due senescence. The mechanism of the increase in
17 senescence is not known at the present time and deserves further study. Finally, the role that
18 changes in allocation of resources plays in plant response to O3 has gained greater
19 prominence. Most studies have shown that allocation of photosynthate to roots is decreased
20 by O3. In some cases, allocation to leaf production has increased. Whether these changes
21 are driven entirely by changes in carbohydrate availability or are controlled by other factors
22 (e.g., hormones) is not known at present.
23 Some potentially significant processes have been investigated since the last Criteria
24 Document, especially detoxification and compensatory processes. The role(s) of
25 detoxification in providing a level of resistance to O3 has been investigated; however, it is
26 still not clear to what degree these processes can provide protection against O3 damage.
27 Data are especially needed on the potential rates of antioxidant production and on the
28 subcellular localization of the antioxidants. Potential rates of antioxidant production are
29 needed to assess whether they are sufficient to detoxify the O3 as it enters the cell. The
30 localization is needed to assess whether the antioxidants are in a location (cell wall or
31 plasmalemma) which permits contact with the O3 before it has a chance to damage
December 1993 5-42 DRAFT-DO NOT QUOTE OR CITE
-------
1 subcellular systems. Ozone exposure has been shown to decrease cold tolerance of foliage in
2 some species. This response could have a major impact on long-lived evergreen species that
3 retain leaves for several years. Various forms of compensation, especially stimulation of
4 production of new leaves and higher photosynthetic capacity of new leaves, have been
5 reported. While these processes divert resources away from other sinks, compensation may
6 counteract the reduction in canopy carbon fixation produced by O3. The quantitative
7 importance of these processes is still in need of investigation. At the canopy level some
8 additional evidence that diurnal stomatal variation is important in O3 uptake. These data will
9 provide important checks on the modelling of stand response to O3.
10 Questions still remaining to be addressed include: How does the plant accumulate the
11 effects of O3 absorption? How important are cumulative O3 absorption and O3 absorption
12 rate? How does the plant integrate injury due to O3 absorption, detoxification,
13 compensation, and other processes into a whole-plant response? What are the roles of
14 growth regulators, especially ethylene? Are they significant? The major problem facing
15 researchers trying to predict long-term O3 effects on plants is how the plant integrates all of
16 the response to O3 into the overall response to the environment, including naturally occurring
17 stresses. Little is now known about how response to O3 changes with increasing age and
18 size. This information is crucial to predicting the long-term consequence of O3 exposure in
19 forested ecosystems.
20
21
22 5.4 FACTORS THAT MODIFY PLANT RESPONSE
23 5.4.1 Modification of Functional and Growth Responses
24 Plant response to oxidants may be modified by various biological, physical, and
25 chemical factors. Biological factors that modify plant response include those within the plant
26 as well as those external to the plant. The genetic make up as well as their development
27 stage play critical roles in the way individual plants respond to O3 and other external
28 stresses. For example, different varieties or cultivars of a particular species are known to
29 differ greatly in their response to a given exposure to O3, while the magnitude of the
30 response of a particular variety, in turn, depends upon environmental factors such as
31 temperature and humidity, soil moisture and nutrition, the presence of pests or pathogens,
December 1993 5.43 DRAFT-DO NOT QUOTE OR CITE
-------
1 and exposure to other pollutants or agricultural spray chemicals. In other words, response
2 will be dictated by the plant's present and past environmental milieu, which also includes the
3 temporal pattern of exposure, and the plant's stage of development. The corollary is also
4 true: exposure to oxidants can modify response to other environmental variables. For
5 example, exposure to O3 reduces the ability of trees to withstand winter injury caused by
6 exposure to freezing temperatures (Davison et al., 1988), and influences the success of pest
7 infestations (Hain, 1987; Lechowiez, 1987). Hence, both the impact of environmental
8 factors on response to oxidants and the effects of oxidants on responses to environmental
9 factors have to be considered in determining the impact of oxidants on vegetation in the field.
10 These interactions are summarized as the involvement of "other stresses" in the scheme
11 shown in Figure 5-6 (Section 5.3). In the following review, the environmental factors are
12 grouped into three categories: biological (including genetic and developmental components),
13 physical, and chemical.
14 Runeckles and Chevone (1992) have recently provided a general review of the
15 interactive effects of environmental factors and O3. The subject is also treated in a National
16 Acid Precipitation Assessment Program report (Shriner et al., 1990).
17
18 5.4.2 Genetics
19 The response of an individual plant within a species and at a given age is affected both
20 by its genetic makeup and the environment in which it grows. This section examines the role
21 of genetics in plant response to O3 and its implication for both managed and natural
22 ecosystems. In addition, major knowledge gaps in the understanding of genetic aspects of
23 O3 responses are pointed out.
24 The responses of plants to O3 are strongly influenced by genetics as was summarized in
25 the air quality criteria document for O3 (U.S. Environmental Protection Agency, 1986).
26 Thus, the plants of a given population or family will not respond to O3 in the same way,
27 even if they are grown in a homogenous environment. This has been amply demonstrated
28 through intraspecific comparisons of O3 sensitivity as determined by foliar sensitivity of
29 ornamental plants, the aesthetic value of which are decreased by visible foliary injury, and of
30 woody plants that are important components of natural ecosystems (Table 5-1). Ornamental
31 plants and plants growing in wilderness areas, for example, have an intrinsic worth, apart
December 1993 5-44 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-1. EXAMPLES OF INTRASPECIFIC VARIATION OF FOLIAR
SYMPTOMS IN OZONE RESPONSE
Species
Ornamental,
Non-Woody
Plants
Petunia sp.
(Pentunia)
Poa pratensis L.
(Kentucky bluegrass)
Genetic
Unit"
Cultivars
Cultivars
Concentration
400 ppb -
4 h/day
400 ppb
300 ppb
Duration
4 days
2h
4h
Range of
Response
20 to 60% (3)
0 to 90%(3)
30 to 60% (3)
Reference
Elkiey and Ormrod
(1979)
Murray et al. (1975);
Wilton et al. (1972)
Trees
Acer rubrum L.
(Red maple)
Fraxinus americana L.
(White ash)
Populations
Half-sib
families
Fraxinus pennsylvanica Half-sib
Marsh. (Green ash) families
Gleditsia triacanthos L. Cultivars
(Honeylocust)
Pinus ponderosa Half-sib
Dougl. ex P. families
Laws and C. Laws
(Ponderosa pine)
Pinus strobus L. Clones
(Eastern white pine)
Pinus taeda L. Half-sib
(Loblolly pine) families
Populus tremuloides Clones
Michx. (Trembling
aspen)
750 ppb
7 h/day
500 ppb
250 ppb
500 ppb
250 ppb
Ambient
1.5 x ambient
300 ppb
250 ppb
ambient +
60 ppb
200 ppb
150 ppb
3 days
7.5h
6h
7.5 h
6h
19 to 34%(2)
0 to 50% (3)
2 to 33% (2)
0 to 40%(3)
2 to 39% (2)
1 growing 0 to 34% (3)
season
3 growing 0 to 28% (2)
seasons
6h
8 h 3 to 29% (2)
1 growing 1 to 42% (1)
season
Townsend and
Dochinger (1974)
Karnosky and Steiner
(1981); Steiner and
Davis (1979)
Karnosky and Steiner
(1981);
Steiner and Davis
(1979)
Karnosky (198 la)
Temple et al. (1992);
0 to 60% (3) Houston (1974)
Kress et al. (1982a);
Adams et al. (1988)
3 h 7 to 56% (1) Karnosky (1977);
6 h 10 to 91 % (1) Berrang et al. (1991)
Cultivars = a variety of agricultural or horticultural crops produced by selective breeding or a vegetatively
propagated tree selection; half-sib seedlings = seedlings with one parent in common; full-sib seedlings =
seedlings in which both parents are in common; clones = vegetatively propagated individual genotypes;
populations = seedlings derived from a common gene pool.
Range of response is expressed as percentage of leaves showing visible symptoms (1); percentage of leaf area
injured (2); or percentage from a leaf injury rating scheme (3).
December 1993
5-45
DRAFT-DO NOT QUOTE OR CITE
-------
1 from any economic value related to growth (Tingey et al., 1990). Considerable genetic
2 variation in O3 sensitivity has also been demonstrated for growth responses of crop plants
3 (Table 5-2). The range of responses displayed for visible foliar injury and growth, biomass
4 or yield vary from species to species and from study to study. However, it is not uncommon
5 to have genotypes varying from no response to well over fifty percent leaf area injured or
6 fifty percent growth or yield reductions in the same study. Additional examples of genetic
7 variation in O3 response are shown in Figure 5-7 for visible foliar injury and in Figure 5-8
8 for growth. From Figure 5-7, we can see that depending on what population has been
9 examined, white ash (Fraxinus americana L.) and green ash (F. pennsylvanica Marsh.) could
10 have been classified as either O3 sensitive or O3 tolerant. Also noticeable from this figure is
11 the large amount of variation in O3 tolerance of individual half-sib (one parent in common)
12 families from a given population. From Figure 5-8, the heterogeneity within a given loblolly
13 pine (Pinus taeda L.) half-sib family in terms of growth is displayed. This variability has
14 some interesting implications. First, since plants of a given species vary widely in their
15 response to O3 exposure, response relationships generated for a single genotype or small
16 group of genotypes may not adequately represent the responses of the species as a whole
17 (Temple, 1990). Second, because of the genetic variability and differential fitness existing
18 among different genotypes in a population of plants, O3 imposes a selective force favoring
19 tolerant genotypes over sensitive ones (Roose et al., 1982; Treshow, 1980). Each of these
20 implications will be discussed in this section.
21
22 Mechanisms and Gene Numbers
23 Little is known about the genetic bases for O3 resistance mechanisms or about the
24 numbers of genes involved in these mechanisms (Pitelka, 1988). Most O3 resistance
25 mechanisms involve a physiological cost which will result in decreased growth and
26 productivity of resistant plants grown under O3 stress. Partial or complete stomatal closure
27 in the presence of O3 is an example of a mechanism of resistance that has been demonstrated
28 for several plants (Engle and Gabelman, 1966; Thorne and Hanson, 1976; Reich, 1987;
29 Sumizono and Inoue, 1986; Tingey and Taylor, 1982) and that involves a high physiological
30 cost as plants that have reduced stomatal conductivity will also have reduced carbon
31 assimilation for growth (Ehleringer, 1991). Tolerance of internal leaf tissues to O3 may
December 1993 5-46 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-2. EXAMPLES OF INTRASPECIFIC VARIATION IN GROWTH
RESPONSES FOLLOWING OZONE EXPOSURES
Species
Crops and
Non-Woody
Plants
Agrostis capillaris L.
(Bentgrass)
Begonia semperflorens
Hort. (Bedding
begonia)
Festuca arundinacea
Schreb. (Fescue)
Lycopersicon
esculentwn L.
(Tomato)
Phaseolus vulgaris L.
(Snapbean)
Plantago major L.
(Common plantago)
Raphanns sativus L.
(Radish)
Silene cucahalus
(Bladder campion)
Solatium tuber o sum L.
(Potato)
Spinacia oleracea L.
(Spinach)
Trees and Other
Woody Plants
Acer rubrum L.
(Red maple)
Abies alba Mill.
(Silver fir)
Pinus elliottii
Engelm. (Slash pine)
Genetic
Unit"
Populations
Cultivars
Cultivars
Cultivars
Cultivars
Populations
Within
Cultivar
Populations
Cultivars
Cultivars
Populations
Populations
Half-sib
families
Concentration
60 ppb
500 ppb-
4 h/day
250 ppb-
4 h/day
400 ppb-
6 h/day
400 ppb
1.5 x ambient
60 ppb-
7h
72 ppb-
7 h
80 ppb-
7 h/day
70 nl/
1-7 h/day
.1 n\l
1-4 h/day -
3 days/week
35 ppb-
12 h/day
150ppb-
6 h/day
130 ppb -
7 h/day
750 ppb -
7 h/day
250 ppb -
7 h/day
3 x ambient
Duration
4 weeks
2 days
4 days
7 days
2h
1 growing
season
mean-
44 days
mean-
54 days
42 days
2 weeks
3 weeks
4 weeks
8 days
38 days
3 days
10 days
3 growing
seasons
Range of
Response
-45 to
+20%(2)
-59 to 0%(2)
-16 to
+ 10% (2)
-53 to -35% (2)
-50 to -4% (2)
-54 to -17%(3)
-26 to -2% (3)
-73 to -44% (3)
-68 to -50% (3)
-24 to 0%(1)
-40 to -5%(2)
-75 to -48% (2)
-10 to 0%(2)
-40 to 0%(2)
-56 to -28% (2)
-36 to -17%(1)
-18 to +3%(1)
-20 to 0%(1)
Reference
Dueck et al.
(1987,)
Reinert and
Nelson (1979);
Reinert and
Nelson (1980)
Flagler and
Younger (1982)
Reinert and
Henderson (1980);
Temple (1990)
Heck et al. (1988);
Temple (1991);
Eason and Reinert
(1991)
Reiling and
Davison (1992a)
Gillespie and
Winner (1989)
Ernst et al. (1985)
Pell and Pearson
(1984);
Ormrod et al.
(1971)
Heagle et al.
(1979)
Townsend and
Dochinger (1974)
Larsen et al.
(1990)
Dean and Johnson
(1992)
December 1993
5-47 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-2 (cont'd). EXAMPLES OF EVTRASPECIFIC VARIATION IN GROWTH
RESPONSES FOLLOWING OZONE EXPOSURES
Species
Pinus taeda L.
(Loblolly pine)
Pinus taeda L.
(Loblolly pine)
Populus tremuloides
Michx. (Trembling
aspen)
Rhododendron
obtusum
(Lindl) Planch.
(Azalea)
Genetic
Unit"
Full-sib
families
Half-sib
families
Clones
Ambient
Cultivars
Concentration
50 ppb-
6 h/day
1.9 x ambient
Ambient -t-
60 ppb
2.5 x ambient
250 ppb
26.4 ppm-h
3 growing seasons
250 ppb-
3 h/day
Duration
28 days
2 growing
seasons
1 growing
season
1 growing
season
8h
92 days
-24 to -
12%(2)
6 days
Range of
Response
-18toO%(l)
-19 to 0%(2)
-27.5 to
+3%(2)
-19 to -2%(2)
-22 to
+30%(2)
-74 to -5% (2)
-43% to 0(2)
Reference
Kress et al.
(1982b);
Shafer and Heagle
(1989)
Adams et al.
(1988);
Qui et al. (1992);
Winner et al.
(1987)
Karnosky et al.
(1992);
Wang et al. (1986)
Sanders and
Reinert (1982)
aCultivars = a variety of agricultural or horticultural crops produced by selective breeding or a vegetatively
propagated tree selection; half-sib seedlings = seedlings with one parent in common; full-sib seedlings =
seedlings in which both parents are in common; clones = vegetatively propagated individual genotypes;
populations = seedlings derived from a common gene pool.
Range of response is expressed as decrease compared to charcoal-filtered-air control plants in terms of growth
(1), biomass (2) or yield (3).
1 involve the production of antioxidant defense compounds (Lee and Bennett, 1982; Gupta
2 et al., 1991) or other types of biochemical defense systems. The extent to which these
3 internal tolerance mechanisms have physiological costs associated with them is not yet
4 understood, but it is likely that increased defense compound production, triggered by O3, will
5 impact the amount of carbon available for growth (Ehleringer, 1991). The genetic regulation
6 of these or other O3 resistance mechanisms has not yet been thoroughly characterized.
7 Whether or not O3 resistance is due to single gene or multi-gene control will affect the
8 rate and the extent of resistance development (Roose, 1991). Rapid stomatal closing in the
9 presence of O3 appears to be under the control of either a single gene or a few genes in
10 onion (Allium cepa L.) (Engle and Gabelman, 1966), some bean (Phaseolus vulgaris L.)
11 cultivars (cultivated varieties) (Knudson-Butler and Tibbitts, 1979), soybean (Glycine max L.)
12 (Damicone and Manning, 1987) and petunia (Petunia spp.) (Elkiey and Ormrod, 1979).
December 1993
5-48
DRAFT-DO NOT QUOTE OR CITE
-------
250-
A. F. americana
4567
Provenance
9 10
250
200
150
.5, 100-
501
0
B. F. pennslvanica
111
III
12345
6 7 89 10 11 12 13 14 15 16
Provenance
Figure 5-7. The average injury index for visible foliar injury after exposure of
1-year-old seedlings to 50 pphm O3 for 7.5 h. Each mean shown represents
the average of five trees per family. There were either four or five half-sib
families for each white ash (Fraxinus americana L.) provenance (geographic
location) and either three or four families for each green ash
(F. pennsylvanica Marsh.) provenance. The specifics of the experimental
design are reported in Karnosky and Steiner (1981).
December 1993
5-49
DRAFT-DO NOT QUOTE OR CITE
-------
113 137 163 188 213 237 263 288 313 337 363 388 413 437 463 488 513
Midpoint of Biomass Class (g)
Figure 5-8. Frequency distribution showing the variability hi O3 response (midpoint of
whole-plant biomass) within one half-sib family of loblolly pine (P. taeda L.)
exposed to increasing levels of O3 under chronic-level field conditions over
several growing seasons (Adams et al., 1988). The arrows show the mean
response for each of the three O3 treatments (sub-ambient, ambient and
above-ambient O3). The specifics of the experimental design are reported
by Adams et al. (1988). This figure was developed by Taylor (1993).
1 Generally, resistance mechanisms appear to be more complex (Karnosky, 1989a) and seem to
2 involve multiple gene control as has been demonstrated in tobacco (Nicotiana tabacum L.)
3 (Aycock, 1972; Huang et al., 1975, Povilaitis, 1967), some bean cultivars (Mebrahtu et al.,
4 1990a,b,c), corn (Zea maize L.) (Cameron, 1975), tall fescue (Festuca arundinacea Schreb.)
5 (Johnston et al., 1983), potato (Solatium tuberosum L.) (DeVos et al., 1982; Dragoescu
6 et al., 1987) and loblolly pine (Pinus taeda L.) (Weir, 1977; Taylor, 1993).
7
8 Genetic Implications of Ozone Effects: Managed Ecosystems
9 Because of the high cost involved in conducting long-term growth studies to determine
10 O3 effects on plants, only a small proportion of the total number of commercial crop
11 cultivars and of the important tree seed sources, families, clones and cultivars have been
December 1993 5-50 DRAFT-DO NOT QUOTE OR CITE
-------
1 adequately examined for O3 sensitivity. Still, a tremendous amount of variation has been
2 found, as was described previously in the first O3 criteria document (U.S. Environmental
3 Protection Agency, 1986) and in Tables 5-1 and 5-2 of this section.
4 Plant breeders and nurserymen working in locations with high O3 concentrations have
5 inadvertently developed selections more tolerant to O3 than those developed in locations with
6 low O3 exposures (Reinert et al., 1982, Roose et al., 1982). "Team" alfalfa (Medicago
1 sativd) and "Kennebec", "Pungo" and "Katahdin" potato were developed at the USDA
8 Research Center at Beltsville, Maryland where 0.120 ppm O3 is commonly exceeded (Lefohn
9 and Pinkerton, 1988; Ludwig and Shelar, 1980). These cultivars have proven to be more
10 O3 tolerant than cultivars developed elsewhere (Reinert et al., 1982). Similarly, cotton
11 (Gossypium spp.) and sugar beet (Phaseolus spp.) cultivars developed in southern California,
12 where O3 levels are among the highest in the county, are more O3 tolerant than cultivars
13 developed in low O3 areas (Reinert et al., 1982).
14 Nurserymen, Christmas tree growers and seed orchard managers have all routinely
15 discarded pollution-sensitive chlorotic dwarf and tipburned white pine trees because of their
16 slow growth in areas with high O3 exposures (Umbach and Davis, 1984). Thus, they have
17 contributed to the selection of more O3-tolerant commercial forests.
18 While these examples suggest that selection of O3-tolerant genotypes is possible, this
19 topic remains a highly debatable issue, and the general consensus of the scientific community
20 is that top priority should be given to solving pollution problems at their source (Karnosky
21 et al., 1989b) and not in selecting pollution-tolerant cultivars.
22 An interesting set of experiments by Barnes et al. (1990) and Velissariou et al. (1992)
23 have described a concern about the modern crop varieties that have been developed in clean-
24 air environments, but that are being routinely grown in areas with elevated O3 exposures.
25 These authors speculated that breeders of spring wheat (Triticum aestivwn L.) grown in
26 Greece had inadvertently selected varieties with increased O3 sensitivity due to their higher
27 rates of stomatal conductivity (Velissariou et al., 1992). Vellisariou et al. (1992) found a
28 significant correlation between year of introduction and stomatal conductance with stomatal
29 conductance increasing with the more modern introductions. The authors suggested that the
30 selection for higher yields had resulted in a higher O3 uptake for the modern spring wheat
31 cultivars, contributing to their increased O3 sensitivity. When they compared the relative
December 1993 5.5! DRAFT-DO NOT QUOTE OR CITE
-------
1 growth rates of spring wheat cultivars released over the period from 1932 to 1980, the
2 modem cultivars had more foliar injury and more growth decrease when grown in the
3 presence of O3 (Barnes et al., 1990; Velissariou et al., 1992).
4
5 Genetic Implications of Ozone Effects: Natural Ecosystems and Biodiversity
6 Air pollutants can affect the genetics of plant populations in two ways: They may
7 increase mutation rates or they may apply selection pressures which may eventually lead to
8 adaptive responses (Cook and Wood, 1976). The issue of O3-induced changes in mutation
9 rate has not been adequately studied yet, but recent evidence by Floyd (1992) suggests that
10 DNA may be affected by O3 to induce mutation in plants. However, there is evidence that
11 O3 may be affecting plant populations via natural selection. According to Bradshaw and
12 McNeilly (1991), there are three stages of selection-driven population change: Stage I.
13 Elimination of the most sensitive genotypes; Stage n: Elimination of all genotypes except
14 the most resistant; Stage HI: Interbreeding of the survivors.
15 The first report of O3 as a selective force in plant populations was that involving lupine
16 (Lupinus bicolof) populations in the greater Los Angeles area (Dunn, 1959). Local
17 Los Angeles area populations were more O3 resistant than populations originating from
18 cleaner-air areas. Berrang et al. (1986, 1989, 1991) have presented evidence for population
19 change in trembling aspen (Populus tremuloides L.). Aspen clones from across the United
20 States were sampled randomly from populations in polluted and non-polluted areas. Aspen
21 from areas with high ambient O3 concentrations were visibly injured to a lesser extent by
22 experimental O3 exposures than clones from areas with low O3 concentrations (Berrang
23 et al., 1986, 1991). Similar results were seen for field trials of O3 injury (Berrang et al.,
24 1989). More recently, growth rate and biomass differences have been reported for aspen
25 clones differing in O3 tolerance (Karnosky et al., 1992). Berrang et al. (1989) suggest that
26 sensitive genotypes are not killed directly by O3, but are eliminated through intraspecific
27 competition for light, nutrients and water with their resistant neighbors. Spatial (population)
28 variation in O3 resistance that is related to background O3 pollution has also been
29 demonstrated in British populations of plantago (Plantago major L.) (Reiling and Davison,
30 1992a,b).
December 1993 5-52 DRAFT-DO NOT QUOTE OR CITE
-------
1 There have been three concerns raised regarding the spatial variation studies of
2 O3 resistance. First, since O3 does not generally show steep concentration gradients, spatial
3 studies must involve populations that are great distances from one another so that it is
4 difficult to determine whether geographical differences in O3 resistance are primarily related
5 to local O3 exposures or to other environmental factors (Reiling and Davison, 1992a).
6 Second, spatial studies are limited by the general absence of historical records of ambient
7 O3 concentrations at the sites where the populations were sampled (Bell et al., 1991). Third,
8 no O3 study has collected plants from the same population over time to demonstrate
9 O3-induced population change over time (Bell et al., 1991) as has been demonstrated for
10 other pollutants. However, Karnosky (1981b; 1989b) studied the O3 symptom expression
11 and survival of over 1,500 eastern white pine trees growing in southern Wisconsin and found
12 that O3-sensitive genotypes had a ten-times-higher rate of mortality than did the O3-resistant
13 genotypes over a 15-year study (Table 5-3). This is direct evidence of Stage I of natural
14 selection occurring. Further evidence of this type was presented by Heagle et al. (1991) who
15 found a population change in O3 sensitivity over two years with white clover (Trifolium
16 repens L.) exposed to O3 in open-top chambers. A high O3 dose at the end of the study
17 caused significantly less foliar injury in plants which survived two seasons of exposure to
18 high O3 concentrations than on plants that had survived low O3 concentrations.
19
20
TABLE 5-3. MORTALITY OF THREE OZONE SENSITIVITY CLASSES OF
EASTERN WHITE PINE (PINUS STROBUS L.) TREES DURING THE PERIOD
FROM 1971 TO 1986
Sensitivty Class11
Resistant
Intermediate
Sensitive
Number Trees
1386
98
57
Number Trees Dead
34
3
14
Percent Mortality
2.4%
3.1%
24.6%
^Resistant = Not showing visible foliar injury during the study; Intermediate = Showing visible injury,
including foliar tipburn during one or two years; Sensitive = Showing visible injury, including foliar tipburn,
short needles and poor needle retention for three or more years of the study.
Source: From Karnosky, (1989b).
December 1993 5.53 DRAFT-DO NOT QUOTE OR CITE
-------
1 The rate of evolution is dependent on: (1) the selection pressure; (2) the magnitude of
2 the genetically controlled variability; and, (3) the number of genes involved (Roose, 1991).
3 Long-lived species, such as trees, will evolve more slowly than annuals or biennials (Barrett
4 and Bush, 1991). Gillespie and Winner (1989) found O3 to be a strong and rapid selective
5 force with radish. Ozone resistance was expressed within one generation following a series
6 of artificial pollinations with various populations from the cultivar "Cherry Belle".
7 Whether or not the loss of some genotypes from plant populations is important is a
8 debatable question. However, it is likely that sensitive genotypes are being lost from natural
9 ecosystems with current O3 exposures. Field studies documenting differential growth rates
10 of O3-sensitive and tolerant genotypes of eastern white pine in natural ecosystems influenced
11 by O3 were summarized in the original air quality criteria document for O3 (U.S.
12 Environmental Protection Agency, 1986). Similar findings have subsequently been reported
13 for O3-sensitive and tolerant Jeffry pine (Pinus Jeffrey, Grev. and Balf.) trees in California
14 (Peterson et al., 1987). It is likely that these growth rate differences affect the competitive
15 ability of O3-sensitive genotypes and increase their mortality rate (Karnosky, 1989b).
16 While some loss of rare alleles (one of a series of genes that are alternative in
17 inheritance) and change in gene frequency is likely with loss of sensitive genotypes, the
18 significance of these effects on biodiversity is unknown (Barrett and Bush, 1991). If the
19 remaining population of O3-resistant plants is less adaptable to subsequent change due to a
20 reduced redundancy, as has been predicted by Gregorius (1989), or if O3 sensitivity is linked
21 to other traits such as rapid growth or high productivity, as has been suggested because of
22 the inherently higher gas exchange rates of some O3-sensitive genotypes (Barnes et al., 1990;
23 Thorne and Hanson, 1976; Turner et al., 1972; Velissarious et al., 1992), then losing these
24 sensitive genotypes is both biologically and economically important. This remains a point of
25 scientific debate. While the evolution of resistance to air pollution is hypothesized to
26 contribute to the loss of genetic variability (Scholz et al., 1989; Karnosky, 1991), other
27 scientists suggest that there is little experimental evidence for concluding that genetic
28 diversity is actually threatened by air pollution and that air pollution has less important
29 implications for plant populations than do factors such as global climate change and habitat
30 fragmentation (Parson and Pitelka, 1991; Taylor and Pitelka, 1991). Clearly, there is a need
31 for additional research in this area of O3 effects in plant biodiversity (Karnosky et al., 1989).
December 1993 5-54 DRAFT-DO NOT QUOTE OR CITE
-------
1 Reproductive Aspects and Related Genetic Implications
2 In the previous discussion in this section, only natural selection at the whole-plant level
3 has been mentioned. This type of selection occurs as plants compete with their neighbors for
4 survival and the ability to reproduce. Selection is thought to also occur during the
5 reproductive process (Feder and Sullivan, 1969; Krause et al., 1975) and this is referred to
6 as gametophytic selection (Mulcahey, 1979; Wolters and Martens, 1987) or fertility selection
7 (Venne et al., 1989). The ability of gametophyte (haploid part of the plant-life cycle)
8 selection to modify the sporophytic generation depends on two critical issues: pollen genes
9 should be expressed after meiosis (cell divisions leading to production of gametes) and those
10 same genes should also be expressed in the sporophytes (diploid part of the plant-life cycle)
11 (Mulcahey and Mulcahey, 1983). This genetic overlap has been demonstrated in some
12 species (Mulcahey, 1979; Searcy and Mulcahey, 1985; Walsh and Charlesworth, 1992).
13 Indirect evidence for O3-induced gametic selection was presented for Scot's pine (Pinus
14 sylvestris L.) by Venne et al. (1989). Based on their studies of the effects of O3 on the
15 pollen germination and tube elongation of some thirty Scots pine clones, they found that
16 O3 could markedly change the relative male contribution to successful fertilization.
17 However, this study did not actually examine offspring as would be needed to positively
18 prove O3-induced gametophytic selection.
19 Studies of O3 effects on pollen germination and tube elongation have generally found a
20 negative impact of O3 on this critical element of reproduction (Table 5-4). Whether or not
21 selection is occurring at the pollen level because of a selective disadvantage of the pollen
22 from sensitive genotypes is a debatable issue. Feder (1986) and Krause et al. (1975) found
23 that the pollen from O3-sensitive genotypes of petunia and tomato (Lycopersicon esculentum
24 L.) was more severely affected by O3 than pollen from tolerant genotypes, suggesting that
25 gametophytic selection could be occurring. Similar results were found for Scots pine clones
26 by Venne et al. (1989). These authors found that the relative male contribution for charcoal-
27 filtered air versus O3-treated conditions was very different and could potentially lead to a
28 strong gametophytic selection response caused by O3. However, Hanson and Addis (1975)
29 did not see any differences in the effect of O3 on the pollen from sensitive and tolerant
30 petunia genotypes, and Benoit et al. (1983) found no apparent differences in the susceptibility
December 1993 5.55 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-4. EXAMPLES OF OZONE EFFECTS ON POLLEN GERMINATION
AND TUBE ELONGATION
Species
Nicotiana tobacum L.
(Tobacco)
Petunia hybrida
(Petunia)
Pinus strobus L.
(Eastern white pine)
Zea mays L.
iCorn)
Pollen
Germination
Decrease
Not tested
No effect
Decrease
Pollen Tube
Elongation
Decrease
Decrease
Decrease
Not tested
Reference
Feder, 1968;
Feder and Shrier, 1990
Feder and Shrier, 1990
Benoitetal. 1983
Mumfordetal. 1972
1 of eastern white pine pollen from O3-sensitive or tolerant genotypes. Clearly, the question of
2 whether O3-induced gametophytic selection is occurring has not yet been resolved.
3 Reduced flowering as the result of prolonged fumigation with O3 has been shown in
4 Bladder campion (Silene cucubalus) (Ernst et al., 1985). Decreased floral initiation and
5 decreased floral productivity under long-term O3 exposures have also been reported in
6 geranium (Pelargonium spp.) and carnation (Dianthus caryophyllus) (Feder, 1970).
7 Ozone-induced impairment of flowering will reduce the fitness of the affected genotypes,
8 populations or species and may result in the eventual loss of these genetic units from the
9 O3-stressed ecosystem. Reduced eastern white pine fecundity in air pollution-stressed
10 ecosystems has been reported by Houston and Dochinger (1977).
11
12 Genetic Summary
13 Plant species, cultivars, populations and individuals within populations display variable
14 responses to O3. Variability in O3 responses between and within species was described in
15 the previous O3 criteria document (U.S. Environmental Protection Agency, 1986).
16 An important component of this variation is genetically controlled. The specific genes
17 controlling O3 response and involved in mechanisms of O3 tolerance are as yet largely
18 unknown. However, control of stomatal conductance and internal biochemical defense
19 systems are among the most commonly described tolerance mechanisms. Ozone tolerance is
20 generally thought to be controlled by multiple genes.
December 1993
5-56
DRAFT-DO NOT QUOTE OR CITE
-------
1 There are implications of genetic variation in O3 response, both for managed and
2 natural ecosystems. These are summarized below along with the relative degree of
3 uncertainty attached to each.
4 It is known, with a great deal of certainty, that plants have a high degree of genetic
5 variation in O3 response. Thus, exposure-response equations and yield-loss equations
6 developed for a single or small number of cultivars, genotypes, families or populations may
7 not adequately represent the response of thespecies as a whole. As a corollary to this,
8 sensitive responder genotypes will not be protected by air-quality standards based on mean
9 responses.
10 The issue of O3 effects on biodiversity via natural selection is a topic of debate within
11 the scientific community. The potential for natural selection for O3 tolerance and associated
12 loss of sensitive genotypes is regional in nature, unlike well known, point-source pollution
13 impacts which occur on local plant populations. However, the intensity of O3 selection is
14 generally thought to be quite low, 0.3 or less (Taylor and Pitelka, 1992), in the majority of
15 the United States. The extent that germplasm has been or is continuing to be affected in
16 terms of allele loss or gene frequency changes by O3 and how this might be impacting the
17 genetic adaptability of populations is an open and important research question.
18 While it is well known that individual plants within a species vary in their O3 tolerance,
19 the physiological costs to tolerant plants are not known in terms of carbon assimilation and
20 allocation. Tolerance mechanisms based on reduced stomatal conductivity in the presence of
21 O3 would likely reduce growth of tolerant plants. Similarly, tolerance mechanisms based on
22 the productivity of antioxidant compounds will likely shunt plant resources away from growth
23 to the production of the defense compounds. The characterization of the extent and types of
24 physiological costs involved in O3 tolerance remains an important research question.
25
26 5.4.3 Environmental Biological Factors
27 The previous criteria document (U.S. Environmental Protection Agency, 1986)
28 discussed pollutant-plant-pest and pollutant-plant-pathogen interactions together, and provided
29 a tabular summary of pathogen effects. However, in light of the numerous studies of insect
30 and pathogen interactions that have appeared in recent years, the topics are dealt with
December 1993 5,57 DRAFT-DO NOT QUOTE OR CITE
-------
1 separately below. Nevertheless, it is worth reiterating several points made in the previous
2 criteria document:
3 • pests and diseases are natural components of managed and natural
4 ecosystems;
5
6 • significant crop and timber losses result from pests and pathogens;
7
8 • the establishment of disease and pest infestations and their subsequent
9 development involve complex interactions between host plant, the
10 environment and the causal organism;
11
12 • the generalized disease (or pest infestation) cycle involves the arrival of the
13 pathogen or pest on the host plant surface or its introduction into the host
14 plant tissues through wounds or as a result of insect feeding activity;
15
16 • growth and development or propagation of the pathogen or pest only occurs
17 if all environmental conditions are favorable;
18
19 • such development leads to various degrees of host tissue destruction or
20 malfunction, and usually culminates in the causal organism entering a
21 reproductive stage and producing propagules (e.g., spores or eggs) that
22 facilitate its spread.
23
24 Ozone may modify any stage of the disease cycle directly, by affecting the causal
25 organism itself, or indirectly by effects on the host plant (Lechowiez, 1987). Conversely,
26 the plant-pest interaction may modify the sensitivity of the host plant to O3.
27 The roots of many members of the pea family (including many important crops such as
28 soybeans, beans and peas) are infected by symbiotic nitrogen-fixing bacteria (Rhizobium spp.)
29 leading to the formation of bacteria-rich nodules that contribute to the nitrogen economy of
30 the plant, through their ability to fix and convert atmospheric nitrogen, N2, to biologically
31 useful forms. Other nitrogen-fixing microorganisms are associated with the roots of several
32 species, and in many cases roots are invaded by species of soil fungi to form mycorrhizal
33 symbioses that assist in root functioning. These symbioses constitute micro-ecosystems and
34 are discussed more fully in Section 5.7 as they relate to forest tree species.
35 Biological interactions also affect the growth of plants in populations (pure stands) and
36 communities (mixtures of species) through the individual plants' competition for available
37 resources (light, CO2, water, nutrients). Such plant-plant interactions are features of all
38 managed and natural ecosystems, but they operate at the indivdual plant level. Hence the
December 1993 5.53 DRAFT-DO NOT QUOTE OR CITE
-------
1 effects of oxidants on these interactions are discussed in both the present section and in
2 Section 5.7 which deals will ecosystem responses.
3
4 5.4.3.1 Oxidant-Plant-Insect Interactions
5 The previous criteria document (U.S. Environmental Protection Agency, 1986),
6 concluded that little was known at that time about O3-insect interactions. Since then the
7 topic has been covered in several reviews: Fluckiger et al. (1988), Hughes (1988), Manning
8 and Keane (1988), and Hain (1987). Relevant studies of the effects of O3 on the feeding
9 preference of herbivorous insects, and on their growth, fecundity, and survival are presented
10 in Table 5-5. As can be readily seen in this summary, the information is widely scattered
11 among a wide range of host plants and pests. Nevertheless, there appears to be a general
12 trend in the observations suggesting that O3-induced changes in the host plants frequently
13 result in increased feeding preference of a range of insect species, although this may or may
14 not be reflected in effects on the growth of the insect.
15 However, in most studies, the effects have been far from clear-cut. For example,
16 variable responses were observed with the aphid, Aphis fabae, on broad bean (Brown et al.,
17 1992), with the aphids, Acyrthosiphon pisum and Aphis rumicis, on pea and dock,
18 respectively, and the beetle, Gastrophysa viridula on dock (Whittaker et al., 1989), with the
19 Mexican bean beetle, Epilachna varivestis, on Corsoy soybean (Endress and Post, 1984), and
20 with the gypsy moth, Lymantria dispar, on white oak (Jeffords and Endress, 1984).
21 Although statistically significant effects were frequently observed, they did not provide any
22 consistent pattern of insect growth response to different levels or patterns of exposure.
23 Brown et al. (1992) observed that the response of Aphis fabae depended upon the
24 dynamics of exposure: growth was stimulated in short-term (< 24 h) continuous exposures
25 or in episodic exposures over several days, whereas longer continuous exposures caused
26 decreased growth. Chappelka et al. (1988c) found that O3 consistently enhanced the feeding
27 preference and larval growth of the Mexican bean beetle on soybean, leading to increased
28 defoliation. Although the cultivar Forrest was significantly more sensitive to O3 than Essex,
29 this difference did not lead to any differences in insect behavior and development. Similarly,
30 clear stimulatory responses were observed with pinworm, Keiferia lycopersicella, on tomato
31 (Trumble et al., 1987), with an aphid, Phyllaphis fagi, and a weevil, Rhynchaenus fagi, on
December 1993 5.59 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-5. OZONE EFFECTS ON INSECT PESTS
I
cr
Host Plant Inse< i
CROP SPECIES
Broad bean aphid
Pea/aphid
Kidney bean/ aphid
Soybean/beetle
(cv Corsoy)
Exposure*1
3 day, 0.085 ppm
< 24 h, 0. 1 ppm
> 24 h, 0. 1 ppm
8 h/day, episodic
4-8 day, var.
14 day, var.
16 day, var.
21 day, 7 h/day, var.
Experimental Conditions
chamber, whole plant
chamber, whole plant
chamber, whole plant
chamber, whole plant
chamber, whole plant
OTC
chamber
OTC
Effect of Ozone on Insect
3-13% decreased growth rate
17% increased growth rate
12% decreased growth rate 15%
increased growth rate
variable effects on growth
15-50% reduction in growth of
insect
variable feeding preference
0.11>0.0>0.05> 0.03 ppm
Reference
Dohmen (1988)
Brown et al. (1992)
Whittaker et al. (1989)
Braun and Fluckiger (1989)
Endress and Post (1985)
Chappelka et al. (1988c)
(cvs Essex. Forrest)
Tomato/pinworm
NATURAL
VEGETATION
Milkweed/monarch
butterfly
Dock/aphid
Dock/beetle
TREES SPECIES
European beech/aphid
2-4 day, 3 h/day,
0.28 ppm
chamber, detached leaf
chamber, whole plant
17-19 day, 7 h/day
0.150-0.178 ppm
15 day, var.
15 day, var.
2 mo., var.
chamber, whole plant
chamber, whole plant
chamber, whole plant
OTC
feeding preference increased and
greater larval growth
80% in crease in larval
development; no effect on
fecundity
no feeding preference but greater
larval growth rate
10% increased growth rate
10% larger egg batches; 4-fold
greater larval survival
75 % increase in number
Trumble et al. (1987)
Bolsinger et al. (1992)
Whittaker et al. (1989)
Whittaker et al. (1989)
Braun and Fluckiger (1989)
-------
TABLE 5-5 (cont'd). OZONE EFFECTS ON INSECT PESTS
3 Host Plant/Insect
cr
Exposure
Experimental Conditions
Effect of Ozone on Insect
Reference
n>
TREE SPECIES
(cont'd)
European beech/weevil 72 h, var.
Cottonwood/beetle
5 h, 0.2 ppm
Ponderosa pine/bark natural
beetle
OTC
OTC
none, field
White oak/gypsy moth 11 day, 7 h/day. var. chamber, leaf disks
2-fold increase in feeding
preference
22-60 % greater consumption of
foliage, but decreased fecundity
increased infestation but in
decreased survival
variable feeding preference
0.15 > 0.03 > 0.09 ppm
Hiltbrunner and Fluckiger (1992)
Jones and Coleman (1988)
Coleman and Jones (1988)
Hain (1987)
Jeffords and Endress (1984)
"var." indicates a range of exposures.
b
"chamber" indicates closed chamber; OTC indicates open-top field chamber.
1
8
O
-------
1 European beech (Braun and Fluckiger, 1989; Hiltbrunner and Fluckiger, 1992), with the
2 monarch butterfly, Danaus plexippus, on milkweed (Bolsinger et al., 1992), and with
3 infestation by the willow leaf beetle, Plagiodera versicolora, on cottonwood (Coleman and
4 Jones, 1988). However, there was less egg-laying by Plagiodera on O^-treated foliage, and
5 treatment had no effect on beetle growth rates and survival (Jones and Coleman, 1988).
6 In view of previous experiments in which it was clearly demonstrated that aphid growth
7 was significantly stimulated by ambient pollutant mixtures containing 63, SO2, and NO2, and
8 in light of other reports of O3-induced stimulations of insect growth, the inhibitory effects of
9 O3 on the growth of Aphis fabae on Broad Bean (dohmen, 1988), Or Kidney bean (Braun
10 and Fluckiger, 1989), may be anomalous. The inhibitory effects on broad bean were only
11 observed at low 03 levels: exposure to higher concentrations resulted in a stimulation of
12 aphid growth, which Dohmen (1988) attributed to the increased rate of leaf senescence of the
13 host plant. The effects observed on kidney bean could not be accounted for by differences in
14 the amino acid composition of the plant sap, although differences in other constituents or
15 direct effects of O3 on the pea aphid itself could not be ruled out (Braun and Fluckiger,
16 1988).
17 A well-established indirect stimulatory effect is the predisposition to bark beetle attack
18 of ponderosa pine injured by exposure to O3. However, the infested trees do not favor good
19 brood production; O3 injury results in a more susceptible but less suitable host (Hain, 1987).
20 In all of these studies, the focus was on direct or indirect effects of 03 on the insect.
21 With the exception of the work of Braun and Fluckiger (1988), any effects on the host plant
22 that were reported were confined to observations on visible symptoms of foliar injury. The
23 only report of an O3-insect interaction affecting the response of the host plant appears to be
24 that of Rosen and Runeckles (1976). This study showed that exposure to sub-acute levels of
25 O3 and infestation with the greenhouse whitefly, Trialeurodes vaporarionm, acted
26 synergistically (i.e., more than additively) in causing leaf injury and accelerated senescence
27 of kidney bean. However, the extent to which other insects with sucking mouthparts, such
28 as aphids. might be involved in similar interactive responses is unknown, as is the nature of
29 any interactions which involve pests that ultimately invade and develop within the host plant,
30 such as those that cause the formation of galls.
December 1993 5-62 DRAFT-DO NOT QUOTE OR CITE
-------
1 The reports of O3-insect-plant interactions are thus scattered among a wide range of
2 host plant and insect species, and only represent a minute fraction of the plant-insect
3 interactions that involve crop and native species. Although there appears to be a trend in the
4 limited data available that suggests that exposures to moderate O3 levels may increase the
5 likelihood of insect attack and its consequences, there is insufficient information to decide
6 whether extrapolation of this generalization is warranted or not. Even if the generalization is
7 valid, it is not possible to generate any quantitative measure of response. Before such
8 estimates will be possible on a broad scale, studies of a much wider range of plant insect-
9 systems will be needed, together with systematic, in-depth studies of individual systems,
10 aimed at determining the long-term effects on both the host plant and the insect. Such
11 studies should include investigations of biological control systems employing beneficial
12 insects which are increasing in use as alternatives to chemical insecticides and herbicides.
13
14 5.4.3.2 Oxidant-Plant-Pathogen Interactions
15 Plant disease is the result of infection by fungi, bacteria, mycoplasmas, viruses and
16 nematodes. Recent reviews of pathogen-plant-O3 interactions have been published by
17 Dowding (1988) and Manning and Keane (1988), and extend the coverage of the previous
18 criteria document (U.S. Environmental Protection Agency, 1986) in which the results of
19 published studies of the effects of O3 on disease development were summarized in tabular
20 form. Interactions involving fungal pathogens occupied most of that review, and more recent
21 studies have maintained this emphasis.
22 The previous criteria document concluded that it was "impossible to generalize and
23 predict effects in particular situations." (U.S. Environmental Protection Agency, 1986;).
24 However, Dowding (1988) has since concluded that pathogens which can benefit from
25 injured host cells or from disordered transport mechanisms are enhanced by pollution insult
26 to their hosts, whereas those that require a healthy mature host for successful invasion and
27 development are depressed by pollutant stress to their host.
28 This conclusion is supported by evidence that the development of diseases caused by
29 obligate parasites such as the rust fungi and bacterial pathogens is usually reduced by
30 O3. As shown by the observations summarized in Table 5-6, reductions in disease
31 development were observed in five of the nine studies of obligate fungal parasites listed
December 1993 5_63 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-6. OZONE-PLANT-PATHOGEN INTERACTIONS
«
o>
1
1
t
!
8
|
i
Host Plant
OBLIGATE FUNGI
Kidney bean
Barley
Cottonwood
Lilac
Oats
Wheat
FACULTATIVE FUNGI
Barley
Cabbage
Pathogen
Uromyces phaseoli
Erysiphe graminis
Melampsora medusae
Microsphaera alni
Puccinia coronata
Erysiphe graminis
Puccinia graminis
Puccinia graminis
Puccinia recondita
Drechslera teres
Gerlachia nivatis
Helminthosporium
sativum
Fusarium oxysporum
Effect of 03 on Disease
increased number of
smaller pustules
reduced infection but
greater spore production
reduced infection and
development
no effect
reduced infection and
development
increased infection and
development
reduced infection and
development
reduced development
reduced infection and
development
increased infection
increased infection
no effect
decreased development
Effect of Disease on 03 Response
reduced injury on severely
diseased leaves
not reported
not reported
not reported
no effect
not reported
reduced leaf injury
no,t reported
not reported
not reported
not reported
not reported
not reported
Reference
Resh and Runeckles (1973)
Heagle and Strickland (1972)
Cojenum et al. (1987)
Hibben and Taylor (1975)
Heagle (197Q)
Tiedemann et al. (1991)
Heagle and Key (1973a,b)
Heagle (1975)
Dohmen (19857)
Tiedemann et al. (1990)
Tiedemann et al. (1990),
Tiedemann et al. (1990)
Manning et al. (1971)
-------
TABLE 5-6 (cont'd). OZONE-PLANT-PATHOGEN INTERACTIONS
0\
Host Plant
FACULTATIVE
Corn
Cottonwood
Geranium
Onion
Potato
Soybean
Wheat
Jeffrey pine
Pathogen
FUNGI (cont'd)
Helminthosporium
maydis
Marssonina brunnea
Botrytis cinerea
Botrytis (3 spp. )
Botrytis cinerea
Altemaria solani
Alternaria solani
Fusarium oxysporum
Gerlachia nivalis
Helminthosporium
sativum
Helminthosporium
sativum
Septoria (2 spp.)
Septoria (2 spp.)
Heterobasidium annosum
Effect of Oj on Disease
increased development
increased infection
decreased infection
increased infection and
development
increased infection and
development
increased infection
increased infection
increased infection
increased infection
no effect
increased infection
increased infection
oncreased infection
increased development
Effect of Disease on O3 Response
not reported
not reported
not reported
not reported
not reported
not reported
not reported
increased leaf injury
not reported
not reported
not reported
not reported
not reported
not reported
Reference
Heagle (1977)
Coleman et al. (1988)
Krause and Weidensaul (1978)
Wukasch and Hofstra (1977a,b)
Manning et al. (1969)
Holley et al. (1985)
Bisessar (1982)
Damicone et al. (1987)
Tiedemann et al. (1990)
Tiedemann et al. (1990)
Tiedemann et al. (1991)
Tiedemann et al. (1990)
Tiedemann et al. (1991)
James et al. (1980)
-------
TABLE 5-6 (cont'd). OZONE-PLANT-PATHOGEN INTERACTIONS
3
if
_
OJ
OS
Os
d
£
H
d
O
g
3
o
d
o
#
o
Host Plant
FACULTATIVL FUNGI
Ponderosa pine
White pine
BACTERIA
Alfalfa
Soybean
White bean
Wild strawberry
NEMATQDES
Begonia
Soybean
Tobacco
aBased on studies using the
Pathogen
(cont'd)
Heterobasidiwn annosum
Vertcidadiella procera
Lophodermium pinastre
Xanthomonas alfalfae
Pseudomonas glycinea
Pseudomonas sp.
Xanthomonas phaseoli
Xanthomonas fragariae
Aphelenchoides
fragariae
Belonolaimus
longicaudatus
Heterodera glycines
Paratrichodorus minor
Pratylenchus penetrans
Meloidogyne hapla
protectant EDU (see Section
Effect of 03 on Disease
increased development
slightly increased
incidence
slightly increased
incidence
reduced development
reduced incidence
reduced infection
no effect
reduced incidence
reduced nematode
reproduction
stimulation or no effect
reduced nematode
reproduction
reduced nematode
reproduction
no effect
possible stimulation4
5.3.2.4.1.3).
Effect of Disease on 03 Response
not reported
not reported
not reported
reduced leaf injury
no effect
reduced leaf injury
reduced leaf injury
no effect
reduced leaf injury
not reported
not reported
reduced leaf injury
not reported
increased leaf injury
Reference
James et al. (1980)
Costonis and Sinclair (1972)
Costonis and Sinclair (1972)
Howell and Graham (1977)
Laurence and Wood (1978a)
Pell et al. (1977)
Temple and Bisessar (1979)
Lawrence and Wood (1978b)
Weber et al. (1979)
Weber et al. (1979)
Weber et al. (1979)
Weber et al. (1979)
Weber et al. (1979)
Bisessar and Palmer (1984)
-------
1 whereas increases were observed in all but four of the studies of facultative fungal pathogens.
2 Similarly, in four of the five bacterial systems, O3 reduced infection or disease development.
3 It should be noted that in three of the four studies of obligate fungi on which exposure to
4 O3 either had no effect or which resulted in stimulated fungal growth the pathogen was a
5 powdery mildews (Erysiphe, Microsphaerd). As discussed by Tiedemann et al. (1991), these
6 species constitute a special case because they are ectoparasites whose hyphae merely
7 penetrate the surface epidermal cells of the host plants leaves rather than the mesophyll
8 tissues within the leaves. They noted that Heagle and Strickland (1972) observed greater
9 pustule development of Erysiphe on exposed barley once infection was established, although
10 the pathogen was sensitive during the early stages of infection. Tiedemann et al. (1991)
11 suggest that the observed stimulations result from a differential weakening of the host's
12 resistance response to the pathogen.
13 In a few of the studies summarized in Table 5-6, effects of disease development on the
14 sensitivity of the host plant to O3 were noted. Heagle and Key (1973b) and Resh and
15 Runeckles (1973) confirmed the earlier observation of Yarwood and Middleton (1954) that
16 infection with obligate rust fungi could reduce the severity of acute injury caused by
17 exposure to O3. However, with Uromyces on bean, the "protection" was only noted on
18 severely infected leaves (Resh and Runeckles, 1973), and Heagle (1970) observed no such
19 effect with crown rust, Puccinia coronata, on oats.
20 Infection with bacterial pathogens and nematodes also tends to reduce the impact of
21 O3, and almost all studies of the interactions of O3 with virus infections appear to do so.
22 The previous criteria document (U.S. Environmental Protection Agency, 1986) reviewed the
23 supporting evidence from numerous studies with a range of host plants and viruses, and
24 noted only two studies in which O3 injury was apparently increased by virus infection
25 (Ormrod and Kemp, 1979; Reinert and Gooding, 1978). However, with tomato infected by
26 mosaic vinises, injury was reduced in the leaves of plants in which viral infection was well
27 established (Ormrod and Kemp, 1979). Two more recent studies have indicated either no
28 effect or variety-dependent increased sensitivity to relatively high O3 levels. Heagle et al.
29 (1991, 1992) found no effects of infection with several viruses on the response of two clonal
30 strains of white clover. On the other hand, Reinert et al. (1988) reported that three cultivars
31 of burley tobacco responded differently to O3 when infected with either tobacco etch virus
December 1993 5.57 DRAFT-DO NOT QUOTE OR CITE
-------
1 (TEV) or tobacco vein mottling virus (TVMR). Although TEV-infection resulted in the
2 protected all cultivars from O3-induced growth suppression, TVMV infection enhanced the
3 suppression of the growth of two cultivars, Burley 21 and Greenville 131, but had no effect
4 on the third, Burley 49.
5 With the exception of one field study demonstrating the suppression of O3 injury on
6 tobacco infected with tobacco mosaic virus (Bisessar and Temple, 1977), the other
7 investigations of virus interactions have all been conducted in laboratory or greenhouse
8 chambers, which raises the question of their relevance to field conditions. As noted in the
9 previous criteria document (U.S. Environmental Protection Agency, 1986), with few
10 exceptions, the reports of viral protection are probably of little commercial significance, but
11 may provide information at the mechanistic level of plant response. The same caveat is
12 equally applicable to the significance of protective effects of other obligate pathogens.
13 No studies appear to have been conducted of interactions involving disease-causing
14 mycoplasmas.
15 As in the case of plant-insect interactions, much more systematic study is needed before
16 it will be possible to provide any quantitative estimates of the magnitude of the interactive
17 effects. The patterns of pollutant modification of plant-pathogen relations suggested by
18 Dowding (1988) are partly supported by the limited evidence available for O3, but studies of
19 a wider range of plant-pathogen systems will be needed before it will be possible to provide
20 quantitative generalizations.
21
22 5.4.3.3 Oxidant-Plant-Symbiont Interactions
23 Exposure to O3 can modify the symbiotic relationships between plants and
24 microorganisms. In the case of Rhizobium, the important nitrogen-fixing symbiont of many
25 leguminous species, the adverse effects of exposure of the host plant reviewed in the
26 previous criteria document (U.S. Environmental Protection Agency, 1986) were all observed
27 at O3 levels of 0.3 ppm or greater. However, Flagler et al. (1987) observed a consistent
28 decline in total nitrogen-fixing activity of nodulated soybean roots with increasing
29 O3 concentrations up to 0.107 ppm (7-h/day seasonal average), with no effect on specific
30 nodule activity. In a greenhouse study of soybean plants exposed at three different growth
31 stages to a 12 h treatment in which the peak O3 concentration (at 6 h) was 0.2 ppm, Smith
December 1993 5-68 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al. (1990) observed a 40% decrease in specific nodule activity. Hence, there is limited
2 evidence to indicate adverse effects on Rhizobial nitrogen-fixation at O3 levels experienced in
3 polluted air.
4 The effects of O3 on mycorrhizal fungal symbioses have been reviewed by Manning
5 and Keane (1988) and McCool (1988). Seasonal exposures averaging 0.079 ppm O3 resulted
6 in a 40% reduction in the growth of the vesicular-arbuscular endomycorrhizal fungus,
7 Glomus fasciculatus, on soybean roots. However, mycorrhizal infection lowered the
8 O3-induced reduction in pod yield from 48 to 25% (Brewer and Heagle, 1983).
9 Once-weekly exposures of tomato plants to 0.3 ppm for 3 h retarded the early development
10 of the same fungus on tomato seedling roots, leading to reduced seedling growth (McCool
11 et al., 1982). Greitner and Winner (1989) reported that the increased availability of nitrogen
12 to alder seedlings resulting from the presence of root nodules containing the nitrogen-fixing
13 actinomycete, Frankia, enabled plants to recover their photosynthetic integrity rapidly after
14 exposure to O3 However, they did not investigate effects on symbiont.
15 In spite of the inconsistencies in the available evidence, it appears that rhizobial and
16 mycorrhizal growth is likely to be impaired as a consequence of long-term exposure to
17 oxidant stress, probably because of reduced allocation of photosynthate to the root system
18 (Chapter 7, U.S. Environmental Protection Agency, 1986). However, the implications of
19 such effects on mycorrhizae are particularly difficult to predict because of our inadequate
20 understanding of the functioning of the tree root-mycorrhiza-soil system.
21
22 5.4.3.4 Oxidant-Plant-Plant Interactions—Competition
23 In the field, the growth of any plant is to some extent dependent upon its ability to
24 compete for resources with its neighbors. Some are better competitors than others for light,
25 water, nutrients, and space. Grime (1977) characterized as "competitors" those with a rapid
26 growth rate associated with a capacity to adjust to rapidly changing conditions. Factors such
27 as light or soil nutrients are not available ad libitum, because of the mutual shading of leaves
28 within the canopy and root competition. Competition may be either intra- or inter-specific,
29 (i.e., the interference may be caused by neighboring members of the same species or by
30 neighboring individual of other species). The planting densities and row spacings adopted
31 for agricultural crops represent compromises between maximizing the number of plants per
December 1993 5.59 DRAFT-DO NOT QUOTE OR CITE
-------
1 unit area and the adverse effects of intra-specific competition. The planting densities and
2 row spacings adopted for agricultural crops represent compromises between maximizing the
3 number of plants per unit area and the adverse effects of intra-specific competition. Weeds
4 are typical interspecific competitors. Interspecific competition also occurs in mixed plantings
5 such as grass-clover forage and pasture plantings, and is an important feature of natural
6 ecosystems.
7 Although competition from weeds may contribute more to crop losses on a global scale
8 than any other factor, no studies appear to have been conducted on the effects of oxidant
9 pollution on such competition. On the other hand, a few crop mixtures have been studied.
10 A consistent finding with grass-clover mixtures has been a significant shift in the mixture
11 biomass in favor of the grass species (Runeckles and Bennett, 1977; Blum et al., 1983;
12 Kohut et al., 1988; Rebbeck et al., 1988; Heagle et al., 1989).
13 As the number of competing species increases, the interactions are more appropriately
14 dealt with at the ecological level, but, as demonstrated by the work of Evans and Ashmore
15 (1992), it is important to recognize that, because of the differential stresses imposed by
16 competition, the impact of O3 on the components of a mixture may not be predictable on the
17 basis of knowledge of the responses of the individual species grown in isolation. A similar
18 caution must be stated about extrapolating to field conditions results obtained in laboratory
19 studies in which competition may be minimal. However, the development and use of field
20 exposure systems have permitted many recent studies of crop species to be conducted at
21 normal planting densities and hence have incorporated interspecific competition as an
22 environmental factor. On the other hand, most forest tree studies have tended to be
23 "artificial" in their use of individual seedlings or saplings or spaced trees, even when
24 exposed in open-air systems (McLeod et al., 1992).
25 The significance of the effects of competitive interactions on the O3 response of the
26 competing species is thus largely unknown except for a few cases involving grass-legume
27 mixtures. However, these are far from typical because they only involve two species, one of
28 which is a legume with unique nitrogen nutrition conferred by the nitrogen-fixing capabilities
29 of Rhizobial symbionts. Hence, at the present time, our lack of knowledge of the effects of
30 O3 on competitive interactions leads to considerable uncertainty in attempting to assess the
December 1993 5-70 DRAFT-DO NOT QUOTE OR CITE
-------
1 impact of O3 on both managed and natural ecosystems, by extrapolation from effects on
2 individual species.
3
4 5.4.4 Physical Factors
5 The physical components of the plant's aerial environment are light, temperature,
6 humidity and surface wetness, while the physical, edaphic components affecting the plant
7 roots are temperature, soil moisture and soil salinity. The previous criteria document (U.S.
8 Environmental Protection Agency, 1986) also included soil fertility under this heading; in the
9 present review, this topic is dealt with separately in the section dealing with chemical factors
10 (Section 5.4.5). The effects of the physical climatic factors (light, temperature, and the
11 availability of water) on plant growth and survival are major determinants of the geographic
12 distribution of the earth's natural vegetation and of the distribution of agricultural lands and
13 the suitability of the crops grown on them. Because of the control that these factors exert
14 over plant growth, their variation, especially in the short term, can be expected to influence
15 the magnitude of plant responses to oxidants. As in the previous criteria document, the
16 factors are discussed individually, although their actions on plant growth and sensitivity are
17 closely inter-related. A brief integration of their effects is presented in Section 5.4.8, which
18 discusses the effects of global climate change.
19 At the tune of the previous criteria document, much of our knowledge of the effects of
20 these factors came from laboratory and greenhouse experimentation which focused the foliar
21 injury response to high exposures to O3 that exceeded those likely to be encountered in
22 ambient air. Since then, more information has become available on growth effects,
23 especially with regard to the key area of the interactions involving drought stress.
24
25 5.4.4.1 Light
26 Light influences plant growth through its intensity, quality (i.e., the distribution of
27 wavelengths), and duration (i.e., day length or photoperiod). Much of the early work on
28 light-oxidant interactions is largely of academic interest since light intensity and daylength
29 are uncontrolled in natural field situations. However, reduced intensities are needed for the
30 production of shade-grown cigar wrapper tobacco and in many commercial greenhouse
31 floriculture operations, in which photoperiod may also be controlled in order to induce
December 1993 5-71 DRAFT-DO NOT QUOTE OR CITE
-------
1 flowering. The general conclusion reported previously (U.S. Environmental Protection
2 Agency, 1986) is that susceptibility to foliar injury is increased by low intensities and short
3 photoperiods, although unpredictable responses had been observed when plants were
4 subjected to increased or decreased intensities during and after exposure to O3. One aspect
5 of increased susceptibility to low light intensities that needs to be emphasized concerns the
6 fact that many studies of oxidant effects have been conducted in controlled-environment
7 chambers in which the light intensities used have rarely approached those of natural sunlight,
8 and hence may have magnified the observed responses. Significant differences in the
9 amounts of foliar injury were observed on soybean plants grown in a growth chamber, a
10 shaded greenhouse or in an open-top chamber in the field when subsequently treated with a
11 standard O3 exposure, although the growing conditions other than light intensity and quality
12 were comparable (Lewis and Brennan, 1977). Factors other than light intensity must have
13 contributed to the observed differences since the descending order of sensitivity was
14 greenhouse-growth chamber-field chamber, although the average light intensities in the
15 greenhouse and growth chamber were 81 % and 18%, respectively, of those in the field
16 chamber.
17 Reduced light intensities have been measured in open-top chambers in the field,
18 resulting form the build-up of dust on the walls. However, Heagle and Letchworth (1982)
19 could detect no significant effects on soybean growth and yield in a comparison of plants
20 grown in unshaded open-top chambers and chambers to which shading cloth was applied.
21 At the mechanistic level, Darrall (1989) has reviewed the effects of light intensity and
22 suggests that, at high intensities, the potential for endogenous oxyradical production is
23 greatest and that this, combined with the production of oxyradicals from 03, might exceed
24 the leaf's detoxification ability. However, at lower intensities, decreased carbon assimilation
25 would limit the availability of energy for use in cellular repair.
26 In most species, light indirectly plays a major role in the opening and closing of
27 stomata. Since stomata therefore tend to close at night and open during the day, light
28 duration to some extent dictates whether or not O3 can be taken up by foliage from the
29 ambient air.
30
31
December 1993 5-72 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.4.4.2 Temperature
2 Temperature affects almost all physical processes and chemical reactions within the
3 plant. Hence it is the temperature within the plant tissues that is important. Although air
4 temperature dictates the overall heat balance in the surrounding air, the temperature of the
5 leaf is also determined by the absorption of infra-red radiation during the photoperiod (which
6 increases the leaf temperature), and the loss of water vapor through transpiration (which
7 provides evaporative cooling). Hence the effects of air temperature per se must be viewed in
8 the context of these other physical factors. It is therefore not surprising that the few early
9 studies of the effects of air temperature alone, using controlled environment chambers, led to
10 variable and conflicting results, as noted in the previous criteria document (U.S.
11 Environmental Protection Agency, 1986). In most of these studies, the relative humidity
12 (RH) and light intensity were held constant. In water-saturated air with a relative humidity
13 of 100%, the absolute humidity (or water vapor pressure) increases with temperature. Such
14 increases occur at all RHs. Therefore, at constant RH, the increase in absolute humidity
15 (or vapor pressure) with temperature, in turn, increases the vapor pressure deficit (VPD)
16 (i.e., the difference between the absolute humidity) (or vapor pressure) and that of
17 completely saturated air at the same temperature. Since VPD controls the rate of evaporation
18 of water, at constant RH the effects of temperature are unavoidably confounded with effects
19 on VPD. In a recent study with tomato seedlings, in which differences in VPD at different
20 temperatures were minimized, Todd et al. (1991) showed that, out of 11 growth variables
21 measured, the only significant modifications of the effects of O3 caused by temperature were
22 on stem fresh weight and specific leaf area (leaf area/leaf dry weight). The authors suggest
23 that VPD probably plays a more important role in determining sensitivity to O3 than
24 temperature.
25 Although transpiration rate is dependent on VPD, it is also regulated by the opening
26 and closing of stomata on the leaf surface, and factors such as O3 that cause stomatal closure
27 will indirectly cause leaf temperature to rise. Such stomatal and temperature changes have
28 been observed during exposure to O3 (Matsushima et al., 1985; Temple and Benoit, 1988).
29 An important O3-temperature interaction affecting trees and other woody perennials is
30 winter hardiness. Several studies have shown that exposures to O3 at realistic levels may
31 reduce the cold- or frost-hardiness of plants, as reviewed by Davison et al. (1988). Using
December 1993 5.73 DRAFT-DO NOT QUOTE OR CITE
-------
1 the pea plant as a laboratory model, Barnes et al. (1988) showed that daily 7-h exposures to
2 0.075 or 0.09 ppm O3 for seven days significantly reduced plant survival after exposure to
3 night-time temperatures that fell from 2 °C to -4 °C over a 2-h period and were than held at
4 -4 °C for a further 4 h.
5 Various responses of coniferous trees to the exposure to O3 during the growing season
6 and freezing temperatures during the following winter have been reported. With Norway
7 spruce, Eamus and Murray (1991) found that the recovery of photosynthetic rates after
8 freezing was slower in O3-treated seedlings. Brown et al. (1987) and Barnes and Davison
9 (1988) observed severe necrosis of the older needle classes of seedlings of some Norway
10 spruce clonal saplings exposed to O3 and then to freezing temperatures although other clones
11 showed no effect. Increased winter injury on plants exposed to O3 was also observed with
12 Sitka spruce (Lucas et al., 1988) and red spruce (Fincher et al., 1989). With loblolly pine,
13 Edwards et al. (1990) observed variable results, but Chappelka et al. (1990) reported that a
14 late winter frost resulted in severe tip die-back of the youngest needles of seedling trees
15 exposed to 1.7 (350 ppm-h) and 2.5 (433 ppnvh) times the ambient (272 ppm-h)
16 O3 concentration during the previous growing season, in contrast to the effects observed on
17 Norway spruce. The response also varied with plant genotype. A reason for the difference
18 may be that, in the study with Norway spruce, the freezing period occurred soon after
19 exposure to elevated O3 levels, while in the loblolly pine study the frost occurred in late
20 winter. The diversity of results led Eamus and Murray (1991) to develop a conceptual
21 framework which recognizes that even in severe winters there are brief periods of mild
22 temperatures that induce partial dehardening. While O3 decreases frost hardiness per se,
23 it also increases the trees' predisposition to dehardening given favorable conditions during
24 winter. Such dehardening puts O3-exposed trees at greater risk from subsequent low
25 temperatures.
26 In a greenhouse study with one-year old red spruce seedlings, Neighbor et al. (1990)
27 reported briefly that decreasing the level of nitric oxide at the time of exposure to
28 O3 prevented the appearance of O3-induced frost injury. They suggest that the effects
29 attributed to O3 are probably due to the combination of O3 with traces of NO above a critical
30 level. However, this effect has not apparently been investigated further.
December 1993 5.74 DRAFT-DO NOT QUOTE OR CITE
-------
1 In a study of the sub-tropical trees, citrus and avocado, in Florida, Eissenstat et al.
2 (1991) found that, although O3 could reduce frost hardiness, the effects were subtle, and the
3 authors concluded that the likelihood that frost resistance is adversely affected by current
4 O3 levels is slight.
5 The general consequences of global warming on O3 responses are discussed in
6 Section 5.4.8.
7
8 5.4.4.3 Humidity and Surface Wetness
9 A review of early investigations led to the conclusion that, in general, high relative
10 humidity (RH) tends to sensitize plants to O3 (U.S. Environmental Protection Agency, 1986).
11 Such a conclusion is supported on mechanistic grounds. Mclaughlin and Taylor (1982)
12 studies indicated that measured O3 uptake by bush bean plants increased with RH, and there
13 are several reports that, at high RH, the rapid decrease in stomatal conductance caused by
14 O3 at lower relative humidities is inhibited (Otto and Daines, 1969; Rich and Turner, 1972;
15 Elkiey and Ormrod, 1979). However, stomatal responses to O3 show considerable variability
16 among species and even among cultivars of the same species (Elkiey and Ormrod, 1979), and
17 hence it is to be expected that the patterns of the O3-RH interaction may not always be as
18 clear. Thus with yellow poplar, five consecutive daily exposures to 0.15 ppm for 7 h at
19 either 40% or 80% RH revealed considerable variation in stomatal conductance (Jensen and
20 Roberts, 1986). At 40% RH, there was a tendency for O3 to cause a decrease hi
21 conductance during the later exposures. Nevertheless, at 80% RH the conductances were
22 generally greater and tended to increase during the later exposures.
23 Surface wetness also influences the foliar uptake of O3 although there appear to have
24 been no studies undertaken to investigate the consequences of such uptake. Until recently, it
25 has been suggested that O3 uptake is reduced when foliage is wet because the stomata may be
26 covered with water (Hicks et al., 1987). However, Fuentes and Gillespie (1992) reported
27 that both wetness from dew or raindrops on the upper surface of maple leaves can
28 significantly increase O3 uptake. While this may partly be due to a stomatal response to
29 resulting increases in RH, the fact that increased uptake occurred in darkness when the
30 stomata were largely closed led them to suggest that direct uptake into the surface water is
December 1993 5.75 DRAFT-DO NOT QUOTE OR CITE
-------
1 the more important mechanism. However, no information is yet available as to the
2 consequences of such deposition.
3
4 5.4.4.4 Drought and Salinity
5 Short- and long-term variations in the availability of soil water have a profound
6 influence on plant growth. In some agricultural situations, the use of irrigation may
7 eliminate drought stress. However, the growth of crops and natural vegetation in many areas
8 will adversely affected by the varying degrees of water shortage that occur, both during a
9 growing season and from year to year. The previous criteria document (U.S. Environmental
10 Protection Agency, 1986) summarized earlier studies and concluded that drought stress
11 reduced the magnitude of adverse effects of 03 including injury and growth or yield
12 reductions. The effect was attributed to an increased rate of stomatal closure in
13 drought-stressed plants in response to O3 that effectively reduced uptake of the pollutant.
14 These conclusions were based almost exclusively on studies with crop species. Since then, a
15 number of studies with tree seedlings and further studies with crops species have shown that
16 the interaction between drought and O3 is more complex and variable than originally thought.
17 Heagle et al. (1988a) summarized the results of investigations into the drought-
18 O3 interaction in six soybean and three cotton studies, and one study each of alfalfa and a
19 clover-fescue mixture. These studies were undertaken as part of the National Crop Loss
20 Assessment Program (NCLAN) (Heck et al., 1984). The results of these investigations are
21 included in Table 5-7. Significant interactions between O3 and drought stress (soil moisture
22 deficit, SMD) were reported only in three soybean and two cotton studies and the alfalfa
23 study. The interaction was usually revealed by the fact that the clear negative relationships
24 between yield and O3 exposure observed with watered plants was either much reduced or
25 could not be demonstrated with drought-stressed plants, bearing in mind that in most of these
26 situations the yields were already depressed by the SMD. As a result, the lack of any
27 significant response to O3 in some cases with such stressed plants reflects the decreased
28 range of yield responses within which an O3 effect could operate. However, as shown in
29 Table 5-7, Heggestad et al. (1988) found with Forrest soybean that SMD significantly
30 enhanced the effects of low O3 exposures. Heagle et al. (I988a) therefore were forced to
31 conclude that the suppression of the response to O3 caused by drought appeared to be
December 1993 5-76 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-7. FIELD STUDIES OF OZONE-DROUGHT STRESS INTERACTIONS IN CROP SPECDZS
I
I— »
1
^
d
g
H
1
d
o
1
0
Estimated Yield Loss (%) per Seasonal Mean
O3 Concentration (ppm)
Crop/Cultivar Year
Soybean
Williams 1982
Williams and 1983
Corsoy 79
Williams
Forrest 1982
Davis 1983
Yield
Davis 1984
Corsoy 79 1986
Young 1986
Response
Yield
Yield
Yield
Root length
Root length
Yield
Yield
Yield
DS
DS
Yield
Yield
Yield
Yield
Significant
Interaction
No
WW
DS
WW
DS
WW
DS
WW
No
WW
DS
No
0.04
7
7
6
0.05
13
13
11
no
0.06
19
18
15
significant O3
[33 36
3
13
4
4
2
0
6
9
21
7
no
7
4
0
11
21
28
12
significant 03
12
8
0
17
0.07
24
24
19
effect
/»
52]3
39
35
16
effect
18
13
0
25
0.08
30
30
23
60
41
21
24
21
1
34
Reference
Heggestad et al. (1985)
Heggestad and Lesser (1990)
Heggestad and Lesser (1990)
Heggestad et al. (1988)
Heggestad et al. (1985)
Heggestad and Lesser (1990)
Heagle et al. 1987)
Heagle et al. (1987)
Irving et al. (1988)
Miller et al. (1989)
-------
S5
8
u
• >. M
81
O K
^1
0(2
S3 *
O.S
81
i§
% g
3 i
o
O
o
2
c/
Cultivar
C
December 1993
•a
u
!
£
282
^, »«H ^/
***iS ^"^
•a -a is
"S -S «
43 « ja
III
H DC tS
82 R s
3^ ?
S
•a
u
t
ss
en I-H vo r-
c« o
55 S5
i
1
as
CSt-
o o
cs
000000
O\O\O\
(S en
-------
1 dependent on the severity of the SMD-induced stress. This conclusion was supported
2 empirically by the fact that the interaction was observed in hot, dry summers, but not in
3 cool, cloudy seasons. However, because different measures of SMD or SMD-induced stress
4 were used in different studies, it is not possible to quantify the relationship between the
5 suppression the O3 response and the level of drought stress. Furthermore, soil conditions at
6 different sites and the depth of the water table in the soil also appeared to influence the
7 interaction through effects on the vertical distribution of root growth (Heggestad et al.,
8 1988). Moser et al. (1988) showed that impact of O3 on the growth and yield of bush bean
9 plants was reduced more by a period of SMD early in the reproductive phase of growth than
10 at a later stage.
11 A retrospective analysis of three years' data involving four soybean cultivars led
12 Heggestad and Lesser (1990) to conclude that, regardless of whether or not drought stress
13 reduced the impact of ozone, in the majority of cases the yield curves showed comparable
14 slopes with increasing O3 exposure. This led them to question the justification for the
15 20% reduction in sensitivity to O3 used by King (1988) in modelling the
16 drought-O3 interaction.
17 Brennan et al. (1987) suggested that the normal experimental protocols used in most
18 NCLAN studies, which called for the use of irrigation to avoid possible complications due to
19 drought, might have biased the yield loss data for soybean because it increased plant
20 sensitivity to O3. However, Heggestad and Lesser (1990) found no evidence to support this
21 suggestion hi view of the comparable estimates of yield losses predicted by the O3-response
22 curves.
23 Bytnerowicz et al. (1988) found no interaction between SMD and O3 in 18 desert
24 annual species. However, moderate SMD rendered the tropical fibre plant, kenaf, less
25 sensitive to O3 although sensitivity was enhanced by sever water stress (Kasana, 1992).
26 A field survey of milkweed plants in two areas in the mid-Ohio River Valley revealed much
27 less foliar injury attributable to O3 in 1988, a dry year in which the maximum concentration
28 recorded nearby reached 0.2 ppm, than in 1989, a year with ample precipitation and a nearby
29 maximum of only 0.12 ppm (Showman, 1991).
30 Although there have been several recent studies of the effects of O3 exposure and
31 drought stress on tree species, they have little in common with respect to the treatments
December 1993 5.79 DRAFT-DO NOT QUOTE OR CITE
-------
1 applied or the measurements made. However, clear demonstrations of significant interactions
2 have been obtained with beech, poplar and loblolly pine seedlings. Davidson et al. (1992)
3 found that although O3 reduced root growth in well-watered plants, SMD reversed this
4 inhibition and led to slight O3-induced stimulations. Drought reduced foliar injury caused by
5 O3 to poplar (Harkov and Brennan, 1980), ponderosa pine (Temple, 1992) and loblolly pine
6 (Meier et al., 1990). In poplar, the effect was attributed to the reduced stomatal conductance
7 observed which reduced O3 uptake. Similar effects on stomatal conductance were observed
8 in Norway spruce and sitka spruce (Dobson et al., 1990). In ponderosa pine, SMD also
9 countered the inhibitory effects of O3 on needle growth and retention (Temple et al., 1993).
10 Tseng et al. (1988), however, observed no effects of O3 on Fraser fir grown under three
11 levels of SMD. No consistent patterns were found with various physiological measurements
12 made on red spruce seedlings subjected to both O3 and drought (Roberts and Cannon, 1992).
13 Lee et al. (1990) observed reduced root conductivity in the second drought cycle following
14 exposure to O3.
15 Thus there is some evidence from tree species to support the view that drought stress
16 may reduce the impact of O3. However, the work with trees provides no additional
17 information to help in resolving the quantitative nature of the drought-O3 interaction.
18 Although drought stress may be the result of insufficient rainfall, conditions of effective
19 SMD may also be induced by excessive soil salinity. Laboratory studies showed that
20 increased salinity could reduce the impact of O$ on injury and yield on various crops, as
21 reviewed in the previous criteria document (U.S. Environmental Protection Agency, 1996).
22 However, in a more recent field study with alfalfa, Olszyk et al. (1988) found no overall
23 interaction between O3 and salinity on growth or yield. Although salinity decreased the
24 number of empty nodes caused by exposure to above ambient levels of Oj, the effect was
25 only statistically significant for the second of four harvests. In general, salinity was found to
26 be a more harmful to alfalfa growth than exposure to O3, but, as pointed out by Olszyk et al.
27 (1988), the amount of information available is insufficient to permit the development of
28 models for estimating losses due to O3-salinity combinations.
29 The bulk of the available evidence supports the view that drought stress may reduce the
30 impact of O3 on plants. However, it must be emphasized that, in terms of growth and
December 1993 5-80 DRAFT-DO NOT QUOTE OR CITE
-------
1 productivity, any "protective" benefit will be offset by the effects of SMD per se, as noted in
2 the previous criteria document (U.S. Environmental Protection Agency, 1986).
3 The O3-water interaction is not confined to the effects of SMD on direct plant response
4 to O3. Numerous studies have shown that O3 may affect various aspects of plant water
5 status, including water use efficiency (WUE), the ratio of the rates of photosynthetic carbon
6 gain and transpirational water loss. For example, Reich et al. (1985) observed that daily
7 exposures to 0.13 ppm O3 for 6.8 h resulted in a 25 % reduction in WUE in well-watered
8 Hodgson soybean, when compared to exposure to 0.01 ppm. Similar findings have been
9 reported for alfalfa (Temple and Benoit, 1988) and radish (Barnes and Pftrrmann, 1992).
10 However, WUE is a complex resultant of both stomatal conductance and the activity of the
11 photosynthetic system, both of which may be independently affected by O3. Genetic or
12 environmentally induced difference in the relative sensitivities of the stomatal and
13 photosynthetic components will dictate the nature and magnitude of any effect of O3 on
14 WUE. Thus, with radish and soybean, Greitner and Winner (1988) observed effects on
15 stomatal conductance and photosynthetic CO2 assimilation that translated into O^-induced
16 increases in WUE, they point out, that this advantageous increase far outweighed the adverse
17 effects of O3 on growth.
18 However, these reports concern herbaceous weedy species, and there appears to be only
19 one report concerning tree species. Johnson and Taylor (1989) reported that exposure to
20 higher than ambient levels of O3 results in adaptation to a more efficient use of water by the
21 foliage of loblolly pine seedlings. The corollary to this observation is the trees exposed
22 continuously to low O3 levels may be more sensitive to recurrent drought stress than those
23 grown under higher exposure levels. The corollary to this observation is that trees exposed
24 continuously to low O3 levels may be more sensitive to recurrent drought stress than those
25 grown under higher exposure levels. As with most studies of tree species, these observations
26 were made on tree seedlings, and their relevance to mature trees has still to be established.
27 It is therefore clear that not only does drought have a pronounced effect on the response
28 of most species to O3, but that O3 also may also modify plant water relations including
29 conferring drought tolerance. However, more study will be needed before it will be possible
30 to generalize about the implications of the latter effect, and its importance to forest
31 ecosystems.
December 1993 5-81 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.4.5 Nutritional Factors
2 All land plants require an adequate supply of essential mineral elements from the soil in
3 order to avoid adverse effects on growth and survival resulting from mineral deficiencies.
4 Two of the essential elements needed for growth are nitrogen and sulfur, and although these
5 are normally obtained from the soil through the root system, the plant's needs can, at least in
6 part, also be met by the uptake of pollutant gases such as NO2 and SO2. Other nutrients
7 such as phosphorus, potassium, magnesium and calcium are generally only available from the
8 soil.
9 A supply of elements such as nitrogen, potassium, phosphorus, sulfur, magnesium and
10 calcium is essential for plant growth, but optimal growth requires that the supply be
11 balanced; with insufficiency (or excess) of any of them, growth will be sub-optimal. Not
12 surprisingly, therefore, nutrient imbalance has been shown to affect response to O3, although
13 the previous criteria document (U.S. Environmental Protection Agency, 1986) concluded that
14 work to that date had not clarified the relationship between soil fertility and sensitivity to O3,
15 largely because of the differences in nutrients and species selected for study and the
16 experimental conditions used. This conclusion is still valid, in spite of the results of a
17 limited number of more recent studies, and is not surprising in view of the vast number of
18 possible permutations and combinations of nutrient elements and their levels that may exert
19 effects on O3 response. A comprehensive summary of the relevant studies is presented in
20 Table 5-8.
21 Most information concerns nitrogen. Inspection of Table 5-8 shows that in three of the
22 13 studies, increased N supply increased susceptibility to foliar injury or enhanced adverse
23 effects of O3 on growth; three of the studies showed the opposite effects; in three studies,
24 injury was greatest at normal N levels and less at lower or higher levels; in one study, injury
25 was least at normal N levels; and in three studies, no interactions were observed.
26 Knowledge of the tissue N levels resulting from the fertilizer treatments, as recommended by
27 Harkov and Brennan (1980), might resolve these contradictions, but these were not reported
28 in most studies. The contradictory evidence for tobacco was attributed by Menser and
29 Hodges (1967) do the different cultivars used.
December 1993 5-82 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-8. OZONE-SOIL NUTRIENT INTERACTIONS
Based in part on Cowling and Koziol (1992)
Species
Response to Increase in
Nutrient Level
Reference
Nitrogen (N)
Loblolly pine
Ponderosa pine
Poplar
Yellow poplar
Ladino clover/tall fescue
Mangel
Radish
Spinach
Tobacco
Decreased reduction of growth
due to O3
No injury or growth interactions
Maximum injury in mid range but
no growth interaction
No growth interaction
No growth interaction
Increased injury
Increased reduction of growth due
toO3
Reversal of reduction of shoot
growth due to O3
Increased injury
Decreased injury
Minimum injury in mid range
Maximum injury in mid range
Maximum injury in mid range
Tjoelker and Luxmoore (1991)
Bytnerowicz et al. (1990)
Harkov and Brennan (1980)
Tjoelker and Luxmoore (1991)
Monies et al. (1982)
Brewer et al. (1961)
Ormrod et al. (1973)
Pell et al. (1990)
Brewer et al. (1961)
Menser and Street (1962)
Macdowall (1965)
Leone et al. (1966)
Menser and Hodges (1967)
Phosphorus (P)
Radish
Tomato
No growth interaction
Increased injury
Ormrod et al. (1973)
Leone and Brennan (1970)
Potassium (K)
Norway spruce
Pinto bean
Soybean
Decreased reduction of CO2
assimilation due to O3
Decreased injury
Decreased injury
Keller and Matyssek (1990)
Dunning et al. (1974)
Dunning et al. (1974)
Sulfur (S)
Bush bean
Decreased injury
Adedipe et al. (1972)
Magnesium (Mg)
Loblolly pine
No growth interaction
Edwards et al. (1992b)
General fertility (N-P-K)
Bush bean
Soybean
Decreased injury
Maximum injury and growth
reduction in mid range
Heck et al. (1965)
Heagle (1979)
December 1993
5-83
DRAFT-DO NOT QUOTE OR CITE
-------
1 The limited evidence for phosphorus, potassium and sulfur consistently indicated a
2 decrease in sensitivity with increased nutrient level. With respect to general fertility, both
3 studies listed in Table 5-8 revealed decreased sensitivity to O3 at high levels of nutrient
4 supply, although with soybean, nutrient-deficient plants also showed decreased sensitivity.
5 Heagle (1979) found that although injury and growth reductions tended to be greatest at
6 nonnal levels of fertility, the effects were dependent upon the rooting medium used; in media
7 containing peat, the impact of O3 on growth was least at the lowest fertility level.
8 Cowling and Koziol (1982) have suggested that, in spite of the apparent contradictory
9 evidence regarding the effects of nutrition on O3 response, there is evidence to support the
10 hypothesis that differences in sensitivity are ultimately linked to changes in the status of
11 soluble carbohydrates in the plant tissues (Dugger et al., 1962). However, this hypothesis
12 has yet to be systematically tested.
13 Nutritional nitrogen and sulfur can also be supplied directly to foliage in the form of
14 nitrogen and sulfur oxides. The interactions of these gaseous pollutants with O3, dealt with
15 in the next section, focus on toxic rather than nutritional effects. However, one example of a
16 beneficial effect concerns nitrogen pentoxide, N2O5. Since N2O5 is produced in trace
17 amounts by high voltage corona discharge O3 generators, it may contaminate O3 produced
18 from air by such generators for use in studies of effects of O3 on vegetation, unless the
19 O3 stream is first passed through a water scrubber. Brown and Roberts (1988) reported that
20 deposition of the nitrate formed by hydration of trace amounts of N2O5 in unscrubbed
21 O3 significantly increased the nitrogen status of the exposed plants, which may have
22 confounded the effects attributed to O3.
23
24 5.4.6 Interactions with Other Pollutants
25 The concurrent or sequential exposure of vegetation to different gaseous air pollutants
26 has been found to modify the magnitude and nature of the response to individual pollutants.
27 Some of the early work on the effects of gaseous pollutant combinations, reviewed in the
28 previous criteria document (U.S. Environmental Protection Agency, 1986), is of academic
29 interest with little relevance to the present review because of the levels of exposure and the
30 exposure profiles used and the fact that the experimental regimes usually involved concurrent
31 exposures to two or more pollutants repeated daily. Lefohn and Tingey (1984) and Lefohn
December 1993 5-84 DRAFT-DO NOT QUOTE OR CITE
-------
1 et al. (1987) reviewed the patterns of co-occurrence of O3, SO2, and NO2 in urban, rural,
2 and remote sites in the United States for the years 1978 to 1982 and found that
3 co-occurrences were usually of short duration and occurred infrequently. They noted that the
4 most frequent types of co-occurrence were either purely sequential or a combination of
5 sequential and overlapping exposures of short duration. Accordingly, the present review will
6 focus on the evidence from experiments which simulated these naturally occurring patterns of
7 combined exposure or at least which used exposure levels in the ranges of those occurring in
8 polluted air. An exception is the co-occurrence of O3 and PAN, which are both components
9 of photochemical oxidant.
10 Over the past decade, the effects of pollutant mixtures have been reviewed by
11 Wolfenden et al. (1992), Shriner et al. (1990), Mansfield and McCune (1988), Torn et al.
12 (1987), Lefohn and Ormrod (1984), Reinert (1984), and Runeckles (1984).
13
14 5.4.6.1 Oxidant Mixtures
15 Because of their photochemical origins, elevated levels of O3 and PAN can occur
16 simultaneously. There appear to have been no further investigations of the effects of
17 simultaneous or sequential exposures since the limited number of studies reviewed in the
18 previous criteria document (U.S. Environmental Protection Agency, 1986). Hence, there is
19 no reason to question the general conclusion, based on the work of Tonneijck (1984) and
20 Nouchi et al. (1984), that the two gases tend to act antagonistically in both concurrent and
21 sequential exposures. Hydrogen peroxide, H2O2, is also a component of photochemically
22 polluted atmospheres. Although Ennis et al. (1990) reported reduced stomatal conductances
23 in red spruce needles exposed to a mixture of 63, SO2 and H2O2, no studies have been made
24 of O3/H2O2 interactions.
25
26 5.4.6.2 Sulfur Dioxide
27 Since SO2 originates from point sources of combustion, the occurrence of high ambient
28 concentrations at a given location is usually episodic because of its dependence upon wind
29 speed, wind direction, and distance from the source. However, aggregations of point sources
30 can lead to more widespread but less marked increases in ambient SC^ levels. Thus the
31 potential exists for elevated O3 exposures to be superimposed on patterns of SO2 ranging
December 1993 5-85 DRAFT-DO NOT QUOTE OR CITE
-------
1 from severe fluctuations to almost steady low-level concentrations. Concern over the
2 importance of O3-SO2 interactions dates from the observations of Menser and Heggestad
3 (1966) that simultaneous exposures of tobacco to SO2 and O3 acted synergistically (i.e., the
4 effects of the mixture were greater than the sum of the responses to either pollutant alone).
5 Indeed, in the Menser and Heggestad study, foliar injury was found to result from exposure
6 to mixtures although exposures to either gas alone at the same concentrations as in the
7 mixtures did not result in injury.
8 Although much of the early work was concerned with foliar injury responses to
9 simultaneous exposures to high levels of O3 and SO2, more recent studies have tended to
10 focus on the consequences of repeated exposures to lower level mixtures or sequences on
11 growth or yield. Several have been aimed at obtaining statistical evidence for the existence
12 of interactions. For example, Ashmore and Onal (1984), studying six cultivars of barley,
13 found that SO2 at 0.065 ppm for 6 h, an exposure that induced no adverse effects, acted
14 antagonistically to a 6-h exposure to 0.18 ppm O3, causing significant decreases in foliar
15 injury, ranging from 46% to as much as 95%. However, only one cultivar, Golden Promise,
16 showed a significant interaction on yield, with SO2 completely reversing the decrease caused
17 by O3 alone. The results could not be explained by effects on stomatal uptake since stomatal
18 conductances were found to be highest in the mixture. In contrast, with pea, Olszyk and
19 Tibbetts (1981) reported that O3+SO2 caused the same degree of stomatal closure as SO2
20 alone. A similar antagonism to that observed on Golden Promise was also observed in a
21 field study of "Arena" barley (Adaros et al., 1991a) and of spring rape (Adaros et al.,
22 1991b). However, with "Tempo" spring wheat, a synergistic interaction was observed: the
23 adverse effect of O3 on yield (-26%) was increased to -38% by SO2 which by itself only
24 reduced yield by 7% (Adaros et al., 1991a). On the other hand, neither Amundson et al.
25 (1987) nor Kohut et al. (1987) observed any interaction in a field study with "Vona" winter
26 wheat. Irving et al. (1988) observed no interaction on field corn.
27 In a series of experiments in which exposure to O3 or an O3/SO2 mixture was preceded
28 by exposures to SO2 alone, an antagonistic response was observed on foliar injury to white
29 bean (Hofstra and Beckerson, 1981). In contrast, the responses of cucumber and radish were
30 synergistic, while there was no interaction on soybean or tomato. However, when followed
December 1993 5-86 DRAFT-DO NOT QUOTE OR CITE
-------
1 by exposure to an O3/SO2 mixture, S02 pretreatment resulted in an increase in injury to
2 white bean, decreases on cucumber and tomato, and no effect on soybean and radish.
3 Field studies with soybean using an air-exclusion system to provide a range of
4 exposures to O3 and SO2 at ambient and sub-ambient levels revealed an antagonistic
5 interaction on yield at low concentrations (Jones et al., 1988). However, Kress et al. (1986)
6 found no interaction in a soybean field study using open-top chambers. No interactions were
7 found with potato (Pell et al., 1988) or with a red clover-timothy forage mixture (Kohut
8 etal., 1988).
9 From the foregoing, it is apparent that no clearer pattern of the interactive effects of
10 O3 and SO2 on crops has emerged since the previous criteria document (U.S. Environmental
11 Protection Agency, 1986). The same is true for the responses of tree species.
12 With tree seedlings, Chappelka et al. (1988a) observed no interaction on white ash.
13 Although a synergistic interaction was found on root growth of yellow poplar (Chappelka
14 et al., 1985), only additive interactions were found on the growth of other parts of the plant.
15 In a unique study, Kargiolaki et al. (1991) noted that SQ2 reduced the accelerated leaf
16 senescence caused by O3 on two poplar clones, but had no effect on other clones. They also
17 observed additive or less than additive interactions on the formation of intumescences, due to
18 hypertrophy of the stems, and bark cracking. They attributed the differences in clonal
19 response to differences in the levels of pollutant-induced ethylene evolution.
20 Sulfur dioxide reversed the inhibition of photosynthesis caused by exposure to (Jj in
21 two lichen species, Flavoparmelia caperata and Umbilicaria mammulata (Eversman and
22 Sigal, 1987).
23 Several studies have attempted to quantify the magnitudes of joint responses to O3 and
24 SO2. The earliest (Macdowall and Cole, 1971) showed that the synergistic injury response
25 of tobacco occurred at concentrations of SO2 less than the threshold for SO2 injury, but not
26 less than the O3 threshold. Oshima (1978), working with kidney bean, found that the
27 synergistic reduction due to intermittent exposures to O3 was linear through a range of
28 O3 concentrations achieved by varying degrees of filtration of ambient air (expressed as 10 to
29 90 ppm-h of concentrations greater than zero), although the threshold for an O3 response was
30 approximately 47 ppm-h.
December 1993 5.37 DRAFT-DO NOT QUOTE OR CITE
-------
1 A selection of statistical models of injury- or yield responses to O3/SO2 is listed in
1 Table 5-9. It is immediately apparent that the models reveal no consistent patterns of
3 response. In part, this is because they were developed on the basis of individual experiments
4 conducted under different environmental conditions at different locations in different years.
5 Although each model was statistically significant, it was based on a unique data set. One
6 study with soybean indicated an antagonistic interaction (Heagle et al., 1983) but another
7 indicated no interaction (Kress et al., 1986). Cucumber (Hofstra et al., 1985) and snap bean
8 (Heggestad and Bennett, 1981) were reported to respond synergistically, while white bean
9 responded antagonistically (Hofstra et al., 1985).
10 All that can be concluded from these studies is that the type of interaction, and whether
11 or not one exists, is probably highly dependent upon species and cultivar, and possibly
12 dependent upon other environmental variables. The available evidence is insufficient to be
13 able to decide in which way, and to what extent, SO2 exposure will influence the effects of
14 O3 on a particular species or cultivar at a particular location. The original observations of
15 synergism (Menser and Heggestad, 1966) certainly is not a general response.
16
17 5.4.6.3 Nitrogen Dioxide
18 As with SO2, most of the few studies of O3/NO2 interactions that have utilized realistic
19 concentrations have involved mixtures of the pollutants. Adaros et al. (1991a) found in a
20 2-year study of two cultivars each of barley and spring wheat that significant interactions
21 could only be detected on wheat yield in one growing season. With both cultivars the
22 interaction was antagonistic. NO2 also reduced the adverse effect of O3 on the yield of
23 spring rape (Adaros et al., 1991b). Foliar injury to sunflower caused by daily exposures to
24 O3 (0.1 ppm, 8 h) was increased by continuous exposure to 0.1 ppm NO2 (Shimizu et al.,
25 1984). Plant dry weight was decreased by O3 + NO2 relative to growth in O3 alone, but
26 since O3 exposure resulted in a slight increase in dry weight relative to the controls, the
27 growth in the mixture and in the controls did not differ significantly.
28 The results of a study of seven tree species exposed to 0.1 ppm Qj and/or 0.1 ppm
29 NO2 for 6 h/day for 28 days (Kress and Skelly, 1982) were reported in detail in the previous
30 criteria document (U.S. Environmental Protection Agency, 1986). However, although
31 several growth interactions were noted in the review, the only statistically significant effect
December 1993 5-88 DRAFT-DO NOT QUOTE OR CITE
-------
oo
TABLE 5-9. SOME STATISTICAL MODELS OF COMBINED OZONE AND
SULFUR DIOXIDE RESPONSES
S"
Co
\o
OJ
Species
Corn
Golden Jubilee
Type of
Interaction
Synergistic
Model
Injury=-11.39+5.4711n(IHT)-9.59[03]+11.81[S02]-86.63[SOJ2+428.95[03][S02]
Reference
Deveau et al.
Cucumber
National Pickling
Snap bean
Maple Arrow
interaction
White bean
Seafarer
Norchip
(IHT=initial plant height, used as a covariate; [O3] and [SOJ: ppm)
Synergistic Injury=2.70-f 1.95 n; atfSOJ=0.10 ppm
Injury=2.40+0.21 n; at[SOJ=0.05 ppm
Injury=2.39+0.39n; at [SO2]=0.03 ppm
Injury =1.86+0.166 n; at [SOJ=0.02 ppm
(n=number of daily 8-h SO2 exposures; O3 exposure: 0.15 ppm, 6 h)
Additive; no Injury=4.44+34.19[O3]+19.98[SO2]([O3] and [SOJ: ppm)
Antagonistic Injury=6.31-0.90 n; at [SO2]=0.10 ppm
Injury=5.95-0.45n; at [SO2]=0.05 ppm
(n=number of daily 8-h SO2 exposures; O3 exposure: 0.15 ppm, 6 h)
Additive; no Yield= 1.27-0.0037[O3]+0.00092[SO2]
interaction (Yield=number of Grade #1 tubers per plant; [Og]: ppm, 10 h/day seasonal mean;
[SOJ: ppm, 3 h/day)
(1987)
Hofstra et al.
(1985)
Deveau et al.
(1987)a
Hofstra et al.
(1985)
Pell et al. (1985)
Antagonistic Polynomial model:
Yield=534.5-3988.6[O3]-479.7[SOJ+ 2661.0[O3][SOJ + l.C
Weibull model:
Yield=531 x expKEOJ/0.133)] x exp[-([SO2]/0.892)]
(Yield=g/m of row; [O3]: ppm, seasonal 7 h/day mean; [SO2]: ppm, seasonal 4 h/day
mean)
Heagle et al.
(1983)
-------
vfa
o
§
TABLE 5-9 (cont'd). SOME STATISTICAL MODELS OF COMBINED OZONE AND
SULFUR DIOXIDE RESPONSES
Species
Amsoy-71 and
Corsoy-79 (pooled)
Tomato
New Yorker
Type of
Interaction
No
interaction
Model
Yield= 1934.4*exp[-([03]/0. 124)2'666]*exp[-([S02]/l .5 1 1)1 -044]
(Yield = kg/ha; [O3]: ppm, seasonal 7 h/day mean; [SO2]: ppm, seasonal 4 h/day mean)
Injury=-75.78 + 20.48m[PI]-29.16[O3]+l,016[O3]2+9.02tSO2]-17.29[SO2]2+258.76
[O3][SO2](PI=plastochron index, used as a covariate; [O3] and [SO2]: ppm)
Reference
Kress et al
Deveau et
(1987)a
. (1986)
al.
aReport includes models for other growth variables.
8
-------
1 was on top growth of pitch pine, in which NO2 reversed a growth stimulation caused by
2 exposure to O3. In contrast, although Yang et al. (1982) also observed an antagonistic
3 interaction on the needle dry weights of two eastern white pine clones, in these cases NO2
4 reversed the adverse effect of O3.
5 There appear to have been only three studies using sequential exposures of O3 and
6 NO2. Runecldes and Palmer (1987) exposed radish, wheat, bush bean and mint daily to
7 0.08 to 0.1 ppm NO2 for 3 h (09:00 to 12:00) or to 0.08 to 0.1 ppm O3 for 6 h (12:00 to
8 18:00), or to the two gases in sequence. With each species except mint, pretreatment with
9 NO2 significantly modified the growth responses to O3. In radish and wheat, the two gases
10 acted conjointly to reduce growth more than O3 alone, while in bean NO2 was antagonistic.
11 In studies with tomato, Goodyear and Ormrod (1988) found that sequential exposure to
12 0.08 ppm O3 for 1 h followed by 0.21 ppm NO2 for 1 h significantly reduced growth.
13 No significant effects were found when the sequence was reversed or the two gases were
14 used as a mixture. However, since the study did not include a treatment with O3 alone,
15 no information was obtained as to how NC^ may have influenced the response to O3.
16 Bender et al. (1991) exposed kidney beans in open-top chambers in the field to the sequence:
17 O3 (08:00 to 16:00, ambient+0.50 ppm) followed by NO2 (16:00 to 08:00,
18 ambient+0.3 ppm), during two growing seasons. No significant treatment effects on growth
19 were observed in 1988, but in 1989 a significant interaction on total plant biomass was noted
20 after 48 days; the overnight NO2 exposures negated the inhibition caused by O3 with a
21 change from -32% to +14%, relative to the controls. This type of response is similar to that
22 observed on bean by Runeckles and Palmer (1987).
23 With such limited information, it is not possible to generalize, particularly since
24 antagonistic and additive responses have been reported even for individual species.
25 However, since on a daily basis changes in NO2 levels tend to lead to maxima at times when
26 O3 levels are lowest, the evidence is sufficiently compelling to indicate that modifications of
27 the O3 response as a result of increased NO2 are highly probable.
28
29 5.4.6.4 Hydrogen Fluoride and Other Gaseous Pollutants
30 The adverse effects of hydrogen fluoride, HF, released from the aluminum smelting
31 process and superphosphate fertilizer manufacture are well documented, but information
December 1993 5.9! DRAFT-DO NOT QUOTE OR CITE
-------
1 about possible HF/O3 interactions are limited to a single study. MacLean (1990) reported
2 that exposures of corn plants on alternate days to 4 h at ljug/m3 fluorine as HF or 0.06 ppm
3 O3 showed reduced rates of senescence, compared with plants exposed only to O3.
4
5 5.4.6.5 Acid Deposition
6 The effects of acidic deposition have been extensively reviewed by Shriner et al.
7 (1990). Although concerns over the possible role of exposures to acid rain or acid fog and
8 O3 in the forest decline syndrome led to several studies with forest tree species, studies have
9 also been conducted on crops. Of over 80 recent reports of studies on over 30 species, more
10 than 75% of the reports indicated no significant interactions between O3 and acidity of
11 simulated acid rain (SAR) or acid fog. The reports are summarized in Table 5-10.
12 In 63 studies, there was either no effect of one or other of the pollutants (usually acid rain)
13 or the effects of both pollutant stresses were simply additive.
14 However, in other studies, statistically significant interactions have been reported for
15 several species, as also shown in Table 5-10. For example, although a large number of
16 studies of loblolly pine revealed no interaction, Qiu et al. (1992) reported significant
17 interactions on foliar and stem biomass with seedling trees of an O3-sensitive family.
18 However, since the study failed to show a significant main effect of acidity of the SAR, the
19 authors question whether the interaction is meaningful.
20 With Norway spruce, antagonistic interactions were noted on stomatal conductance
21 (Barnes et al., 1990a) and dark respiration (Barnes et al., 1990b). In contrast, Eamus and
22 Murray (1991) reported greater than additive effects of O3 and acid mist on photosynthetic
23 rates. However, no interactions were noted in nine other investigations (Table 5-10).
24 Kohut et al. (1990) observed significant interactions on needle and shoot growth of red
25 spruce. In both cases the inhibition caused by O3 and SAR at pH 5.1 was reversed by more
26 acidic rain at pH 3.1. However, there were unexplained inconsistencies in the trends since
27 the combination of intermediate O3 levels and low pH resulted in the greatest reductions in
28 dry matter. Percy et al. (1992), also working with red spruce, observed an unexplained
29 statistically significant interaction on the thickness of the needle epidermal cell cuticular
30 membrane: at intermediate O3 exposures, increased acidity led to reduced membrane
31 thickness while at lower or higher O3 levels led to thicker membranes.
December 1993 5-92 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-10. REFERENCES TO REPORTS OF INTERACTION OR NO
INTERACTION BETWEEN OZONE AND ACID RAIN OR ACID FOG
Species
Tree species:
CONIFERS:
Jeffrey pine
Loblolly pine
Ponderosa pine
Shortleaf pine
Slash pine
White pine
Douglas fir
Norway spruce
Red spruce
Sequoia
Totals
HARDWOODS:
Green ash
White ash
European beech
Paper birch
Sugar maple
Red oak
Yellow poplar
Totals
Crop species:
Interaction
No. References
0 —
1 41
0 —
1 52
2 9, 13
1 47
0 -
1 3, 4
2 31, 39
1 63
9
0 —
0 —
1 14
1 29
0 -
0
3 11, 12, 27
5
No.
1
13
2
1
0
3
1
9
5
0
35
1
1
1
0
2
2
1
8
No Interaction
References
62
1, 10, 16-21, 26, 32, 43, 47, 55
65,66
8
—
44, 46, 56
25
2, 5-7, 15, 24, 30, 36, 50
33, 34, 38, 40, 62
—
23
23
35
—
44,45
44,45
48
FORAGES AND FIELD CROPS:
Alfalfa
Sorghum
Soybean
Wheat
Totals
HORTICULTURAL CROPS:
Snap bean
Celery
Corn
Pepper
1 59
1 51
1 67
0 —
2
0 -
0 -
0 —
0 —
4
0
4
1
5
1
1
1
2
42, 53, 59, 64
—
28, 37, 53, 57
53
53
60
60
58,60
December 1993
5-93
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-10 (cont'd). REFERENCES TO REPORTS OF INTERACTION OR NO
INTERACTION BETWEEN OZONE AND ACID RAIN OR ACID FOG
Interaction No Interaction
Species No. References No. References
HORTICULTURAL CROPS (cont'd'):
Strawberry 0 — 2 60, 61
Tomato 0 — 2 53, 60
Avocado 1 22 0 —
Citrus 1 22 0 —
Totals 2 9
Others:
Ivy 0 - 1 30
Lichen (Lobaria) 0 — 1 54
Totals 0 2
TOTALS: 19 63
References:
1. Adams and O'Neill (1991). 2. Barnes and Brown (1990). 3. Barnes et al. (1990a). 4. Barnes et al.
(1990b). 5. Blank et al. (1990a). 6. Blank et al. (1990b). 7. Blaschke and Weiss (1990). 8. Boutton and
Flagler (1990). 9. Byres et al. (1992). 10. Carter et al. (1992). 11. Chappelka et al. (1985). 12. Chappelka
et al. (1988). 13. Dean and Johnson (1992). 14. Eamus and Murray (1991). 15. Ebel et al. (1990).
16. Edwards and Kelly (1992). 17. Edwards et al. (1990). 18. Edwards et al. (1991). 19. Edwards et al.
(1992). 20. Edwards et al. (1992). 21. Edwards et al. (1992). 22. Eissenstat et al. (1991). 23. Elliott et al.
(1987). 24. Fuhrer et al. (1990). 25. Gorissen et al. (1991). 26. Hanson et al. (1988). 27. Jensen and Patton
(1990). 28. Johnston and Shriner (1986). 29. Keane and Manning (1988). 30. Kerfourn and Garrec (1992).
31. Kohut et al. (1990). 32. Kress et al. (1988). 33. Laurence et al. (1989). 34. Lee et al. (1990).
35. Leonard! and Langebartels (1990). 36. Magel et al. (1990). 37. Norby et al. (1986). 38. Patton et al.
(1991). 39. Percy et al. (1992). 40. Pier et al. (1992). 41. Qiu et al. (1992). 42. Rebbeck and Brennan
(1984). 43. Reddy et al. (1991). 44. Reich and Amundson (1985). 45. Reich et al. (1986). 46. Reich et al.
(1987). 47. Reich et al. (1988). 48. Roberts (1990). 49. Sasek et al. (1991). 50. Senser (1990). 51. Shafer
(1988). 52. Shelburne et al. (1993). 53. Shriner and Johnson (1987). 54. Sigal and Johnston (1986).
55. Somerville et al. (1992). 56. Stroo et al. (1988). 57. Takemoto et al. (1987). 58. Takemoto et al.
(1988a). 59. Takemoto et al. (1988b). 60. Takemoto et al. (1988c). 61. Takemoto et al. (1989). 62. Taylor
et al. (1986). 63. Temple (1988). 64. Temple et al. (1987). 65. Temple et al. (1992). 66. Temple et al.
(1993). 67. Troiano et al. (1983).
1 Shelburne et al. (1993) reported that, in two growing seasons, needle biomass of
2 shortleaf pine was significantly reduced in tree seedlings receiving the highest O3 exposures
3 (2.5 x ambient) and SAR at pH 3.3. However, there were no effects at lower O3 exposure
4 levels or higher pHs.
5 A three-year study of slash pine revealed a significant interaction on stem volume
6 increment in each year (Dean and Johnson, 1992). This was attributed to a high rate of
December 1993 5-94 DRAFT-DO NOT QUOTE OR CITE
-------
1 increase observed with increasing acidity in trees exposed to an intermediate O3 level
2 (2 x ambient). In contrast, at higher or lower O3 exposures, acidity of the SAR applied had
3 little effect. Although another study with slash pine indicated a significant interaction on
4 photosynthetic rates, no information was provided about its nature (Byres et al., 1992).
5 The mineral status (K, Ca and manganese, Mn) of white pine showed antagonistic
6 interactions between O3 and SAR (Reich et al., 1988). Increased acidity nullified the
7 increase in foliar K and the decreases in root CA caused by O3, while increased O3 nullified
8 the increase in root Mn that resulted from increased acidity.
9 Temple (1988) reported a synergistic response of root growth of giant sequoia. Yellow
10 poplar showed no interactions hi one study (Table 5-10), but a greater than additive response
11 of root growth was observed by Chappelka et al. (1985). Chappelka et al. (1988b) found
12 that although neither O3 nor the pH of SAR caused any significant main effects on growth, at
13 intermediate O3 levels increased acidity caused significant decreases in stem and leaf
14 biomass. Jensen and Patton (1990), on the other hand, reported significant antagonistic
15 interactions on yellow poplar leaf and shoot growth. Based on estimates from growth models
16 derived from experimental data, increased acidity (pH 5.5 to pH 3.0) of SAR reduced the
17 decreases caused by O3 by almost 50%.
18 Adverse effects of O3 on the leaf area and shoot, leaf, and root biomass of paper birch
19 were reversed by increased acidity of SAR (Keane and Manning, 1988). Similarly, in both
20 avocado and lemon trees, Eissentstat et al. (1991) found that increased acidity offset the
21 negative effects of O3 on leaf growth.
22 Although there are four reports of no interactions on alfalfa, Takemoto et al. (1988b)
23 observed significant interactions on leaf drop. In charcoal-filtered air, leaf drop increased by
24 a factor of six as the pH of the fog treatment changed from 7.24 to the extremely acid pH
25 1.68, the lowest level recorded in the field in southern California. In unfiltered air, in
26 contrast, leaf drop only increased 20%.
27 Several studies with soybean revealed no significant interactions. However, Troiano
28 et al. (1983) reported a 42% reduction in seed yield between charcoal-filtered and unfiltered
29 air with SAR at pH 2.8, versus a 6% reduction at pH 4.0. Increased acidity thus multiplied
30 the effect O3, due largely to a stimulation of seed yield caused by increased acidity. Shafer
31 (1988) observed a stimulation of shoot growth of sorghum at pH 2.5 of SAR over growth at
December 1993 5.95 DRAFT-DO NOT QUOTE OR CITE
-------
1 pH 5.5 as a result of which greater growth occurred at low O3 exposure levels, although
2 there was no effect of acidity at the highest 03 level (0.3 ppm).
3 In summary, although the majority of studies have not demonstrated the existence of
4 interactions between O3 and SAR, where statistically significant interactions on growth or
5 physiology have been reported, the interactions were mostly antagonistic. The only
6 synergistic interactions reported are in two studies of yellow poplar and single studies of
7 sequoia and shortleaf and slash pines. In most cases where significant interactions were
8 noted, the authors have had difficulty in providing any mechanistic explanation. It appears
9 that, although the effects may have passed normally accepted tests of statistical significance,
10 they may nevertheless have been spurious findings. Overall, it appears that exposure to
11 acidic precipitation is unlikely to result in significant enhancement of the adverse effects of
12 O3 in most species. In the few cases of antagonistic interactions, the suggestion was made
13 that these may have reflected a beneficial fertilizer effect due to the nitrate and sulphate
14 present in the SAR applied.
15
16 5.4.6.6 Heavy Metals
17 Interactions of O3 with several heavy metal pollutants were reviewed in the previous
18 criteria document (U.S. Environmental Protection Agency, 1986). The limited data for
19 pollutants such as cadmium (Cd), nickel (Ni), and zinc (Zn) almost invariably showed that
20 they enhanced the adverse effects of O3, usually additively, but occasionally more than
21 additively. To the results with Cd, Ni, and Zn on garden cress, lettuce, pea, tomato, and
22 aspen, reviewed at that time, should be added similar findings with Zn on Pinto bean
23 (Mcllveen et al., 1975), with increased Zn resulting in significantly increased foliar injury
24 and decreased infection mycorrhizal establishment. However, in a study of the effects of Oj,
25 Ni, and copper (Cu) on tomato, Prokipcak and Ormrod (1986) found that as the levels of
26 both O3 and Ni increased, the interaction changed from additive to less than additive.
27 Complex interactions were observed when the treatments included both Ni and Cu.
28 No information appears to be available about possible interactions with lead. Although
29 qualitatively heavy metals appear to increase plant sensitivity to O3, the limited information
30 available precludes defining any quantitative relationships.
31
December 1993 5-96 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.4.6.7 Mixtures of Ozone with Two or More Pollutants
2 Pollutant-pollutant interactions are not limited to mixtures or sequences of two
3 pollutants. Several studies have been made of interactions of O3 with various combinations
4 of SO2, NO2 and acid rain. However, in some of these investigations, no treatment with
5 O3 was included in the experimental design and therefore no information was obtained on
6 effects on the response to O3. Some studies using only repeated daily exposures to high
7 levels (>0.3 ppm) of one or more pollutants are excluded from this review.
8 Adaros et al. (1991b), hi a field study of spring rape using open-top chambers, found
9 no significant interactions between O3 and NO2 (sequential exposures) and SO2 (continuous
10 exposures). In a two-year study on spring barley and spring wheat, some statistically
11 significant interactions were noted but they were scattered through the different growth
12 measurements, cultivars and years with no consistent pattern (Adaros et al., 1990a).
13 Additive effects with no interactions were observed in studies of shore juniper (Fravel et al.,
14 1984), radish (Reinert and Gray, 1981), and azalea (Sanders and Reinert, 1982). Yang et al.
15 (1982) reported a less than additive interaction on injury to white pine.
16 No significant three-way interactions were found in studies of soybean (Norby et al.,
17 1985), yellow poplar (Chappelka et al., 1985, 1988b), nor on other hardwood species (Davis
18 and Skelly, 1992; Jensen and Dochinger, 1989; Reich et al., 1985) exposed to O3, SO2 and
19 SAR.
20 No information was collected on interactions in the few published studies involving
21 03, SO2, NO2 and SAR.
22 The limited data make it difficult to draw any firm conclusions, but in general, the
23 consequences of such exposures appear to be largely dictated by the dominant individual
24 two-way interaction.
25
26 5.4.7 Interactions with Agricultural Chemicals
27 Agricultural chemicals are used for the control of insect pests, diseases, and weeds, and
28 for the control of growth in specialized situations such as the selective thinning of fruit on
29 orchard trees. The potential for some agricultural chemicals to modify plant response to
30 O3, first noted with certain fungicides on Pinto beans (Kendrick, 1954), led to numerous
31 field and laboratory studies. As noted in the previous criteria document (U.S. Environmental
December 1993 5.97 DRAFT-DO NOT QUOTE OR CITE
-------
1 Protection Agency, 1986), protection against O3 injury was found to be conferred by
2 applications of numerous commercial fungicides, herbicides and growth regulators.
3 The available information is derived from studies involving a number of different
4 commercial chemicals and species. No comprehensive and systematic studies have been
5 reported, but the weight of evidence indicates that certain fungicides are consistent in
6 providing protection. In particular, there have numerous reports of protection conferred by
7 applications of benomyl (benlate; methyl-l-[butylcarbamoyl]-2-benzimidazolecarbamate).
8 In addition to the studies reviewed in the previous criteria document (U.S. Environmental
9 Protection Agency, 1986), benomyl protection of grape (Musselman and Taschenberg, 1986)
10 and bean cultivars (Pell, 1976; Pellisier et al., 1972) has also been reported. It is of interest
11 to note that although several nematocides were found to increase sensitivity of tobacco and
12 Pinto bean to O3, applications of benomyl overcame this response and conferred resistance
13 (Miller et al., 1976). However, benomyl was found to increase the injury caused by PAN
14 (Pell and Gardner, 1979). It should also be noted that many of the effective fungicides are
15 carbamates and have been used as antioxidants in other applications such as rubber
16 formulations.
17 The need to distinguish between protective action against O3 injury and fungicidal
18 activity per se is shown by a study of fentin hydroxide (Du-Ter; tetraphenyltin hydroxide) on
19 potato (Holley et al., 1985). The fungicide reduced foliar injury in the field and also the
.20 colonization of injured leaf tissue by the early blight fungus, Alternaria soktni. However,
21 yield increases appeared to result from the reduction of disease rather than from diminished
22 O3 injury.
23 The triazoles are a family of compounds with both fungicidal and plant growth
24 regulating properties. Fletcher and Hofstra (1985) reported on the protective action of
25 triadimefon (l-[4-chlorophenoxy]-3,3-dimethyl-l-[lH--l,2,4-triazo-l-yl]-2-butanone),and
26 Musselman and Taschenberg (1986) found that triadimefon and the triazole, etaconazole
27 (l-[[2,4-dichlorophenyl]-4-ethyl-l,3-dioxolan-2-yl]methyl-lH-l,2,4-triazole), were as
28 effective as benomyl in protecting grape from oxidant injury; cultivar differences were noted,
29 with the fungicides being more effective on Concord than on Ives foliage. Seed treatment
30 with triazole S-3307 ([E]-l-[4-chlorophenoxy]-3,3-dunethyl-2-[l,2,4-triazol-l-yl]-
31 l-penten-3-ol) resulted in a 50% reduction in the size of wheat plants but provided complete
December 1993 5-98 DRAFT-DO NOT QUOTE OR CITE
-------
1 protection from an excessive exposure to 0.5 ppm O3 for 6 h that resulted in severe necrosis
2 on the leaves of untreated plants (Mackay et al., 1987).
3 A range of commercial plant growth regulating compounds was studied by Cathey and
4 Heggestad (1972). The plant growth retardants, CBBP (Phosfon-D;
5 2,4-dichloro-benzyltributyl phosphonium chloride) and SADH (Alar; succinic acid
6 2,2-dimethyl-hydrazide) and several of its analogs, were found to be more effective than
7 benomyl in reducing O3 injury on petunia.
8 Conflicting reports of the effects of herbicide-O3 interactions were reviewed in the
9 previous criteria document (U.S. Environmental Protection Agency, 1986). Recent studies
10 of metolachlor (2-chloro-N-[2-ethyl-6-methlphenyl]-N-[2-methoxy-1 -methylethyl] acetamide);
11 Mersie et al., 1989) and atrazine (2-chloro-4-ethylamino-6-isopropylamino-.s-triazine); Mersie
12 et al., 1990) revealed species-dependent effects: metolachlor sensitized com to O3 but
13 offered protection to bean and soybean. The effects of atrazine on corn were additive to
14 those induced by exposure to 0.2 ppm O3 for 6 h/day twice weekly for three weeks, but
15 antagonistic to exposures to 0.3 ppm. Mersie et al. (1990) also observed a protective action
16 of the commercial antioxidant, n-propyl gallate, on corn.
17 In spite of reports to the contrary (Teso et al., 1979), Rebbeck and Brennan (1984)
18 found that the insecticide, diazinon (O,O-diethyl-O-[2-isopropyl-4-methyl-6-pyrimidinyl]
19 phosphorothioate), did not protect alfalfa from O3 injury in a greenhouse study.
20 Our knowledge of the interactions of these different types of agricultural chemical with
21 O3 is still too fragmentary to be able to draw any general conclusions other than to note the
22 general efficacy of the carbamate fungicides. As noted in the previous criteria document
23 (U.S. Environmental Protection Agency, 1986), it is premature to recommend their use
24 specifically for protecting crops from the adverse effects of 63, rather than for their primary
25 purpose.
26
27 5.4.8 Factors Associated with Global Climate Change
28 This section focuses solely on the ways in which features of global climate change may
29 be expected to affect the impact of oxidants on vegetation. It is not intended to provide a
30 comprehensive review of the issues and components of climate change per se.
December 1993 5.99 DRAFT-DO NOT QUOTE OR CITE
-------
1 The magnitudes and causes of some of the changes in features of the global climate that
2 have beea observed or are predicted to occur are currently the subject of controversy.
3 However, there is clear evidence of increases in mean CO2 levels (Keeling et al., 1989)
4 which together with other anthropogenic emissions of radiatively active gases may contribute
5 to the upward trend in mean surface level temperatures observed over the past century
6 (Jones, 1989) and changes precipitation patterns throughout the world (Diaz et al., 1989).
7 In addition, depletion of the stratospheric O3 layer in the polar regions, caused by
8 halofluorocarbons, results in increased penetration of the atmosphere by solar ultraviolet-B
9 (UV-B) radiation (280 to 320 nm wavelengths). However, the intensity of UV-B radiation
10 reaching the earth's surface may be attenuated by O3-pollution in the lower troposphere
11 (Briihl and Crutzen, 1989). Differences in the degree of this attenuation probably contribute
12 to the discrepancies between recently observed trends in surface-level UV-B intensities
13 (Scotto et al., 1988; Blumenthaler and Ambach, 1990).
14 Independent of any effects of ambient temperature, CO2 level affects plant water
15 relations through effects on stomatal aperture and conductance, leading to effects on leaf and
16 canopy temperature and the uptake of gaseous pollutants. The effects of UV-B on numerous
17 growth processes have been reviewed by Tevini and Teramura (1989) and Runeckles and
18 Krupa (1993). Individual interactive effects of O3 and several of these features of global
19 climate change have been reviewed in the previous sections. However, it is important to
20 recognize that, because of the interactions among the different components of climate change
21 themselves, a holistic approach is essential, which includes their potential for modifying plant
22 response to oxidants. Overall reviews of the interactions involving the factors of climate
23 change and O3 have been presented by Krupa and Kickert (1989) and Ashmore and Bell
24 (1991).
25 The effect of increased CO2 in stimulating photosynthetic rates may also lead to
26 increased leaf area, biomass and yield (Allen, 1990). Increased CO2 also leads to stomatal
27 closure. However, with regard to water use, the result of decreased stomatal conductance in
28 reducing transpiration is partly offset by the increase in leaf and canopy temperature resulting
29 from reduced evaporative cooling, and the increase in leaf area. The net result is that
30 increased CO2 may lead to only slight increases in water use efficiency that are attributable
31 more to increased photosynthetic activity than to reduced transpiration (Allen, 1990). On the
December 1993 5-100 DRAFT-DO NOT QUOTE OR CITE
-------
1 other hand, since the primary route of entry into the leaf of a gaseous pollutant such as O3 is
2 through the stomata, increased CO2 levels would be expected to decrease the impact of O3 by
3 reducing uptake as a consequence of reduced stomatal conductance.
4 Allen (1990) provides a simulation of the effect of doubling the average ambient CO2
5 level from 340 to 680 ppm on soybean yield, based on the Weibull response model to 03 and
6 SO2 of Heagle et al. (1983), and the model of stomatal conductance developed for soybean
7 by Rogers et al. (1983):
8
9 gg = 0.0485 - 7.00 x 10"5[CO2] + 3.40 x 10"8[CO2]2,
10
11 where gs is stomatal conductance (m/s), and [COJ is CO2 concentration (ppm). According
12 to this model, a doubling of the CO2 level would reduce gs by a factor of 0.69, effectively
13 reducing the O3 and SO2 concentrations to 0.038 and 0.018 ppm respectively. At the current
14 340 ppm CO2 level, the Weibull model predicts a yield of 340.5 g/m of row. Reduced
15 pollutant entry at 380 ppm CO2 gives a predicted yield of 390.6 g/m of row, an increase of
16 14.7%. This is a conservative estimate since it ignores the direct effect of the increased CO2
17 level on soybean growth.
18 Although the calculation makes numerous assumptions, it is qualitatively supported by
19 evidence from the few studies published to date on CO2/O3 interactions. Barnes and
20 Pfirrmann (1992) reported that an increased CO2 level of 765 ppm countered the adverse
21 effects of O3 on photosynthesis, shoot growth rate, leaf area, and water use efficiency of
22 radish. Protection against the adverse effects of O3 on soybean by elevated CO2 was also
23 reported by Kramer et al. (1991). The yield loss due to O3 at ambient CO2 was 11.9%,
24 whereas in the presence of ambient+150 ppm CO2, the loss was only 6.7%.
25 Although these studies support the prediction of Allen (1990), they were conducted in
26 growth chambers (Brown and Pfirrmann, 1991) or open-top field chambers (Kramer et al.,
27 1991; Mulchi et al., 1992), as were the studies on which Allen's model was based. Hence
28 the plants would not have been subjected to the environmental conditions typical of the open
29 field, particularly with respect to wind speed and its effects on transpiration and temperature.
30 Nevertheless they support the view that increased CO2 levels will reduce adverse effects of
31 O3 on crops.
December 1993 5-101 DRAFT-DO NOT QUOTE OR CITE
-------
1 It is unclear as to whether such CO2-induced reductions of the impact of O3 also apply
2 to the long term growth of trees, and it is equally unclear as to how increased CO^ will
3 affect the impact of O3 on ecosystems. These uncertainties arise because of the numerous
4 compensatory feedback mechanisms that play important roles in both long-term perennial
5 growth and in the behavior of ecosystems. Such feedbacks include changing demands for
6 nutrients, increased leaf area and potential water loss, and changes in litter quality and
7 quantity. For example, in terms of the effects of increased CO^ alone, long-term studies of
8 several species suggest that although photosynthesis may be demonstrably stimulated, there
9 may be little or no net response at the ecosystem level (Bazzaz, 1990).
10 The consequences of global warming as a feature of climate change are difficult to
11 assess since, as discussed in Section 5.4.4, the information on the effects of temperature on
12 O3-response is conflicting. However, as Ashmore and Bell (1991) point out, concerns over
13 the effects of O3 on sensitivity to freezing temperatures will become increasing unimportant
14 as warming occurs.
15 Various models of climate change scenarios have indicated that changed precipitation
16 patterns will lead to increased drought in some mid-latitude regions of the world. The bulk
17 of the evidence reviewed in Section 5.4.4 suggests that this would reduce the impact of 03.
18 However, because of the major direct impact of drought per se, such protection would be of
19 little practical significance.
20 Greater certainty surrounds the likelihood that global warming will increase the
21 incidence and severity of losses caused by pests and diseases. Concurrent increases may also
22 favor the competitiveness of many weed species. At present, it is not possible to quantify
23 such changes or to determine how they would influence the interactions discussed in
24 Section 5.4.3.
25 With regard to possible interactions of O3 and UV-B, Runeckles and Krupa (1993)
26 point out that, because of the episodic nature of O3 pollution, including its typical diurnal
27 pattern, surface level exposures to UV-B will also be episodic. They have described various
28 possible O3/UV-B scenarios that need to be considered. With low surface O3 levels and
29 increased UV-B irradiation due to stratospheric O3 depletion, effects of UV-B will
30 predominate. On the other hand, elevated surface O3 levels will cause increased attenuation
31 of UV-B resulting in reduced surface intensities. With no stratospheric 03 depletion, this
December 1993 5-102 DRAFT-DO NOT QUOTE OR CITE
-------
1 condition implies that surface effects of O3 will predominate over the effects of the effects of
2 UV-B; with stratospheric O3 depletion, the resulting surface level irradiation will be
3 dependent upon the concentration and thickness of the surface O3 layer and both O3 and
4 UV-B effects may occur.
5 To date there have been no experiments conducted specifically to simulate these
6 different scenarios. However, Booker et al. (1992) exposed soybean in field open-top
7 chambers within which lamps were suspended to provide increased intensities of UV-B. The
8 O3 treatments were ambient and 1.5 x ambient. No significant O3/UV-B interactions were
9 noted; the effects on growth were solely attributable to the O3 exposure. However, increased
10 UV-B irradiation resulted hi increases in the foliar content of UV-absorbing constituents.
11 In contrast, Miller and Pursley (1990) reported that a preliminary experiment revealed a less
12 than additive interaction of O3 and UV-B on soybean growth.
13 It is clear overall that the effects of O3 on vegetation will be modified to some degree
14 by various components of the complex mix of factors that constitute climate change.
15 Considerably more research will need to be undertaken before quantitative assessments of the
16 magnitudes of the changes will be possible.
17
18 5.4.9 Summary
19 Since the previous criteria document (U.S. Environmental Protection Agency, 1986),
20 additional studies have been published on a wide range of biological, physical, and chemical
21 factors in the environment that interact with plant response to O3.
22 Biological components of the environment of individual plants include pests, pathogens
23 and plants of the same or other species in competition. With regard to insect pests, although
24 only a very limited number of plant-insect systems has been studied, there is a general trend
25 in the observations that suggests that some pests have a preference for and grow better when
26 feeding on plants that have been impacted by O3. Unfortunately, because we have no
27 knowledge of how the vast majority plant-insect systems will be affected by O^, it is not
28 possible to offer any quantitative overall assessment of the consequences of such interactions
29 on the growth of crops and natural vegetation. At best, we may conclude that there is a
30 reasonable likelihood that some insect pest problems will increase as a result of increased
31 ambient O3 levels, but there is no evidence to suggest that O3 may trigger pest outbreaks.
December 1993 5-103 DRAFT-DO NOT QUOTE OR CITE
-------
1 Plant-pathogen systems are also affected by O3 but here too, the available evidence is
2 far from representative of the wide spectrum of plant diseases. Nevertheless, the suggestion
3 of Dowding (1988) that diseases caused by obligate pathogens tend to be diminished by
4 O3 while those caused by facultative pathogens tend to be favored, is generally supported by
5 the limited evidence available. In terms of its broader implications, this suggests that
6 continued exposure to O3 may lead to a change in the overall pattern of the incidence and
7 severity of specific plant diseases affecting crops and forest trees. However, it is not
8 possible with the limited evidence currently available to predict whether the net consequences
9 would be more harmful or less.
10 A major level of uncertainty concerns the effects of O3 at the population and
11 community levels within natural ecosystems. Very few studies have been conducted on
12 multi-species systems, and Woodward (1992) has pointed out the hazards of attempting to
13 extrapolate from responses of the individual plant to responses of a population of such plants.
14 This is borne out by the observations of Evans and Ashmore (1992) who showed that the
15 behavior to O3-exposure of a species growing in mixture with other species is not predictable
16 from its behavior when grown in isolation. This has serious implications with regard to
17 complex natural ecosystems, and identifies a serious gap in our knowledge of the effects of
18 O3 that can only be filled by a substantial research effort.
19 With regard to the physical environment, the combination of light, temperature and
20 water availability largely determines the success of plant growth because of the influence of
21 these factors on the processes of photosynthesis, respiration and transpiration. For
22 agricultural crops, perhaps the most important of these potential interactions with
23 O3 concerns water availability and use. There is consistent evidence that drought conditions
24 tend to reduce the direct adverse effects of O3 on growth and yield. Conversely, the ready
25 availability of soil water tends to increase the susceptibility of plants to O3 injury. However,
26 a lack of water should not be viewed as a potentially protective condition, because of the
27 adverse effects of drought per se. The combination of drought conditions and exposure to
28 O3 is likely to result hi adverse effects on growth and yield that are largely the result of lack
29 of water. However, with perennial trees, there is evidence that prolonged exposures to
30 O3 may lead to greater water use efficiency which would enable such trees to be better able
31 to survive drought conditions.
December 1993 5-104 DRAFT-DO NOT QUOTE OR CITE
-------
1 In contrast with crop species, with tree species the relative roles of light, temperature
2 and water are shifted somewhat because of the differences in plant form. In particular, the
3 photosynthetic function of the leaves is carried out by a much smaller proportion of the
4 plant's biomass. Conversely, a larger demand is placed on temperature-dependent
5 respiratory processes to maintain and support the tissues of the stem and root systems.
6 In addition, in temperate regions, the perennial habit brings with it the requirement for
7 storage of carbohydrates and other reserves, in order to permit survival during the winter
8 season and to facilitate renewed spring growth. Hence, with tree species it becomes
9 important to distinguish between the immediate effects of exposure to 03 and the longer-term
10 consequences of these effects.
11 Of particular importance in northern latitudes and at higher elevations is the
12 demonstrated role of O3 in adversely affecting cold hardiness, by reducing carbohydrate
13 storage. Independent of effects on winter hardiness, there is also evidence to indicate that
14 adverse effects on storage may also be a component of changes in growth occurring in
15 subsequent seasons (Hogsett et al., 1989; Anderson et al., 1991; Sasek et al., 1991).
16 However, it is not yet possible to assemble these observations into a general quantitative
17 model.
18 The plant's environment also contains numerous chemical components, ranging from
19 soil nutrients and other air pollutants to agricultural chemicals used for pest, disease and
20 weed control. With regards to plant nutrients and their influence on plant response to
21 O3, the available evidence is highly fragmentary and frequently contradictory, and hence
22 does not permit the drawing of any general conclusions. A large number of studies have
23 been conducted on the effects of O3 in conjunction with other gaseous air pollutants such as
24 SO2 and NO2, although the information obtained in several of the studies is of no more than
25 academic interest because of the unrealistic exposure conditions used. Although there is
26 clear evidence to show that O3 and SO2 may act synergistically in increasing foliar injury in
27 some species, the available evidence indicates that this type of response is not universal.
28 Several empirical models of the O3-SO2 interaction have been developed, but they have little
29 in common and are highly specific to the crop and exposure conditions used. Furthermore,
30 the frequently observed lack of interaction implies that in many cases the impact of 03 is
December 1993 5_105 DRAFT-DO NOT QUOTE OR CITE
-------
1 probably best assessed on its own. The same is true of the situation with regard to
2 combinations of O3 and acid rain or acid fog, and O3 and NO2.
3 Numerous agricultural chemicals have been found to influence the responses of plants
4 to O3. In particular, several fungicides have been shown to confer protection against visible
5 injury, although none has been adopted for commercial application for this purpose. On the
6 other hand, the experimental chemical, ethylenediurea, has been found consistently to
7 provide protection of a wide range of species, both in the laboratory and in the field.
8 Since increased tropospheric O3 is a component of global climate change, results from
9 studies on the interactions of O3 with increased levels of CO2 and UV-B radiation are
10 beginning to appear. Initial work with CO2 suggests that increased CO2 levels may
11 ameliorate the effects of O3. However, it is too soon to be able to generalize on the outcome
12 of this interaction. At the present time, no investigations of the compound interactions
13 involving O3, CO2, UV-B, increased temperature and changed soil moisture status have been
14 reported.
15 In conclusion, in spite of the amount of work carried out to date on the interactions of
16 O3 with environmental factors, we are left with a very fragmented understanding from which
17 to draw conclusions. This is probably inevitable in view of the vast scope of the possible
18 interactions between O3 and all the other environmental variables. It is also a result of the
19 fact that most of the published work consists of studies resulting from personal interests of
20 the investigators, rather than from coordinated programs of research that focus on systematic
21 investigations.
22
23
24 5.5 EFFECTS-BASED AIR QUALITY EXPOSURE INDICES
25 5.5.1 Introduction
26 5.5.1.1 Biological Support for Identifying Relevant Exposure Indices
27 The effects of O3 on individual plants and factors that modify plant response to O3 are
28 complex and vary with species, environmental conditions, and soil and nutrient conditions.
29 Because of the complex effect of Oj and its interactions with physical and genetic factors that
30 influence response, the development of exposure indices to characterize plant exposure and to
31 quantify the relationship between O3 exposure and ensuing plant response has been and
December 1993 5-106 DRAFT-DO NOT QUOTE OR CITE
-------
1 continues to be a major problem. At best, experimental evidence of O3's effect on biomass
2 production can refine our knowledge of those factors of O3 exposure that affect our ability to
3 predict plant response using exposure indices. The impacts of measured O3 concentrations
4 on plant response are discussed and evaluated to determine the key factors of exposure that
5 account for the variations in plant response and, if possible, to develop measures of pollutant
6 exposure which relate well with plant response.
7 Considerable evidence of the primary mode of action of 63 on plants (e.g., injury to
8 proteins and membranes, reduction in photosynthesis, changes in allocation of carbohydrate,
9 and early senescence), which eventually impacts biomass production, identifies O3 uptake as
10 the correct characterization of plant exposure (Section 5.3). Ozone uptake is controlled by
11 canopy conductance, stomatal conductance, and ambient O3 outside the leaf (see Figure 5-3).
12 Any factor that will affect stomatal conductance (e.g., light, temperature, humidity, soil and
13 atmospheric chemistry and nutrients, time of day, phenology, and biological agents) will
14 affect O3 uptake and, consequently plant response (i.e., yield or biomass). Biochemical
15 mechanisms describe the mode of action of O3 on plants as the culmination of a series of
16 physical, biochemical, and physiological events leading to alterations in plant metabolism.
17 Ozone-induced injury is cumulative, resulting in net reductions in photosynthesis, changes in
18 allocation of carbohydrate, and early senescence, which ultimately lead to reductions in
19 biomass production. In most cases, increasing the duration of exposure increases the effect
20 of O3 on plant response. Peak concentrations occurring during daylight when stomatal
21 conductance is high have more influence in determining the impact of O3 on plant response
22 than lower concentrations or night concentrations because of a greater likelihood of
23 intracellular impairment.
24 From a lexicological perspective, duration and peak concentrations above some level
25 have value in determining plant response but interact with other factors such as respite time,
26 temporal variation, phenology, canopy structure, physiological processes, environmental
27 conditions, and soil and nutrient conditions in different fashions, depending upon species.
28 Effects occur on vegetation when the amount of pollutant absorbed exceeds the ability of the
29 plant to detoxify O3 or repair the initial impact (Tingey and Taylor, 1982).
30 Although O3 uptake integrates the above factors with atmospheric conditions and relates
31 well with plant response, it is difficult to measure. Several empirical models to predict
December 1993 5407 DRAFT-DO NOT QUOTE OR CITE
-------
1 stomatal conductance have been developed for particular species (Losch and Tenhunen, 1981)
2 but have not been used to estimate O3 uptake or to develop exposure indices.
3
4 5.5.1.2 Historical Perspective on Developing Exposure Indices
5 For almost seventy years, air pollution specialists have explored alternative
6 mathematical approaches for summarizing ambient air quality information in biologically
7 meaningful forms that can serve as surrogates for dose for vegetation effects purposes. Some
8 of the indices introduced have attempted to incorporate some of the factors (directly or
9 indirectly) described above. Recognizing the importance of duration and peak concentrations
10 in conjunction with stomatal conductance, the optimum exposure index can be written as:
11
12 Index = Ei=1na Wj x f(Cj) (5-1)
13
14 where Q is the hourly mean concentration, f(C{) is some function of Q, and Wj is some
15 weighting scheme that relates ambient condition and internal O3 flux. The optimal weights
16 are difficult to develop because of the complex relationship among exposure, environmental
17 condition, and species.
18 Equation (5-1) represents a taxonomy of exposure indices that have been proposed as
19 surrogates of dose in the literature. The exposure indices differ in the ways in which the
20 values are assigned to Wj. Based on the weighting function, the exposure indices can be
21 arranged into the following categories (description from Lee et al., 1989):
22
23 • One Event: Wj=0 for all CA except for the few concentrations where Wj=l.
24 Examples of such indices are the second highest daily maximum 1-h
25 concentration (2ndHDM), the maximum of 7-h (P7) and 1-h (PI) maximum
26 daily averages, and the 90th or higher percentiles of hourly distribution;
27
28 • Mean: w—0 for all CA outside the period of interest, P, and wi=vi/'Ei=ini
29 Vj for all Cj inside the period P where VA is a function of CA or some
30 environmental variable. Examples are the seasonal mean of 7-h daily
31 means (M7) (Heagle et al., 1979); the effective mean, denoted me, where
32 mev is the index in Equation (5-1) with f(Cj)= Ci"1/v and w—1 for some
33 parameter v (Larsen and Heck, 1984); the solar-weighted mean where vt is
34 the hourly solar radiation value (Rawlings et al., 1988);
35
December 1993 5-108 DRAFT-DO NOT QUOTE OR CITE
-------
1 • Cumulative: w—1 for all C{. An example is the seasonal sum of hourly
2 concentrations (i.e., total exposure, denoted as SUMOO);
3
4 • Concentration Weighting: w^gCC;) where g() is a monotonically non-
5 decreasing function. Examples are the seasonal sum of hourly
6 concentrations at or above a threshold level such as 0.06 ppm (SUM06) or
7 0.08 ppm (SUM08); the seasonal sum of the difference between an hourly
8 concentration above a threshold level less the threshold value, such as
9 0.08 ppm (AOT08); the total impact with w—C/"1"17^ for some v (Larsen
10 et al., 1983); the index with the allometric function, g(Ci)=Cia, a>0; the
11 index with sigmoidal weighting function, g(Ci) = l/[l+M x exp(-AxCi)],
12 where M=4,403 and A=126, denoted as W126 by Lefohn et al. (1988a),
13 and M=500 and A=100, denoted SIGMOID by Lee et al. (1989); total
14 hours with concentrations at or above a threshold level, such as 0.08 ppm
15 (HRS08), g(Ci)=0 for Cj<0.08 ppm and w^l/Q for Cj>0.08 ppm;
16
17 • Multicomponent: wi=g(Ci, i). Examples are indices that incorporate
18 several characteristics of exposure and crop development stage, including
19 the phenologically weighted cumulative impact indices (Lee et al., 1987).
20
21
22 Oshima (1975) and Oshima et al. (1976) proposed an exposure index, where the
23 difference between the value above 0.10 ppm and 0.10 was summed. This is referred to as
24 the AOT10 exposure index with f(C£)=CrO. 10 and w£=0 for C{ < 0.10 ppm and wt= 1 for
25 Cj^O.10 ppm in Equation (5-1). Alternatively, Lefohn and Benedict (1982) introduced an
26 exposure index based on the hypothesis that if the higher O3 concentrations had greater value
27 in predicting adverse effects on agricultural crops than the lower values, then the higher
28 hourly mean concentrations should be given more weight than the lower values. This index
29 summed all hourly concentrations equal to and above a 0.10 ppm threshold level. This index
30 is referred to as the SUM10 exposure index with f(Cj)=Cj and W;=0 for C^O.10 ppm and
31 wj=l for Cj^O.10 ppm.
32 A 6-h long-term seasonal mean O3 exposure index was used by Heagle et al. (1974).
33 Also, Heagle et al. (1979) reported the use of a 7-h experimental period mean. The 7-h
34 (0900 to 1559 h) mean, calculated over an experimental period, was adopted as the statistic
35 of choice by the U.S. EPA's NCLAN program (Heck et al., 1982). The 7-h daily daylight
36 period was selected by NCLAN because the index was believed to correspond to the period
37 of greatest plant susceptibility to O3 pollution. In addition, the 7-h period of each day
38 (0900 to 1559 h) was assumed to correspond to the time that the highest hourly
December 1993 5-109 DRAFT-DO NOT QUOTE OR CITE
-------
1 O3 concentrations would occur. However, not all monitoring sites in the United States
2 experience their highest O3 exposures within the 0900 to 1559 h 7-h time period (Lefohn and
3 Jones, 1986; Lefohn and Irving, 1988; Logan, 1989). Toward the end of the program,
4 NCLAN redesigned its experimental protocol and applied proportional additions of O3 to its
5 crops for 12-h periods. The expanded 12-h window reflected NCLAN's desire to capture
6 more of the daily O3 exposure. In the published literature, the majority of NCLAN's
7 experiments were summarized using the 7-h experimental-period average.
8 As additional evidence began to mount that higher concentrations of O3 should be given
9 more weight than lower concentrations (summarized in U.S. Environmental Protection
10 Agency, 1986), concerns about the use of a long-term average to summarize exposures of
11 O3 began appearing in the literature (Lefohn and Benedict, 1982; Tingey, 1984; Lefohn,
12 1984; Lefohn and Tingey, 1985; Smith et al., 1987). Specific concerns were focused on the
13 fact that the use of a long-term average failed to consider the impact of peak concentrations.
14 The 7-h seasonal mean contained all hourly concentrations between 0900 to 1559 h; this
15 long-term average treated all concentrations within the fixed window in a similar manner.
16 A large number of hourly distributions within the 0900 to 1559 h window could be used to
17 generate the same 7-h seasonal mean, ranging from those containing many peaks to those
18 containing none. Larsen and Heck (1984) pointed out that it was possible for two air
19 sampling sites with the same daytime arithmetic mean O3 concentration to experience
20 different estimated crop reductions.
21 In the late 1980s, the focus of attention turned from the use of long-term seasonal
22 means to cumulative indices (i.e., exposure indices that sum the products of concentrations
23 multiplied by time over an exposure period). As indicated previously, the cumulative index
24 parameters proposed by Oshima (1975) and Lefohn and Benedict (1982) were similar. Both
25 parameters gave equal weight to the higher hourly concentrations, but ignored the
26 concentrations below a subjectively defined minimum threshold (e.g., 0.10 ppm). Besides
27 the cumulative indices proposed by Oshima (1975), Oshima et al. (1976), and Lefohn and
28 Benedict (1982), other cumulative indices were suggested, including (1) the number of
29 occurrences of daily maximum hourly averaged concentrations greater than a threshold level
30 (Ashmore, 1984) and (2) the use of exponential functions (Nouchi and Aoki, 1979; Larsen
31 and Heck, 1984) to assign unequal weighting to O3 concentrations.
December 1993 5-110 DRAFT-DO NOT QUOTE OR CITE
-------
1 A possible disadvantage of applying an integrated exposure index, as defined by
2 Oshima (1975) or Lefohn and Benedict (1982), is that the use of an artificial threshold
3 concentration as a cutoff point eliminates any possible contribution of the lower
4 concentrations to vegetation effects. Although this disadvantage may not be important when
5 considering O3 exposures that occur in the California South Coast Air Basin, where repeated
6 high concentrations are experienced from day-to-day and there are relatively short periods
7 between episodes, it is important when assessing the typical exposures experienced in other
8 parts of the United States.
9 Recognizing the disadvantage, Lefohn and Runeckles (1987) suggested a modification to
10 the Lefohn and Benedict (1982) exposure index by weighting individual hourly mean
11 concentrations of O3 and summing over time. Lefohn and Runeckles (1987) proposed a
12 sigmoidal weighting function that was used in developing a cumulative integrated exposure
13 index. The index included the lower, less biologically effective concentrations in the
14 integrated exposure summation.
15 None of the exposure indices mentioned above fully characterize the potential for plant
16 uptake of O3 because the indices, being measures of ambient condition, ignore the biological
17 processes controlling the transfer of O3 from the atmosphere through the leaf and into the
18 leaf interior (U.S. Environmental Protection Agency, 1986, 1992). Early studies with beans
19 and tobacco, reviewed in the previous criteria document (U.S. Environmental Protection
20 Agency, 1986), showed that short-term higher-peak exposures induced more injury than
21 longer-term lower-peak exposures of the same total exposure, indicating that concentration
22 has more value than exposure duration in eliciting a response. Other studies with soybean,
23 tobacco, and bean, conducted prior to 1983 and described in U.S. EPA (1986), showed that
24 the foliar injury response to subsequent peak exposures varies with temporal pattern.
25 Predisposition to low levels of O3 for a few days increases plant sensitivity to subsequent
26 peaks (Johnston and Heagle, 1982; Heagle and Heck, 1974; and Runeckles and Rosen,
27 1977). Tobacco plants exposed to two consecutive days of peak exposures showed greater
28 injury on the first day (Mukammal, 1965). Plants exposed to a series of successive short
29 exposures suffered more injury than did those plants that received a continuous uniform
30 exposure, with all plants receiving equal total exposure (Stan and Schicker, 1982).
December 1993 5411 DRAFT-DO NOT QUOTE OR CITE
-------
1 When yield or growth are considered, "not only are concentration and time important
2 but the dynamics of the O3 exposure are also important" (U.S. Environmental Protection
3 Agency, 1986). Musselman et al. (1983) and Hogsett et al. (1985a) were among the first to
4 demonstrate that plants exposed to variable concentrations showed greater effect on plant
5 growth than those exposed to a fixed or daily peak concentration of equal total exposure but
6 lower peak concentrations. Musselman et al. (1986), in a subsequent experiment, exposed
7 kidney bean plants to either a simulated ambient or a uniform concentration that had equal
8 total exposure and peak concentration (at two levels of 0.30 and 0.40 ppm) and found that
9 the effects of the two distributions did not differ significantly. Consequently, when peak
10 concentrations and total exposures are equal, the diurnal distribution of concentrations
11 appears to be unimportant.
12 More recent studies with bean (Kohut et al., 1988), soybean (Heagle et al., 1986), and
13 tobacco (Heagle et al., 1987) (reviewed in U.S. Environmental Protection Agency, 1992)
14 showed conflicting evidence of no significant differences in response to different exposure
15 patterns of equal total exposure but varying peak concentrations. The value of peak
16 concentrations in influencing response is inconclusive in these studies because of low
17 statistical power. For the study with beans, plants exposed to peak exposures showed
18 significant impairment in the early harvests, but at the final harvest, O3 effects on growth
19 and yield were not statistically significant. For the NCLAN studies with soybean and
20 tobacco, differences in yield between the constant and proportional 7-h O3 addition exposures
21 were not significant, even though the proportional-addition treatments had greater peak
22 concentrations. In reanalysis of the soybean and tobacco studies, Rawlings et al. (1988)
23 stated that the differences between the constant and proportional O3 additions were relatively
24 small, thus limiting the power of the comparison test. However, 12-h exposures caused
25 greater effects than 7-h exposures but the decrease in yield loss was not directly proportional
26 to the increased length of exposure (Rawlings et al., 1988).
27 Considerable research since the publication of the previous criteria document (U.S.
28 Environmental Protection Agency, 1986) has been directed at developing measures of
29 exposure that were consistent with then-current knowledge of the mode of action of 63 on
30 plants and on factors including concentration, duration, and temporal dynamics of exposure
31 influencing response. A number of retrospective studies of existing data to evaluate and
December 1993 5-112 DRAFT-DO NOT QUOTE OR CITE
-------
1 compare exposure indices based on statistical fit have been summarized in the literature
2 between 1986 and 1988 and reviewed by the U.S. EPA (1992) (Rawlings et al., 1988;
3 Adomait et al., 1987; Cure et al., 1986; McCool et al., 1986, 1987; Smith et al., 1987; Lee
4 et al., 1987, 1988; Lefohn et al., 1988a; Tingey et al., 1989; Musselman et al., 1988).
5 These studies support the conclusion that: (1) O3 effects are cumulative; (2) peak
6 concentrations are more important than lower concentrations in eliciting a response; and
7 (3) plant sensitivity to O3 varies with time of day and crop development stage. Exposure
8 indices that cumulate the exposure and preferentially weight the peaks yield better statistical
9 fits to response than the mean and peak indices.
10 Because the mean exposure index treats all concentrations equally and does not
11 specifically include an exposure duration component, the use of a mean exposure index for
12 characterizing plant exposures appears to be inappropriate for relating exposure with
13 vegetation effects (U.S. Environmental Protection Agency, 1992). In particular, the
14 weighting of the hourly O3 concentrations of the mean is inconsistent with the weighting
15 function of plant exposure to O3 in Equation (5-1), which attempts to relate 03 flux with
16 ambient condition. The total exposure index includes an exposure duration component but
17 does not adequately relate pollutant exposure with plant response because the index weights
18 all concentrations equally and focuses on the lower concentrations, whose impact on
19 vegetation is minimal.
20 Evidence supporting the use of peak-weighted, cumulative indices in relating
21 O3 exposure and plant response is growing. However, it is unlikely that the empirical
22 modeling of plant response will determine the optimal weighting function of hourly
23 O3 concentrations for use in characterizing plant exposure, which vary with environmental
24 factors and species. The development and comparison of exposure indices based on
25 statistical fits is difficult because only a limited number of experiments have been specifically
26 designed to test and evaluate the various exposure indices.
27 While much research has been conducted on O3 effects on crops and trees between
28 1988 and the current period, our overall understanding of the mode of action of C^ on plants
29 and factors that modify plant response remains unchanged since the previous criteria
30 document (U.S. Environmental Protection Agency, 1986) and its supplement (U.S.
31 Environmental Protection Agency, 1992). Additional studies further support the value of
December 1993 5-113 DRAFT-DO NOT QUOTE OR CITE
-------
1 concentration, duration, and temporal pattern of exposure in describing plant exposure and its
2 relation to plant response. Studies that applied two or more different exposure patterns of
3 equal exposure but possibly different peak concentrations are reviewed in Section 5.5.2.2 to
4 substantiate the value of exposure structure in influencing the magnitude of plant response.
5 Recent papers that report results from replicate studies over time and/or space are
6 summarized in Section 5.5.2.3 to test the value of duration and its relation with plant
7 response. In addition, a few recent studies that provide additional insight to those factors
8 that modify plant response are reviewed in Section 5.5.2.4.
9
10 5.5.2 Developing Exposure Indices
11 5.5.2.1 Experimental Design and Statistical Analysis
12 Controlled and field exposure-response studies, where extraneous factors influencing
13 response are controlled or monitored, allow one to study concentration, duration, respite
14 time, temporal fluctuations at general and specific stages of crop development in influencing
15 response. These studies provide insight on the efficacy of exposure indices in explaining
16 variation of response. A small number of experiments have been designed specifically to
17 study the components of exposure and have applied two or more different patterns of
18 exposure which measure the same SUMOO values. These designs provide the best evidence
19 to determine whether plants respond differentially to temporal variations in
20 O3 concentrations. However, they may have limited application in developing a statistical
21 relationship between O3 exposure and plant response. Other design considerations, including
22 the number, kind, and levels of O3 exposure, the patterns of randomization, the number of
23 replicates used in the experiment, and experimental protocol, determine the (1) strength of
24 the statistical analysis that is applied to the treatment mean comparison tests and (2) range of
25 ambient and environmental conditions over which generalizations may be made. These
26 designs have been successfully used to test the value of components of exposure, particularly
27 concentration, in influencing response (Musselman et al., 1983, 1986; Hogsett et al., 1985a).
28 Different approaches that include either a mean separation procedure and/or a regression
29 procedure have been used to identify those important components of exposure that influence
30 response.
December 1993 5-114 DRAFT-DO NOT QUOTE OR CITE
-------
1 To identify the importance of exposure in contributing to variation of plant response,
2 the majority of pollutant effects studies use regression-based designs that apply a single
3 pattern of exposure at varying concentration levels. However, using these designs, the
4 application of the results is limited; the order of plant response (i.e., plant yield) with respect
5 to exposure is unchanged with different measures of exposure. The relative position and
6 spacing between exposure levels is a function of how the exposure index weights the hourly
7 03 concentrations and governs the statistical fit to response. The regression approach has
8 been used to compare and evaluate various exposure indices but the ability to discriminate
9 among indices is low for these studies. By their nature, those studies that have used
10 regression-based designs that utilize data from single patterns of exposure cannot distinguish
11 between mean exposure indices and sums constructed from means (i.e., mean x duration)
12 and, consequently cannot be used to test the value of duration in explaining the variation of
13 response.
14 Evidence to substantiate the value of duration in explaining the experimental variation
15 of plant response may be obtained when combining data from replicate studies of the same
16 species and cultivar over time and/or space. Pooling of data from replicate studies of the
17 same species to evaluate duration effects and/or to compare various exposure indices assumes
18 that the primary cause of biological response is pollutant exposure. This assumption may or
19 may not be valid, particularly when plants from replicate studies are grown under varying
20 environmental, edaphic, and/or agronomic conditions that tend to mask the treatment effects
21 during the growth of the plant (Section 5.3). Hence, it is more difficult to substantiate the
22 importance of exposure-dynamic factors from retrospective analyses of combined data from
23 replicate studies of the same species than from experiments designed specifically to address
24 the components of exposure. The comparison of environmental conditions, as well as the
25 yields of plants exposed to charcoal-filtered air over replicate studies, is a simple check of
26 interaction but does not ensure that O3 effects on response can be isolated. In addition, when
27 the main effect of O3 is insignificant, the data may be limited for determining the value of
28 duration or other components of exposure in predicting response. Nonetheless, if an air
29 pollutant is the primary source of variability in plant response, the relationship between
30 exposure and response should be consistent when data sets for the same crop are combined
31 over several years and/or locations.
December 1993 5_H5 DRAFT-DO NOT QUOTE OR CITE
-------
1 Sets of replicate studies of equal and varying duration are readily available in the
2 published literature but only a few reports have combined the data to specifically test the
3 value of duration in explaining variation of plant response or to evaluate exposure indices
4 based on statistical fit. Lefohn et al. (1988a) were the first to fit a common response model
5 to combined data from two replicate studies of varying duration using various exposure
6 indices. Greater yield losses occurred when plants were exposed for the longer duration,
7 indicating that the duration component of exposure was important in influencing response and
8 that a cumulative-type index was able to describe adequately the relationship between
9 exposure and yield. More recent papers have reported results of the two years of replicate
10 studies and a few papers have used the regression approach with and without variance
11 components for sites and/or years to evaluate various exposure indices based on the adequacy
12 of fit of a common response model.
13 A number of the papers relevant to the study of components of exposure influencing
14 plant response report only the mean and total exposure (i.e., sum of all hourly average
15 concentrations—SUMOO) indices. Because exposure indices weight hourly 03 concentrations
16 differently, it is almost impossible to transform one index into another. One must return to
17 the original data, which in many cases are not available, to generate alternative exposure
18 indices. Therefore, unless adequate information is given to allow calculation of exposure
19 indices, the analysis of reported results from individual and combined data to evaluate
20 different exposure indices is not possible, although it may be possible to perform
21 retrospective evaluation of the structure of exposure in altering plant response.
22 Another concern relates to the experimental design, particularly the number, kind, and
23 levels of exposure used in the study. Generalization of experimental results is largely
24 dependent on the degree to which atmospheric and biospheric conditions mimic those of the
25 target population. However desirable the need to mimic the real world, understanding the
26 relationship between exposure and the ensuing response (i.e., plant yield) and to identify
27 those components of exposure in influencing response may require the use of exposure
28 regimes with temporal pattern, concentration, and/or structure that are not observed in
29 nature. Research goals are difficult to attain when the investigator requires exposure levels
30 between charcoal-filtered and near-ambient conditions, but the mathematician who is
December 1993 5-116 DRAFT-DO NOT QUOTE OR CITE
-------
1 modeling the experiment requires higher than ambient levels of O3 to better describe the
2 nature of plant response to O3.
3 For example, the O3 exposures utilized by the NCLAN program have been described as
4 producing artificial regimes that do not mimic actual conditions. Lefohn and Foley (1992)
5 have compared the O3 exposures for the charcoal-filtered (CF) treatments from 13 NCLAN
6 data sets with clean monitoring sites in the United States. The authors concluded that the
7 exposures in the CF treatments were lower than those experienced under ambient conditions
8 at clean O3 monitoring sites and that the CF treatments did not mimic those exposures that
9 occurred under ambient conditions at the cleanest sites in the United States. Most of these
10 clean monitoring sites, as described by Lefohn and Foley (1992), are not representative of
11 the crop-growing regions of interest, yet, it is doubtful that the ambient conditions inside the
12 CF chambers would be experienced at any sites in the United States. The application of
13 treatment levels so low that they are almost never observed anywhere may result in
14 statistically significant differences that occur when growth comparisons are made between
15 plants grown in CF and nonfiltered (NF) treatments that are not relevant. Ideally, each
16 exposure in the design should be representative of ambient conditions experienced in the
17 geographic area of interest and the lowest treatment level (i.e., CF) should be set at the
18 cleanest of these sites. Because treatment levels are derived from ambient conditions
19 experienced at the research site, CF chambers may not be identical to ambient conditions at
20 the cleanest site in the region of interest, but may represent ambient conditions observed
21 somewhere in the world.
22 In addition to the CF concentration regimes, Lefohn et al. (1988a) have reported that
23 the highest treatments have a tendency to display bimodal distributions that are unrealistic
24 (Figure 5-9). At this time, there is no evidence to suggest whether or not these higher
25 ambient exposures provide realistic information on the impact of C^ on plant response.
26 Studies that utilize exposures with peak concentrations above 0.40 ppm may not provide
27 realistic evidence of O3 impact on plant response in the United States. These studies provide
28 limited evidence for substantiating the value of peak concentrations in influencing response.
29 Consequently, these studies are not included in this section.
30
December 1993 5-117 DRAFT-DO NOT QUOTE OR CITE
-------
oo
3
tn
ve
^ 5s £
&3 w T3
B il 8
as
a
£1
§
s a
»ff
rf „
FB.
B'
00
P?l
oo ^ g
S3
ll
I
i
- 63
Number of Occurrences
200 400 600
Number of Occurrences
800
50
250-
300^
50
j_
250-
300-
100
_i_
150
Number of Occurrences
50 100
50
00
X
>
3
a-
0'
2
O
N
0
3
CD
100-
-
-
-
150-
-
-
-
200-
250-
300-
d>
150
^
X
>
g;
-------
1 5.5.2.2 Studies with Two or More Different Patterns of Exposure
2 Experiments that focus on the structure of exposure have shown that plant response is
3 differential to temporal patterns of O3 exposure. For crop species, there is evidence to
4 suggest that plant response is influenced more by higher concentrations than lower
5 concentrations or exposure duration. Greater response to concentration occurred when plants
6 were predisposed to low concentrations for a few days or when peaks occurred just prior to
7 or at maximum leaf expansion (U.S. Environmental Protection Agency, 1978, 1986). Plants
8 exposed to two (or more) different exposure patterns of equal exposure (i.e., same SUMOO
9 value) showed greater foliar injury response to:
10
11 1. The short-term, high-concentration exposure than the longer-term exposure
12 with lower peak concentrations (Heck et al., 1966; Heck and Tingey,
13 1971; Bennett, 1979; Nouchi and Aoki, 1979; Amiro et al., 1984;
14 Ashmore, 1984; Tonneijck, 1984); and
15
16 2. The exposure that predisposes plants to low O3 concentrations for a few
17 days prior to a high O3 concentration than exposures that have a set diurnal
18 pattern of O3 concentrations or less than two days of respite time between
19 high concentrations (Heck and Dunning, 1967; Johnston and Heagle, 1982;
20 Heagle and Heck, 1974; Runeckles and Rosen, 1977; Mukammal, 1965;
21 Stan and Shicker, 1982).
22
23
24 The studies that applied the same exposure using different patterns of exposure have
25 been reviewed in previous criteria documents (U.S. Environmental Protection Agency, 1978,
26 1986, 1992) and substantiate the role of concentration, temporal dynamics, respite time, and
27 predisposition in influencing the magnitude of plant response to O3.
28 Musselman et al. (1983) and Hogsett et al. (1985) were among the first to demonstrate
29 that variable concentrations produced greater effect on plant growth than fixed or set diurnal
30 patterns of exposure of equal total exposure with lower peak concentrations (Table 5-11).
31 Musselman et al. (1986), in a subsequent experiment, exposed kidney bean plants to either a
32 simulated ambient or a uniform concentration of equal total exposure and equal peaks (at two
33 levels of 0.30 and 0.40 ppm) and found the effects of the two distributions were not
34 significantly different (Table 5-11). Consequently, when peak concentrations and total
35 exposures are equal, the diurnal distribution of concentrations appears not to be an important
36 factor (U.S. Environmental Protection Agency, 1992).
December 1993 5419 DRAFT-DO NOT QUOTE OR CITE
-------
t—'
rO
O
TABLE 5-11. A SUMMARY OF STUDIES REPORTING THE EFFECTS OF OZONE
FOR TWO OR MORE EXPOSURE PATTERNS ON THE GROWTH,
PRODUCTIVITY, OR YBELD OF PLANTS
Species
Glycine max L.
Merrill cv. Davis,
Forrest, Bragg, and
Ransom
Medicago saliva L.
Facility'
OTC in
pots
OTC in
pots
Total
Number
Chambers Exposure Patterns
24 16 combinations of CF or
NF over 4 quarters
(31 -days/quarter)
8 Episodic (E), Daily Peak
(DP)
Exposure Concentration (ppm)/Exposure
Duration (ppm-h)
124 days M7 (ppm): CF range from
.016 to .038 over the 4 quarters,
NF range from 0.96 to 0.98 over
the 4 quarters
133 days Equal SUM07 (ppm-h):
DPH=113, DPL=63, EH=117,
EL=72
Equal SUMOO (ppm-h):
DPH=183, DPL=140, EH=193,
Variable Effect Reference
Total Forrest: greater effect in Q3 than Heagle et al.
seed in other quarters. Davis: no (1991)
weight consistent effect Ql, significant
but similar effects for Q2, Q3, and
Q4. Ransom: no significant 03
effects in Ql or Q2, and equal
responses in Ql, Q3, and Q4.
shoot 91 and 67% reductions for EH andHogsett et al.
dry wt DPH (*). Significant difference (1985a)
between E and DP regimes.
Treatment means are ordered
CF
-------
TABLE 5-11 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF OZONE
FOR TWO OR MORE EXPOSURE PATTERNS ON THE GROWTH,
3
sr
£2 Species
Phaseolus vulgaris L.
cv. Calif. Dark Red
Kidney Bean
Pinus ponderosa Laws,
PRODUCTIVITY, OR YBELD OF PLANTS
Facility
GC in
pots
OTCin
Total
Number
Chambers
10
15
Exposure Patterns
Initial exposure of .3 ppm
for 3-h and second exp of
.3 ppm at 2-6 (or 1-5)
days after initial exp.
Episodic (E), High Elev
Exposure
Duration
2-6 days
in 1984
and
1-5 days
in 1985
111 days
Concentration (ppm)/Exposure
(ppm-h)
Equal maximum concentration of
0.30 ppm.
Equal SUMOO (ppm-h):
Variable
total dry
weight
leaf
Effect6
Reductions due to the second
exposure were significant when
exposures were 3-6 days apart in
1984 and 5 days apart in 1985.
At low level, 64 and 0%
Reference
McCool et al. (1988)
Hogsett and Tingey (1990)
Populus tremuloides pots
michx.
(HE), Daily Peak (DP)
DPH=HEH=EH=170,
HEL=EL=122.
stem, reductions for E and HE(*). At
stem, high level, 42 %, 75 %, and 7 %
root, dry reductions for E, HE, and DP(*).
weight Treatment means are ordered
E
-------
1 One recent study exposed bean plants to two consecutive exposures of 0.30 ppm for
2 3 h per day in the rapid vegetative growth stage and showed greater reductions in total dry
3 weight when exposures were three to six days apart (McCool et al., 1988) (Table 5-11),
4 consistent with earlier results on the role of predisposition in influencing response (e.g.,
5 Hogsett et al., 1988). Predisposition to a high concentration above the level that causes
6 visible injury may increase plant sensitivity over tune (Mukammal, 1965). As a result, the
7 subsequent response to a high concentration following recovery may be greater than
8 experienced in prior exposures. In future modeling efforts, this phenomenon may have to be
9 taken into consideration by the weighting of hourly concentrations for properly characterizing
10 plant exposure.
11 Sensitivity of plants to O3 is a function of stomatal conductance and varies with
12 phenology. To test the role of phenology, Heagle et al. (1991) applied 16 patterns of
13 exposure in combinations of either charcoal-filtered (CF) or nonfiltered (NF) for each quarter
14 of the experimental period (31 days/quarter) (Table 5-11). The authors concluded that plant
15 response was differential to phonological stage of development and that exposures during
16 mid- to late-growth stages caused a greater yield loss than early exposures. For crops,
17 foliage appears to be most sensitive to O3 just prior to or during maximum leaf expansion
18 (U.S. Environmental Protection Agency, 1978). These results are consistent with earlier
19 studies (Lee et al., 1987) that reported better statistical fits to response using exposure
20 indices that preferentially weighted hourly O3 concentrations during the period of anthesis to
21 seed fill.
22 There is very limited information on the nature of seedling response to O3 (see
23 Section 5.3.7) and much less is known about the role of exposure-dynamic factors (e.g.,
24 concentration, duration, respite time, temporal variation) in influencing biomass response in
25 long-lived species. In a study by Hogsett and Tingey (1990), ponderosa pine and aspen
26 seedlings were exposed to three different exposure patterns of equal SUMOO over a 5-mo
27 growing season (Table 5-11). One pattern simulated ambient conditions at high-elevation
28 sites remote from urban influence. A second pattern (i.e., episodic) was representative of
29 ambient conditions at low-elevation sites at rural or near-urban locations. A third pattern
30 was artificial and had a diurnal pattern rising to a peak at 2 ppm, repeated daily for the
31 length of the experimental period. Growth reductions in both ponderosa pine and aspen were
December 1993 5-122 DRAFT-DO NOT QUOTE OR CITE
-------
1 greatest in the episodic exposure pattern, which had the largest peak concentrations of the
2 three patterns. The smallest growth reductions in both species were observed with the high-
3 elevation pattern that had peak concentrations less than 0.10 ppm. The authors concluded
4 that temporal pattern and concentration were important in influencing long-term growth
5 response of tree seedlings, just as in crops, and, consequently, should be considered in
6 measures of exposure.
7 When yield is considered, a number of exposure-dynamic factors, including
8 concentration, temporal pattern, predisposition and respite times, as well as phenological
9 stage of plant development, have been shown to influence the impact of O3 on plant
10 response. Evidence from studies of kidney bean (Musselman et al., 1983), alfalfa (Hogsett
11 et al., 1985a), tobacco (Heagle et al., 1987), soybean (Heagle et al., 1986), ponderosa pine
12 and aspen (Hogsett and Tingey, 1990) suggests that concentration and temporal variation of
13 exposure are important factors hi influencing biomass production and, consequently become
14 considerations in measures of exposure. Because the SUMOO index weights all
15 concentrations equally, the SUMOO is inadequate for characterizing plant exposure to
16 O3 (Lefohn et al., 1989). Other factors, including predisposition time (McCool et al., 1988)
17 and crop development stage (Heagle et al., 1991), contribute to variations in biological
18 response, which suggests the need for weighting O3 concentrations to account for
19 predisposition time and phenology. However, the roles of predisposition and phenology in
20 influencing plant response vary with species and environmental conditions and are not
21 understood well enough to allow specification of a weighting function for use in
22 characterizing plant exposure.
23
24 5.5.2.3 Combinations of Years, Sites, or Species: Comparisons of Yield Losses with
25 Different Exposure Durations
26 Duration has not been a focus in experimental designs of studies that applied two or
27 more exposure regimes over the growing season. Several lines of evidence suggest that the
28 ultimate yield depends upon the cumulative impact of repeated peak concentrations (U.S.
29 Environmental Protection Agency, 1986, 1992) and that O3-induced reductions in growth are
30 linked to reduced photosynthesis, which is impaired by the cumulative O^ exposure (Reich
31 and Amundsen, 1985; Reich, 1987; Pye, 1988). In the Agency's reviews of the literature
32 (U.S. Environmental Protection Agency, 1986, 1992), EPA concluded that, "When plant
December 1993 5_123 DRAFT-DO NOT QUOTE OR CITE
-------
1 yield is considered, the ultimate impact of an air pollutant on yield depends on the integrated
2 impact of the pollutant exposures during the growth of the plant." As a measure of plant
3 exposure, the appropriate index should differentiate between exposures of the same
4 concentration but of different duration. For example, a mean index calculated over an
5 unspecified time cannot accomplish this (Lefohn et al., 1988a; Hogsett et al., 1988; Tingey
6 et al., 1989, 1991; U.S. Environmental Protection Agency, 1986, 1992).
7 The paper by Lefohn et al. (1988a), reviewed previously in U.S. EPA (1992) along
8 with published criticisms and responses, was the first to fit a common response model to
9 combined data from two replicate studies of unequal duration (71 and 36 days for the 1982
10 and 1983 wheat studies, respectively, conducted at Ithaca, NY) to specifically test the
11 importance of duration in influencing plant response. Greater yield losses occurred in 1982
12 which can partially be attributed to the longer duration. Because the mean index ignores the
13 length of the exposure period, the year-to-year variation in plant response was minimized by
14 the use of several cumulative indices rather than the mean. Lefohn (1988) and Lefohn et al.
15 (1988b) concluded that duration has value in explaining variation in plant response and that a
16 cumulative-type index was preferred over a mean or peak index based on statistical fit.
17 When O3 effects are the primary cause of variation in plant response, plants from
18 replicate studies of varying duration showed greater reductions in yield or growth when
19 exposed for the longer duration (Lee et al., 1991; Olszyk et al., 1993; Adaros et al., 1991a)
20 (Table 5-12 Part A). Using NCLAN data for wheat, cotton, kidney bean, and potato (from
21 Penn State) from replicate studies with markedly different exposure durations, Lee et al.
22 (1991) showed that year-to-year variations in the magnitude of relative yield loss were
23 minimized by the use of exposure indices that are cumulative and weight peak concentrations
24 more than low concentrations, indicating that O3 effects are cumulative (Figure 5-10).
25 Olszyk et al. (1993), using the two NCLAN cotton studies summarized by Temple et al.
26 (1985) and Lee et al. (1991), in addition to cotton studies replicated at four sites in
27 California's San Joaquin Valley over two years, tested and compared various exposure
28 indices (SIGMOID, SUM06, M7, and 2ndHDM) based on statistical fit of a common
29 response model. A Weibull response model with variance components was fit to the
30 combined data and used to test for a common response (Gumpertz and Rawlings, 1991,
31 1992; Gumpertz and Pantula, 1992). The likelihood ratio test of parallel exposure-response
December 1993 5-124 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-12. A SUMMARY OF STUDIES REPORTING THE EFFECTS OF OZONE ON THE GROWTH,
PRODUCTIVITY, OR YIELD OF PLANTS FOR TWO OR MORE REPLICATE STUDIES HAVING
EQUAL TOTAL EXPOSURES AND EITHER VARYING DURATIONS (PART A) OR SIMILAR
DURATIONS (PART B)
Part A
Species
Brassica napus L.
var. napus
cv. callypso
Gossypium
Hirsutum L.
cv. Acala SJ2
T- •!• 3
Facility
OTC in
pots
OTC
Total No.
of Plots
'87: 18
'88: 24
'89: 16
'81: 12
'82: 12
'87:
'88:
'89:
'81:
'82:
Duration
05-13 to 08-10 (89)
05-02 to 08-24 (113)
05-08 to 08-01 (84)
07-06 to 09-15 (72)
06-04 to 09-09 (98)
'8V:
'88:
'89:
'81:
138
'82:
111
Concentration (ppm)/ Exposure (ppm-h)
M24 (MS) in ppb range from 5 (9) to 16 (43).
M24 (M8) in ppb range from 3 (5) to 16 (48).
M24 (M8) in ppb range from 6 (5) to 22 (62).
M7 (SUM06) range from 18 ppb (0 ppm-h) to
(68).
M7 (SUM06) range from 12 ppb (0 ppm-h) to
(71).
Variable
seed
dry
weight
lint
dry wt
'87: 27%
'88: 18%
'89: 11%
45% and
57% and
Effect0
reduction at M8=43 ppb
reduction at M8 = 48 ppb
reduction at MS = 62 ppb
(***).
(***).
(***).
66 % reductions at M7= 1 1 1 ppb .
60 % reductions at
SUM06=68 ppm-h.
Reference
Adaros et al.
(1991c)
Lee et al.
(1991),
Olszyk et al.
(1993)
Hordeum vulgare
L.
cf. Arena and
Alexis
Phaseoulus
vulgaris L.
cv. Calif Dark
Red
Kidney Bean
OTC in '88: 24 '88: 04-29 to 08-15 (108)
pots '89: 16 '89: 05-08 to 08-15 (99)
OTC '80: 20 '80: 08-20 to 09-10 (22)
'82: 20 '82: 08-11 to 10-06 (57)
'88: M8 (max 8-h mean) in ppb range from 5 (15) to seed
48 (89). dry wt
'89: M8 (max 8-h mean) in ppb range from 11 (27)
to 62 (101).
'80: M7 (SUM06) range from 24 ppb (0 ppm-h) to
139 (19).
'82: M7 (SUM06) range from 19 ppb (0 ppm-h) to
110(40).
seed
dry wt
Arena: 14% (*) and 6% (NS) reductions at Adaros et al.
M8=48ppb. (1991b)
Alexis: No reductions at M8=48 ppb
(NS).
13% and 59% reductions at M7= 110 ppb. Lee et al.
28% and 8% reductions at (1991)
SUM06= 19 ppm-h.
C, Phaseolus vulgaris OTC in
Trt L. pots
H cf. Rintintin
O
o
1^ Solanum OTC
O Tuberosum L.
" cv. Norchip
O
S Triticum aestivum OTC
,j L., cv. Vona
'881:4 '88 I: 06-15 to 08-04 (51) I. M8 (Max) in ppb range from 3 (19) to 48 (70).
'88 H: 6 '88 H: 07-24 to 08-29 (37) n. M8 (Max) in ppb range from 2 (19) to 50 (105).
'89 HI: 8 '89 ffl: 06-04 to 07-25 (52) ffi. M8 (Max) in ppb range from 6 (26) to 109
(159).
'85: 15 '85: 06-14 to 08-22 (70)
'86: 39 '86: 06-20 to 08-20 (62)
'82: 20 '82:05-18 to 07-17 (61)
'83: 12 '83: 06-12 to 07-17 (36)
'85: M7 (SUM06) range from 22 ppb (0 ppm-h) to
85 (47).
'86: M7 (SUM06) range from 24 ppb (0 ppm-h) to
88 (38).
'82: M7 (SUM08) range from 21 ppb (0 ppm-h) to
95 (40).
'83: M7 (SUM08) range from 26 ppb (0 ppm-h) to
96(21).
pod I. 2% reduction at M8=48 ppb (NS). Bender et al.
dry H. 0% reduction at M8=50 ppb (NS). (1990)
weight m. 0% (NS) and 47% (*) reductions at
M8=50 and 109 ppb.
tuber 42% and 25% reductions at M7 = 85 ppb. Lee et al.
weight 32% and 27% reductions at 12-H (1991)
SUM06=38 ppm-h.
seed 74% and 49% reductions at M7=95 ppb. Lefohn et al.
dry wt 49% and 62% reductions at (1988),
SUM08=21 ppm-h. Lee et al.
55% and 60% reductions at 7-h (1991)
SUM06=22 ppm-h.
-------
I
u>
TABLE 5-12 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF OZONE ON THE
GROWTH, PRODUCTIVITY, OR YIELD OF PLANTS FOR TWO OR MORE REPLICATE STUDIES
HAVING EQUAL TOTAL EXPOSURES AND EITHER VARYING DURATIONS (PART A)
OR SIMILAR DURATIONS (PART B)
Part A
Species
Facility
Total No.
of Plots
Duration
Concentration (ppm)/ Exposure (ppm-h)
Variable
Effect"
Reference
Triticum aestivum OTC in '88: 6
L., cv. Star and pots '89: 10
Turbo
'88: 04-27 to 08-23 (118)
'89: 05-09 to 08-15 (98)
'88: M8 (Max,SUM06) in ppb range from 4 (58,0) seed
to 51 (106,8.2). dry wt
'89: M8 (Max,SUM06) in ppb range from 10 (34,0)
to 113 (162,87).
Star: 20% (*) and 9% (NS) reductions at Adaros et al.
M8 = 51 ppb. (1991a)
Turbo: 25% (*) and 31% (*) reductions at
M8 = 51 ppb.
Triticum aestivum OTC in
L., cv. Star and pots
Turbo
'88:24 '88: 04-29 to 08-15 (108) '88: M8 (max 8-h mean) in ppb range from 5 (15) to seed Star: 26% (*) and 12% (*) reductions at Adaros et al.
'89:16 '89:05-08 to 08-15 (99) 48(89) dry wt M8 = 48ppb. (1991b)
'89: M8 (max 8-h mean) in ppb range from 11 (27) Turbo: 34% (*) and 17% (*) reductions at
to 62 (101) M8 = 48ppb.
Ul
t—*
N)
PartB.
Glycine max L.
Merr.
cv. Davis
Qlycine max L.
Merr.
cv. Williams
OTC in
pots
OTC
'77:8 '77: 06-17 to 10-10(116) '77: M7 (max) in ppb range from 27 (78) to 154
'78: 8 '78: 06-28 to 10-21 (116) (277).
'78: M7 (max) in ppb range from 28 (84) to 131
(241).
'81: 31 '81: 07-20 to 09-22 (65)
'82: 31 '82: 07-14 to 09-22 (71)
'83: 31 '83: 07-23 to 09-23 (63)
'81: M7 in ppb range from 15 to 64.
'82: M7 in ppb range from 17 to 99.
'83: M7 in ppb range from 19 to 132.
seed 47% and 37% reductions at M7=131 ppb. Cure et al.
dry wt (1986),
Heagle et al.
(1983)
bean 28%, 20%, and 32% reductions at Heggestad
dry M7=64ppb. and
weight 43 % and 41 % reductions at M7=99 ppb Lesser
in '82 and '83 (1990),
Heggestad
et al. (1988)
O Medicago saliva OTC
O L. cv. WL-514
25
O pjnus rigida OTC in
H Mill. pots
'84:30 '84: 03-16 to 10-10 (209) '84: M12 in ppb range from 16 to 109.
'85:30 '85: 03-23 to 10-09 (201) '85: M12 in ppb range from 10 to 94.
Exp. 1: 4 Exp. 1: 13 weeks
Exp. 2: 4 Exp. 2: 13 weeks
1: M8 in ppb range from 0 to 200 ppb (U).
2: M8 in ppb range from 0 to 200 ppb (U).
top 29% (*) and 25% (*) reductions at Temple et al.
drywt M12=94ppb. (1988)
total 49% and 46% reductions at M8=200 ppb. Schier et al.
dry wt (1990)
-------
§
TABLE 5-12 (cont'd). A SUMMARY OF STUDDZS REPORTING THE EFFECTS OF OZONE ON THE
GROWTH, PRODUCTIVITY, OR YIELD OF PLANTS FOR TWO OR MORE REPLICATE STUDIES
HAVING EQUAL TOTAL EXPOSURES AND EITHER VARYING DURATIONS (PART A)
OR SIMILAR DURATIONS (PART B)
PartB
Species
Pinus taeda L.
Total No.
Facility of Plots Duration
GC in '86: 15 '86: 09-15 to 12-04 (81)
pots '87: 15 '87: 07-27 to 10-15 (81)
Concentration (ppm)/ Exposure (ppm-h)
'86: SUMOO in ppm-h range from 0 to 99 (U).
'87: SUMOO in ppm-h range from 0 to 99 (U).
Variable Effect6
total 43 % and 28 % reductions at
dry wt SUMOO = 99 ppm-h
Reference
Shafer et al.
(1993)
Picea rubens OTC in '87: 12
Sarg. pots '88: 12
£j Pisum sativum L. ZAPS '86: 14
cv. Puget '87: 14
O Populus OTC in '88: 18
I* tremuloides pots '89: 18
(L: Michx clones
'87: 05-30 to 12-15 (199) '87: SUMOO in ppm-h are 32, 61, 91 and 119.
'88: 06-01 to 12-01 (184) '87: SUMOO in ppm-h are 36, 70, 101 and 135.
'86: last 58 days M12 and D25 (# days with 1-h cone >25 ppb)
'86: last 52 days used in simple linear regression
'88: 07-19 to 09-27 (71) '88: SUMOO in ppm-h are 5.0, 10.0, and 19.4 (U).
'89: 07-20 to 09-20 (64) '89: SUMOO in ppm-h are 7.7, 15.4, and 26.4 (U).
total
dry wt
averaged across all families.
Individual families show similar
reductions, e.g., 35% and 33% reductions
at SUMOO = 99 ppm-h for family 5.56,
14% and 12% reductions at
SUMOO=99 ppm-h for family 1.68.
0% (NS) reduction in biomass after 1st yr, Alscher et al.
8% (*) reduction at SUMOO= 135 ppm-h (1989)
after 2nd year of exposure Amundsen
etal.
(1991)
pea 0 % reductions at M12 =100 ppb
fresh wt based on linear regression models
stem and 36% (*) and 40% (*) reductions at
leaf dry SUMOO= 19.4 ppm-h.
wts
Runeckles
etal.
(1990)
Karnosky
etal.
(1992)
tj Triticum aestivum OTC
O L-. cv. Albis
O
CH Triticum aestivum OTC
Q L., cv. Albis
'86: 12 '86: 05-06 to 07-31 (86) '86: M24 (max) in ppb range from 12 (61) to 47
'87:16 '87: 04-27 to 08-10 (92) (181).
'88: 16 '88: 05-04 to 08-01 (89) '87: M24 (max) in ppb range from 12 (54) to 45
(175).
'88: M24 (max) in ppb range from 17 (65) to 45
(148).
'89: 24 '89: 05-16 to 08-14 (91) '89: M7 (SUM06) range from 18 ppb (0 ppm-h) to
'90: 24 '90: 05-14 to 08-09 (88) 62 (3.8).
'90: M7 (SUM06) range from 17 ppb (0 ppm-h) to
71 (5.6).
seed '86: 61% reduction at M24=47 ppb. Fuhrer etal.
dry wt '87: 27% reduction at M24=45 ppb. (1989)
'88: 65% reduction at M24=45 ppb.
seed 29 % and 22 % reduction at M7=62 ppb. Fuhrer et al.
dry wt 29% and 17% reduction at (1992)
SUM06=3.8 ppm-h.
-------
V
oo
TABLE 5-12 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF OZONE ON THE
GROWTH, PRODUCTIVITY, OR YIELD OF PLANTS FOR TWO OR MORE REPLICATE STUDffiS
HAVING EQUAL TOTAL EXPOSURES AND EITHER VARYING DURATIONS (PART A)
OR SIMILAR DURATIONS (PART B)
Part B Total No.
Species Facility8 of Plots Duration
Triticum aestivum OTC
L., cv. Severn,
Potomac, Oasis,
MD55 18308
'84: 20 '84: 05-14 to 06-22 (40)
'85: 20 '85: 05-06 to 06-15 (41)
Concentration (ppm)/ Exposure (ppm-h) Variable Effect
'84: M4 (AOT03) in ppb (ppb-h) range from 32 (0) seed
to 93 (10). dry wt
'85: M4 (AOT03) in ppb (ppb-h) range from 30 (0)
to 86 (9).
31 % (*) and 956 (NS) reductions at
M4 = 86ppb.
Reference
Slaughter
et al.
(1989)
*GC = controlled environmental growth chamber or CSTR, OTC= open-top chamber, ZAPS=zonal air pollution system.
bU = Uniform.
c* = significant at the 0.05 level, NS= not significant.
-------
1.00ft
Cotton
Wheat
.08 .12 i 20 40 60
M-7 (ppm) Sum06 (ppm-h)
Kidney Bean
0.03 0.06
M-7 (ppm) Sum06 (ppm-h)
Potato
20 40
M-7 (ppm) Sum06 (ppm-h)
0.03 0.06 0.09 0 20 40
M-7 (ppm) Sum06 (ppm-h)
Figure 5-10. Comparison of the Weibull exposure-response functions and its predicted
relative yield loss (PRYL) curves (relative to 0 ozone) using M-7 and
daytime SUM06 for replicate years of National Crop Loss Assessment
Network Program's data for (A) and (B) cotton (var. Acala SJ-2), (C) and
(D) wheat (var. Vona), (E) and (F) kidney bean (var. California light red),
and (G) and (H) potato (var. Norchip), respectively. Mean dry weights
and the Weibull exposure-response functions for replicate studies are given
in the top portion of the graphs (Lee et al., 1991).
December 1993
5-129
DRAFT-DO NOT QUOTE OR CITE
-------
1 curves was statistically significant for M7 and 2ndHDM for at least one set of cotton data,
2 indicating significant differences in the magnitude of response across years and/or sites.
3 On the other hand, the SIGMOID and SUM06 indices resulted in consistent patterns of
4 response for both sets of cotton data, as well as between sets of cotton data (Figure 5-11).
5 The authors concluded that the peak-weighted, cumulative indices minimized the temporal
6 and spatial variations in crop yield and better predicted cotton yield responses than the M7 or
7 2ndHDM indices. The mean and peak indices did not differentiate between exposure seasons
8 of differing duration and could not account for year-to-year differences in response.
9 The results of European studies with wheat (Adaros et al., 1991a,b), spring rape
10 (Adaros et al., 1991c), barley (Adaros et al., 1991b) and kidney beans (Bender et al., 1990),
11 using data from replicate studies with varying duration, are less conclusive as to the role of
12 duration in determining plant response (Table 5-12 Part A). Exposures are reported using a
13 mean index. Adaros et al. (1991a) showed a greater reduction in above ground dry weight
14 when exposed for the longer duration for wheat cultivar, Star, but not for cultivar, Turbo
15 (Figure 5-12). Adaros et al. (1991b), in another two-year study with barley (cv. Arena and
16 Alexis) and wheat (cv. Star and Turbo) involving mixtures of O3, SO2, and NO2, showed
17 greater reductions in yield when exposed for the longer duration for all species and cultivars
18 except barley cv. Alexis (Table 5-12 Part A). Ozone effects were insignificant in both years
19 for barley cv. Alexis. The authors did not attribute the differential response in growth and
20 yield to any single factor but the data suggested that 03 effects are cumulative. When
21 O3 exposure is the primary source of response, the mean exposure index of unspecified
22 duration could not account for the year-to-year variation in response.
23 The role of duration in influencing growth or yield is unclear for the other studies
24 because of the following limitations in the data:
25
26 1. Treatment levels were below the levels to induce injury or damage to
27 kidney bean plants in two of the three years. None of the years produced a
28 significant O3 effect at or below 70 ppb concentration (Bender et al.,
29 1990). Similarly, the study with barley showed no significant O3 effects.
30
31 2. Differences in growing conditions and varying kinds of interactions
32 between O3, SO2, and NO2 resulted in different sizes of control plants of
33 spring rape over years and affected the magnitude of response to O3.
December 1993 5-130 DRAFT-DO NOT QUOTE OR CITE
-------
3
en
Predicted Relative Yield Loss (%)
Predicted Relative Yield Loss (%)
-------
120
90 -
f 60
30
Star
Total Biomass
1988
1989
Grain Yield
1988
I
Turbo
XV
I
0.015 0.03 0 0.015
Ozone Concentration (ppm)
0.03
0.045
Figure 5-12. Relative effect of ozone on growth and yield of spring wheat cultivars (var.
Star and Turbo) from two growing seasons (Adaros et al., 1991a).
Compared to 1987, yield of control plants increased by 32% in 1988 and by
94% in 1989 (Adaros et al., 1991c). Consequently, the evidence of duration
as the primary cause of differences in response over years was difficult to
substantiate.
When durations were nearly equal, plant response to O3 were similar for 2- or 3-year
studies with alfalfa (Temple et al., 1988), pea (Runeckles et al., 1990), soybean (Heagle
et al., 1983; Heggestad and Lesser, 1990; Cure et al., 1986), wheat (Fuhrer et al., 1989,
1992), aspen clones (Karnosky et al., 1992), loblolly pine (Shafer et al., 1993), and pitch
pine (Schier et al., 1990) (Table 5-12 Part B). For example, year-to-year variations in wheat
yield response to O3 were small for the three years having durations between 86 and 92 days
to allow pooling of the data to fit a common Weibull model using Rawling's solar-radiation
December 1993
5-132
DRAFT-DO NOT QUOTE OR CITE
-------
weighted mean index (Fuhrer et al., 1989) (Figure 5-13). Different growing conditions were
reported in a number of studies (Shafer et al., 1993; Fuhrer et al., 1989) but found no
interaction between O3 and climatic effects. On the other hand, Slaughter et al. (1989)
reported reductions in wheat grain yield of 69% and 9% in a two-year study having equal
durations, which the authors attribute to differences in rainfall and temperature.
Environmental conditions in 1985 favored greater photosynthate partitioning for grain
development rather than for vegetative growth, resulting in larger plants in 1985. Air
pollution effects may not have been the primary source of variation in response and,
consequently, the data are unable to substantiate the role of duration in influencing response.
\.
w3^ X-NX-
"X^\
S--P
00 0.02 0.04 0.06 0.08
Year
& - 1986
— -B— 1987
......4..... 1988
— e — Combined
^^~^
i
0.10 0.1
Mean Weighted Ozone Concentration (ppm)
Figure 5-13. Weibull exposure-response curves for the relative effect of ozone on grain
yield of spring wheat for three years, individually and combined (Fuhrer
et al., 1989).
December 1993
5-133 DRAFT-DO NOT QUOTE OR CITE
-------
1 These studies report plant response as a function of a mean exposure index and do not
2 evaluate or compare various exposure indices based on statistical fit. In a series of three
3 papers that examined the response of spring wheat to O3 at higher elevations, Grimm and
4 Fuhrer (1992a,b) and Fuhrer et al. (1992) conducted a 2-year study in which the flux of
5 O3 was determined in open-top chambers. Plants were exposed to O3 for periods lasting
6 44 and 50 days in 1989 and 1990, respectively and flux measurements were taken repeatedly
7 over the experimental period. In addition to O3 flux, exposures were characterized using
8 M7, M24, SUM06, and the solar radiation-weighted mean index (Rawlings et al., 1988).
9 The quadratic response curves relating the various indices with grain yield showed
10 year-to-year variations were minimized using the mean O3 flux index (Figure 5-14). The
11 other three exposure indices showed slightly greater yield losses in 1989 than in 1990, in
12 contrast with longer exposure in 1990 and drier conditions in 1989. The authors concluded
13 that the O3 flux related well with yield because the mean flux incorporated environmental
14 factors, canopy structure, and physiological processes, which affected the uptake of O3 from
15 the air to the leaf interior. The measures of ambient condition ignored these factors and
16 consequently, were unable to account for all of the year-to-year variability hi wheat response.
17 The authors suggested that O3 flux was a surrogate of Fowler and Cape's (1982) "pollutant
18 absorbed dose" and appeared to be the relevant measure for use hi relating exposure and
19 plant response.
20 Alscher et al. (1989) and Amundson et al. (1991) report on the impact of Oj on
21 growth, injury, and biomass response of 2-year-old red spruce seedlings after 1 and 2 years
22 of exposure, respectively. Exposures were characterized using the M12 (or M7), M24, and
23 SUMOO indices. No significant O3 effects on biomass were detected hi 1987 (Alscher et al.,
24 1989) because stomatal conductances in red spruce are inherently low and, consequently,
25 result in low rates of pollutant uptake (Seiler and Cazell, 1990). However, in the second
26 year, O3 reduced leaf and root starch, increased foliar antioxidant content, and reduced
27 biomass of 1988 fixed-growth foliage. However, O3 effects on biomass were slight hi the
28 second year. The authors concluded O3 effects are cumulative because the onset of damage
29 occurred in the second year rather than the first year of exposure.
30 When the concentrations are above the levels hi which injury and damage occur, plant
31 response is influenced by exposure duration. The results of these studies are hi general
December 1993 5-134 DRAFT-DO NOT QUOTE OR CITE
-------
1
1.2
1.0
0.8
0.6
0.4
0.2
0,
(a)
0 20 40 60 80 100 120 140 160 180
7 h day1 (ng nf)
0 20 40 60 80 100 120 140 160 180
Radiation weighted mean (ng nf)
• 1.2
1.0
0.8
0.6
0.4
0.2
0
20
40 60 80 100 120 140
Sum06 (mg rrt3 h)
20 40 60 80 100
Ozone flux (pg nf mlri1)
120
Figure 5-14. Quadratic exposure-response curves for the relative effect of ozone on
grain yield of spring wheat in 1989 and 1990 using four different exposure
indices (a-d) (Fuhrer et al., 1992).
1
2
3
4
5
6
7
8
9
10
11
12
13
agreement that O3 effects are cumulative and the ultimate impact of long-term exposures tp
O3 on crops and seedling biomass response depends on the integration of repeated peak
concentrations during the growth of the plant. Consequently, the mean or peak indices are
inappropriate since the length of exposure is unspecified and these indices cannot differentiate
among exposures of the same concentration but varying in duration. These results support
the conclusion that an appropriate O3 index should cumulate hourly concentrations in some
fashion to reflect the nature of O3 on plant response. Fuhrer et al. (1992) suggested that the
weighting function should reflect the relationship between ambient condition and internal
O3 flux, consistent with the mode of action of 03 on plants and with earlier findings that
peak-weighted, cumulative indices give better predictions of plant response than mean or
peak indices.
December 1993
5-135
DRAFT-DO NOT QUOTE OR CITE
-------
1 5.5.2.4 Comparisons of Measures of Exposure Based on Reanalysis of Single-Year,
2 Single-Species Studies
3 Studies cited in previous sections focused on the role of the structure of exposure in
4 influencing plant response but do not specifically identify the weighting function for use in
5 characterizing plant exposure to O3. In addition to these types of studies, other studies have
6 focused on comparison of measures of exposure based on reanalysis of single-year, single
7 species studies. The variety of statistical approaches used to relate exposure and plant
8 response range from informal description of the distributions of O3 concentrations associated
9 with response to more formal regression-based procedures.
10 The regression approach is designed to select those exposure indices that properly order
11 and space the treatment means along the horizontal axis to optimize the fit of a linear or
12 curvilinear model. However, because the experimental designs are not designed to evaluate
13 various indices, the power of the regression approach to identify the important exposure-
14 dynamic factors influencing plant response is less than desirable (Lefohn et al., 1992a).
15 Consequently, these retrospective studies provide less substantiating evidence of the role of
16 exposure-dynamic factors (e.g., concentration, duration, temporal pattern, and respite time)
17 than those studies with experimental designs and analyses that focus on specific components
18 of exposure.
19 Most of the early retrospective studies reporting regression results using data from the
20 NCLAN program and/or from Corvallis, Oregon (Lee et al., 1987, 1988; Lefohn et al.,
21 1988a: Tingey et al., 1989) or using data collected by Oshima (U.S. Environmental
22 Protection Agency, 1986; Musselman et al., 1988) were in general agreement and
23 consistently favored the use of cumulative peak-weighted exposure indices. These studies
24 have been previously reviewed by the Agency (U.S. Environmental Protection Agency,
25 1992). Lee et al. (1987) suggested that exposure indices that included all the data (24 h)
26 performed better than those that used only 7 h of data; this is consistent with the conclusions
27 of Heagle et al. (1987) that found plants receiving exposures for an additional 5 h/day
28 showed 10% greater yield loss than those exposed for 7-h/day. In a subsequent analysis
29 using more of the NCLAN data, Lee et al. (1988) found the "best" exposure index was a
30 general phenologically weighted, cumulative-impact index (GPWCI) with sigmoid weighting
31 on concentration and a gamma weighting function as surrogate of time of increased plant
32 sensitivity to O3. For most cases, Lee et al. (1987) computed their exposure indices based
December 1993 5-136 DRAFT-DO NOT QUOTE OR CITE
-------
1 on the daylight exposure periods used by the NCLAN investigators. The exposure indices
2 with minimum RSS were those indices that (1) cumulated hourly O3 concentrations over the
3 growth of the plant, (2) gave preferential weighting to peak concentrations, and
4 (3) phenologically weighted the exposures to emphasize concentrations during the plant
5 growth stage. The paper by Tingey et al. (1989) is a summarization of the results in Lee
6 et al. (1988) and shows the limitations of the mean index.
7 Lefohn and Foley (1992) characterized the NCLAN exposures that had a SUM06 level
8 closest to those that predicted a 20% yield loss, using the exposure-response equations as
9 reported in Lee et al. (1991) and Tingey et al. (1991). Lefohn and Foley (1992)
10 characterized the hourly average concentrations using percentiles, HRS06, HRS10, SUM06,
11 and W126 for each of 22 NCLAN studies. The authors noted that the frequent occurrence in
12 many cases of high hourly concentrations (^:0.10 ppm) may have been partly responsible for
13 the 20% yield loss. The number of hourly average concentrations ranged from 0 to 515 with
14 only one of the 22 NCLAN experiments experiencing no hourly average concentrations
15 ^0.10 ppm, while the remaining experiments experienced multiple occurrences ^0.10 ppm.
16 The repeated occurrences of high hourly average concentrations were a result of the NCLAN
17 protocol (Table 5-13). As a result of their analysis, Lefohn and Foley (1992) and Lefohn
18 et al. (1992b) stressed that because the NCLAN experiments contained peak hourly average
19 concentrations, it is important that any index selected to characterize those regimes
20 responsible for growth reduction adequately capture the presence of these peak concentrations
21 when attempting to predict biological responses using actual ambient air quality data.
22 For example, Tingey et al. (1991), using mostly NCLAN data, identified 24.4 ppm-h
23 as the SUM06 value, calculated over a 3-mo period, that would protect 50% of the NCLAN
24 crops analyzed at the 10% yield reduction level. There are monitoring sites in the United
25 States that experience 3-mo cumulative SUM06 values greater than 24.4 ppm-h, but do not
26 experience frequent occurrences of hourly average concentrations > 0.10 ppm. For
27 example, 24% (1987), 10% (1988), 30% (1989), 25% (1990), and 31% (1991) of the rural
28 agricultural sites listed in the U.S. EPA AIRS database experienced 3-mo cumulative SUM06
29 values greater than 24.4 ppm-h, but experienced fewer than 11 hourly average concentrations
30 equal to or greater than 0.10 ppm. Lefohn and Foley (1992) noted that agricultural crops
31 grown at a site experiencing a 3-mo cumulative SUM06 value greater than 24.4 ppm-h, but
December 1993 5437 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-13. SUMMARY OF OZONE EXPOSURES THAT ARE CLOSEST TO THOSE PREDICTED FOR
20% YIELD REDUCTION PER SUM06 EXPOSURE RESPONSE MODELS USED BY
1
VO
Ul
1
oo
M
u
i
g
0
s
O
LEEET
Experiment Chamber
A80SO- Corsoy NF+0.03-1
A83SO - Amsoy NF+0.03-1
A83SO - Corsoy NF+0.03-1
A85SO - Corsoy_79 D NFx2.00-lD
A85SO - Corsoy_79 W NFx2.00-lW
A86SO - Corsoy_79 D NFx2.5-lD
A86SO - Corsoy_79 W NFx2.0-lW
B83SO - Corsoy _79 D NF-1D
B83SO - Corsoy_79 W NF+0.03-1W
B83SO - Williams D NF+0.03-1D
B83SO - Williams W NF+0.03-1W
I81SO - Hodgson NF+0.06-1
R81SO - Davis NF-1
R82SO - Davis NF+0.02-1
R83SO - Davis Dry NF+0.02-1D
R83SO - Davis Wet NF+0.02-1W
R84SO - Davis Dry NF+0.015-1D
R84SO - Davis Wet NF+0.015-1W
R86SO - Young Dry NFxl .3- ID
R86SO - Young Wet NFxl .3-1 W
SORGHUM
A82SG - Dekalb NF+0.10-1
WHEAT
A82WH - Abe NF+0.03-1
A82WH - Arthur_71 NF+0.06-1
Min.
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
10
0.000
0.001
0.001
0.000
0.000
0.002
0.002
0.002
0.002
0.002
0.002
0.004
0.003
0.001
0.002
0.002
0.006
0.006
0.003
0.003
0.001
0.002
0.002
AL. (1991) IN SELECTED
30
0.011
0.014
0.014
0.008
0.011
0.016
0.015
0.006
0.006
0.006
0.006
0.007
0.015
0.013
0.015
0.015
0.018
0.018
0.013
0.013
0.010
0.015
0.015
50
0.026
0.028
0.028
0.023
0.026
0.035
0.033
0.018
0.019
0.019
0.019
0.015
0.026
0.026
0.030
0.030
0.030
0.029
0.024
0.023
0.023
0.027
0.027
Percentiles
70 90
0.045
0.049
0.049
0.051
0.063
0.085
0.065
0.037
0.049
0.049
0.049
0.031
0.043
0.047
0.055
0.054
0.047
0.046
0.047
0.046
0.055
0.047
0.053
0.077
0.083
0.083
0.110
0.114
0.137
0.105
0.063
0.084
0.084
0.084
0.083
0.066
0.080
0.089
0.087
0.077
0.075
0.089
0.087
0.145
0.079
0.109
95
0.090
0.098
0.098
0.129
0.134
0.161
0.124
0.074
0.097
0.098
0.097
0.090
0.075
0.091
0.104
0.101
0.089
0.089
0.107
0.101
0.160
0.094
0.121
NCLAN EXPERIMENTS (ppm)
99
0.111
0.123
0.123
0.160
0.162
0.207
0.161
0.087
0.118
0.118
0.118
0.105
0.088
0.123
0.126
0.119
0.113
0.110
0.137
0.129
0.185
0.113
0.144
SUM
Number Number of Occurrences 06 08
Max ofObs. >0.06 S0.08 >0.10 (ppm-h)
0.123 1,344
0.168 1,992
0.168 1,992
0.194 2,352
0.199 2,352
0.279 2,040
0.242 2,040
0.111 1,512
0.135 1,512
0.137 1,512
0.135 1,512
0.132 1,680
0.145 2,664
0.203 2,160
0.155 2,640
0.138 2,640
0.140 2,496
0.159 2,496
0.206 2,568
0.198 2,568
0.223 2,040
0.149 1,344
0.170 1,344
263
467
467
657
729
784
719
184
359
364
359
323
421
471
721
698
512
486
597
573
599
300
373
113
223
223
495
547
654
510
51
198
204
198
191
79
218
378
359
208
193
345
323
557
130
293
35
90
90
319
358
515
271
5
70
66
70
29
6
56
163
140
59
62
175
136
516
43
186
21.1
39.16
39.1
67.5
75.1
92.1
69.6
13.5
30.1
30.5
30.1
26.7
30.2
39.0
61.6
58.7
41.2
38.9
53.7
50.2
79.1
24.1
37.4
10.7
22.1
22.1
56.2
62.5
83.2
55.1
4.4
18.9
19.5
18.9
17.4
7.0
21.4
37.7
35.0
19.9
18.6
36.2
32.8
76.3
12.5
31.8
W126
(ppm-h)
17.7
33.2
33.2
63.0
70.0
88.6
63.7
10.6
25.8
26.0
25.8
22.9
2.6
33.1
53.1
50.7
34.8
32.6
47.7
44.1
78.2
19.8
35.3
-------
TABLE 5-13 (cont'd). SUMMARY OF OZONE EXPOSURES THAT ARE CLOSEST TO THOSE PREDICTED FOR
20% YIELD REDUCTION PER SUM06 EXPOSURE RESPONSE MODELS USED BY
1
£2 Experiment
A83WH - Arthur_71
BTI82WH - VONA
BTI83WH - VONA
CORN
A81MA - PAG 397
A81MA - Pioneer
COTTON
R82CO - Stoneville
R85CO (McNair) Dry
V R85CO (McNair) Wet
i— '
^> PEANUT
R80PN - NC-6
2 TOBACCO
3 R83TO - McNair 944
LEE ET AL. (1991)
Chamber
NF+0.06-1
NF-1
NF-1
NF +0.06-2
NF +0.06-2
NF-1
NFxl.99-lD
NFxl. 33-1 W
NF +0.015-1
NF +0.020-1
Min.
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
10
0.004
0.011
0.006
0.000
0.000
0.003
0.003
0.003
0.004
0.003
30
0.019
0.025
0.021
0.008
0.008
0.018
0.012
0.012
0.017
0.018
50
0.032
0.034
0.036
0.020
0.020
0.029
0.024
0.024
0.029
0.037
IN SELECTED
Percentiles
70 90
0.054
0.042
0.049
0.052
0.052
0.044
0.052
0.041
0.043
0.061
0.108
0.057
0.071
0.111
0.111
0.065
0.117
0.073
0.066
0.089
95
0.123
0.064
0.083
0.126
0.126
0.074
0.154
0.091
0.076
0.104
NCLAN EXPERIMENTS (ppm)
99 Max
0.159 0.186
0.072 0.098
0.097 0.116
0.150 0.187
0.150 0.187
0.087 0.152
0.221 0.291
0.129 0.166
0.091 0.112
0.121 0.155
Number
of Obs.
1,296
1,464
864
1,968
1,968
2,856
3,000
3,000
2,688
1,968
Number of Occurrences
>0.06 feO.08 aO.10
365
114
165
552
552
390
810
487
369
611
295
2
51
461
461
64
609
226
101
288
186
0
4
306
306
7
407
118
5
117
SUM
06 08
(ppm-h)
37.4
7.6
12.4
57.5
57.5
28.2
92.9
41.4
27.2
50.7
32.5
0.2
4.7
51.0
51.0
5.8
78.9
23.5
8.8
28.4
W126
(ppm-h)
35.6
6.2
9.8
55.1
55.1
22.7
88.2
35.9
22.0
42.6
-------
1 with infrequent high hourly average concentrations (e.g., >0.10 ppm), might experience less
2 yield reduction than predicted using NCLAN experimental results. For rural forest sites,
3 21% (1987), 23% (1988), 54% (1989), 50% (1990), and 52% (1991) of the sites exhibited
4 3-mo cumulative SUM06 values greater than 24.4 ppm-h, but fewer than 11 hourly average
5 concentrations equal to or greater than 0.10 ppm. Tables 5-14 and 5-15 illustrate that sites
6 that experience 3-mo, SUM06 values >24.4 ppm do not necessarily have peaks, whereas
7 sites that experience values <24.4 ppm-h do have peaks.
8 Krupa et al. (1993) tested the importance of ambient O3 frequency distributions in
9 eliciting a response based on comparisons of non-filtered (NF) and charcoal-filtered (CF)
10 treatments using the NCLAN data. Eight cases where yields in NF were significantly lower
11 than CF were matched with corresponding cases where the yields in CF and NF were not
12 significantly different. The frequency distributions of NF treatments resulting in significant
13 yield loss had a greater proportion of concentrations >50 ppb than those of NF treatments
14 resulting in no significant yield loss, indicating that these intermediate concentrations may
15 contribute to determining the magnitude of response. The authors conclude that
16 concentrations between 50 and 87 ppb are more important than concentrations > 100 ppb in
17 eliciting plant response. These conclusions are difficult to substantiate with the selected data
18 because: (1) the NF treatments had few occurrences of concentrations > 100 ppb, and
19 consequently, the relative importance of peak concentrations > 100 ppb cannot be reliably
20 tested with any degree of statistical power; and (2) the selected cases showing no significant
21 yield loss for NF do not closely match the cases showing significant yield loss for NF in
22 terms of environmental conditions, soil and nutrient condition, exposure periods and
23 durations, sites, species and cultivars, and harvests (for clover studies). Because these
24 factors interact with O3 effects in determining plant response, differences in these factors
25 confound the comparison of frequency distributions between cases having significant and
26 non-significant yield losses for NF. Another concern is the inconsistency of harvests for the
27 two clover studies conducted at Raleigh, NC, in 1984 and 1985, which have six and seven
28 harvests, respectively (Heagle et al., 1989), not 12 as reported by Krupa et al. (1993).
29 Reich (1987) reviewed 44 studies on 45 species to study the effects of O3 on net
30 photosynthesis (Pn) and growth of crops and tree species. Plants responded differently to
31 equivalent total exposures (i.e., SUMOO) when peak concentrations differed widely, with
December 1993 5-140 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-14. SUMMARY OF PERCENTILES FOR OZONE MONITORING SITES IN 1989
1
BUT WITH SECOND
» » *. -•- .A.JL f. .m, A. *.•_(. m^ m_»_L V M. ^S ,i » M. JL. •*- .M_M.m. J -*f_^ ^V AV«rA ^ -*. M.M^ t*J WAV .», W T J. .B_M_* *t^l M-J ^** **~W«T I^MAJLl. MM.
HOURLY MAXIMUM CONCENTRATION .> 0.125 ppm
t . Maximum Number of Observ.
\O Percentiles
S
Ut
i
£
O
>
M
i.
O
o
H
O
d
AIRS Site
060010003
060371301
060374002
060375001
060830008
060830010
060833001
090010113
090091123
220191003
220330003
220330004
220331001
220470002
220770001
230052003
471630009
481410027
481990002
482010024
482010062
482011034
482011037
490350003
490353001
Name
LIVERMORE.CA
LYNWOOD,CA
LONG BEACH, CA
HAWTHORNE, CA
SANTA BARBARA, CA
SANTA BARBARA, CA
SANTA BARBARA CO, CA
BRIDGEPORT, CT
NEW HAVEN, CT
WESTLAKE, LA
BATON ROUGE, LA
BATON ROUGE, LA
E BATON ROUGE, LA
IBERVILLE PAR, LA
NEW ROADS, LA
CAPE ELIZABETH, ME
KINGSPORT, TN
EL PASO, TX
KOUNTZE, TX
HARRIS CO, TX
HOUSTON, TX
HOUSTON, TX
HOUSTON, TX
SALT LAKE CO, UT
SALT LAKE CITY, UT
Min.
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.000
.001
.001
.000
.000
.000
.000
.000
.000
.000
.000
10
.000
.000
.010
.000
.010
.010
.010
.002
.003
.003
.001
.002
.003
.005
.001
.017
.001
.010
.000
.000
.000
.000
.000
.001
.002
30
.010
.010
.020
.020
.020
.020
.020
.011
.010
.013
.009
.008
.012
.014
.011
.027
.005
.020
.010
.010
.010
.010
.010
.008
.014
50
.030
.020
.020
.030
.030
.030
.030
.022
.019
.022
.021
.016
.022
.023
.021
.034
.017
.030
.020
.020
.020
.010
.010
.029
.029
70
.040
.030
.030
.040
.040
.040
.040
.033
.029
.033
.034
.028
.034
.034
.033
.042
.032
.040
.030
.030
.030
.030
.030
.042
.041
90
.050
.050
.050
.060
.050
.050
.050
.048
.045
.052
.059
.047
.056
.057
.052
.055
.054
.050
.050
.060
.050
.050
.050
.056
.053
95
.060
.070
.060
.060
.060
.060
.060
.059
.056
.061
.069
.057
.066
.068
.062
.064
.062
.060
.060
.070
.070
.060
.060
.062
.061
99
.090
.100
.080
.080
.080
.080
.080
.091
.091
.082
.094
.078
.092
.093
.083
.093
.078
.080
.080
.110
.110
.100
.110
.083
.079
Max.
.140
.140
.160
.190
.190
.220
.140
.156
.156
.137
.168
.138
.171
.149
.141
.146
.125
.260
.130
.230
.170
.220
.250
.125
.140
Uncorrected
SUM06 (ppm-h)
17.0
18.1
13.6
18.1
17.1
13.3
12.3
16.5
12.9
12.2
17.4
8.4
14.4
15.9
12.0
16.7
13.4
14.9
10.6
19.2
16.8
14.0
16.3
17.4
13.0
Over 7-mo
Period
5,067
4,793
4,876
4,894
4,823
4,663
5,077
4,865
4,502
4,811
4,964
4,791
4,890
5,040
4,964
4,627
4,252
4,484
4,630
4,728
4,600
4,595
4,729
4,585
4,544
-------
TABLE 5-15. SUMMARY OF PERCENTILES FOR OZONE MONITORING SITES IN 1989
(APRIL THROUGH OCTOBER) WITH A MAXIMUM THREE-MONTH SUM06 VALUE > 24.4 ppm-h
1
fa
BUT WITH SECOND HOURLY MAXIMUM CONCENTRATION < 0.125 ppm
»T* Percentiles
S
W
i
§
d
§>
3
*p
d
> j
2!
H
O
0
O
AIRS Site
040132004
060070002
060170009
060430004
060710006
061011002
120094001
170190004
170491001
180970042
240030014
240053001
310550032
350431001
360310002
370270003
370810011
371470099
390030002
391510016
420070003
420770004
470090101
510130020
510610002
511870002
550270001
551390007
Name
SCOTTSDALE, AZ
CHICO, CA
SOUTH LAKE TAHOE, CA
YOSEMITE NP, CA
SAN BERNARDINO CO, CA
YUBA CITY, CA
COCOA BEACH, FL
CHAMPAIGN, IL
EFFDSfGHAM CO, DL
INDIANAPOLIS, IN
ANNE ARUNDEL, MD
ESSEX, MD
OMAHA, NE
SANDOVAL CO, NM
ESSEX CO, NY
LENOIR, NC
GUILFORD CO, NC
FARMVILLE, NC
ALLEN CO, OH
CANTON, OH
NEW BRIGHTON, PA
ALLENTOWN, PA
SMOKY MT NP, TN
ARLINGTON CO, VA
FAUQUIER CO, VA
SHEN NP (DKY RDG), VA
HORICON, WI
OSHKOSH, WI
Min.
.000
.000
.000
.000
.000
.000
.002
.000
.000
.001
.000
.000
.002
.000
.016
.000
.004
.000
.000
.000
.000
.000
.000
.000
.000
.004
.002
.002
10
.006
.010
.020
.008
.020
.000
.017
.008
.009
.006
.006
.002
.021
.010
.033
.007
.010
.010
.007
.008
.008
.003
.025
.001
.009
.027
.019
.016
30
.018
.020
.030
.022
.040
.020
.024
.020
.023
.021
.021
.010
.030
.020
.042
.019
.023
.023
.022
.019
.021
.016
.036
.010
.021
.037
.029
.028
50
.031
.030
.040
.035
.050
.030
.032
.029
.036
.034
.032
.024
.037
.030
.050
.032
.034
.034
.032
.030
.032
.028
.044
.023
.033
.045
.037
.038
70
.045
.040
.050
.049
.060
.040
.042
.039
.046
.046
.045
.038
.047
.040
.056
.045
.046
.044
.043
.042
.043
.039
.053
.037
.045
.054
.047
.048
90
.062
.060
.060
.065
.070
.060
.059
.065
.063
.063
.064
.059
.062
.060
.067
.062
.063
.062
.060
.060
.062
.060
.065
.059
.061
.065
.062
.063
95
.071
.070
.070
.072
.080
.070
.068
.072
.070
.072
.073
.069
.067
.060
.073
.067
.070
.070
.068
.070
.070
.070
.070
.071
.069
.071
.070
.070
99
.084
.080
.080
.083
.090
.080
.077
.078
.081
.085
.090
.089
.075
.070
.086
.078
.083
.083
.086
.088
.087
.087
.081
.088
.084
.082
.088
.084
Max.
AQ7
.100
.100
.111
.100
.100
.094
.088
.104
.103
.120
.121
.098
.090
.106
.092
.113
.100
.107
.110
.102
.102
.098
.116
.122
.100
.111
.121
Uncorrected
SUM06 (ppm-h)
— _
33.5
44.8
37.6
70.5
29.0
28.7
32.0
25.3
25.4
25.5
25.2
24.9
25.1
45.6
25.8
27.7
26.4
24.5
26.3
29.4
25.1
35.9
25.7
24.6
59.0
24.6
27.9
Over 7-mo
Period
5,070
4,690
4,768
4,853
4,856
4,623
5,012
5,091
4,600
4,592
4,360
5,028
4,160
5,059
4,070
4,806
4,853
4,833
4,854
4,875
5,055
5,040
4,764
5,029
5,050
4,454
4,142
4,206
-------
1 greater loss of net photosynthesis (Pn) for increasing concentrations (Figure 5-15). Short-
2 term, high concentrations above 0.40 ppm (e.g., 0.50 ppm for 8 h) caused rapid and
3 significant reduction in Pn. Longer-term exposures for weeks to lower concentrations had a
4 significant effect on Pn; the observed reductions were less severe than at the higher
5 concentrations. Based on short-term, high concentration studies, SUMOO alone was an
6 inadequate descriptor of exposure for predicting response. However, for assessing the effects
7 of long-term, low concentrations typical of ambient condition, SUMOO may be adequate, as
8 the response of field-grown plants to SUMOO was roughly linear. SUMOO explained much,
9 although not all, of the variation in net photosynthesis and growth of conifers, hardwood
10 trees, and agricultural crops (Figures 5-16 through 5-18). Unexplained variation can be
11 attributed to biological variation, inherent experimental error, experimental conditions, and
12 differences in O3 uptake. Imputed O3 uptake calculated as the product of SUMOO and mean
13 diffusive conductance (ks) for each species better correlated with Pn and growth than
14 SUMOO.
15 Kickert and Krupa (1991) criticized Reich's findings on the basis of insufficient
16 reporting of statistical model parameters, possible non-normality of Pn and growth variables,
17 exclusion of ks terms for imputing O3 uptake for each species, and with no implication for
18 any individual plant species. However, Reich's synthesis of Pn and growth, using the
19 SUMOO index, would not necessarily be invalidated by non-normality of the variables and
20 provides evidence on the importance of peak concentrations and duration in eliciting a
21 response. Reich's use of a mean diffusive conductance to impute O3 uptake is questionable
22 as leaf diffusive conductance measurements vary with time of day, season, and environmental
23 condition. Consequently, numerous measurements of conductance are required to weight
24 hourly O3 concentrations to calculate O3 uptake over the growth of a plant.
25 Pye (1988) reviewed 15 studies on 26 seedling species and found reductions in biomass
26 response increased with SUMOO (Figure 5-19). Seasonal sum of hourly concentrations
27 values ranged from 4 to 297 ppm-h. However, there was substantial variation in response.
28 Pines, poplars, sycamore, ash, and maple are relatively sensitive. Both concentration and
29 duration are important factors governing impact on growth and photosynthesis, but they
30 probably are not equally important. The biomass data suggest a nonlinear response to
December 1993 5.143 DRAFT-DO NOT QUOTE OR CITE
-------
-------
40 80 120
Dose (ppm-h)
160
10 20 30
Uptake (mg/cm2)
40
Figure 5-16. Percent reduction in net photosynthesis and biomass growth of coniferous
species in relation to (a) total exposure (SUMOO) and (b) estimated total
ozone uptake (Reich, 1987).
1 fumigation and the presence of convexity of response implies that for similar mean
2 O3 exposures, damage will be greater when O3 concentrations are more variable.
3 There is limited information for assessing the relative performance of exposure indices
4 for relating to vegetation effects. Lefohn et al. (1992a) reported that it was not possible to
5 differentiate among the SUMOO, SUM06, W126, and SUM08 exposure indices because the
6 indices were highly correlated with one another in the experiment (Figure 5-20). However,
7 results based on biological experiments, reported by Musselman et al. (1983) and Hogsett
8 et al. (1985a) have shown that different exposure regimes with similar SUMOO values
9 resulted in those exposures experiencing capture peak concentrations exhibiting the greater
10 effects. The authors demonstrated that plants exposed to variable concentrations showed
December 1993
5-145
DRAFT-DO NOT QUOTE OR CITE
-------
20
20 40 60 80 100 120
Dose (ppm-h)
10 20 30 40
Uptake (mg/cm2)
so
Figure 5-17. Percent reduction in net photosynthesis and biomass growth of hardwood
species in relation to (a) total exposure (SUMOO) and (b) estimated total
ozone uptake (Reich, 1987).
1 greater effect on plant growth than those exposed to a fixed or daily peak concentration of
2 equal SUMOO but lower peak concentrations.
3 Building upon the above cited results that biologically showed that the importance of
4 the higher hourly average concentrations, Lefohn et al. (1989) concluded that the SUMOO
5 index did not appear to perform adequately. Using air quality data, Lefohn et al. (1989)
6 showed that die magnitude of the SUMOO exposure index was largely determined by the
7 lower hourly average concentrations instead of the biologically relevant higher hourly
8 average concentrations (Figure 5-21). Figure 5-21 illustrates that the slope of the curve that
9 described the cumulative frequency for the SUMOO index (referred to as TOTDOSE) was
10 greater than the slope of the curve for the W126 index until approximately 0.06 ppm.
December 1993
5-146
DRAFT-DO NOT QUOTE OR CITE
-------
<£ -20
c
o
O
-40
-60
-80
C
20
0
-20
-40
-60
-80
10
Crops
20
0
-20 \
-40
-60
-80
Crops
20 40 60 80 100 120
Dose (ppm-h)
10 20 30 40
Uptake (mg/cm2)
so
Figure 5-18. Percent reduction in net photosynthesis and biomass growth of agricultural
crops in relation to (a) total exposure (SUMOO) and (b) estimated total
ozone uptake (Reich, 1987).
1 Thereafter, the reverse was true. This occurred because the W126 index weighted the higher
2 concentrations more heavily than the lower ones, while the TOTDOSE index did not.
3 Supplementing the results in Lefohn et al. (1989), Lefohn et al. (1992a), using loblolly
4 pine data exposed at Auburn, AL, to varying levels of O3 over 555 days (Qiu et al., 1992)
5 reported that the magnitude of the SUMOO values in the CF chamber, although experiencing
6 hourly average values greater than those at the South Pole or Pt. Barrow, AK, was about
7 50% less than the SUMOO values experienced at the South Pole and Pt. Barrow, AK.
8 In a similar analysis using ambient data, Lefohn et al. (1992a) identified a separate set
9 of ambient sites that experienced SUMOO values similar to those of the ambient treatments at
December 1993
5-147 DRAFT-DO NOT QUOTE OR CITE
-------
50
25
5-25
c
6
-75
-100
!F
ft
0 25 50 75 100 125 150 175 275 300
Ozone Exposure (nil1 hr)
Figure 5-19. Percent reduction in biomass growth of tree seedlings in relation to total
exposure (Pye, 1988).
1 Auburn; these ambient sites experienced fewer hourly concentrations above 0.07 ppm than
2 did the ambient chambers. Similar to the results cited above, the authors noted that the
3 magnitude of the SUMOO index was unable to capture the occurrence of the higher hourly
4 average concentrations in the ambient treatments. The authors indicated that the SUMOO
5 index was inadequate because of the observed inconsistences of the SUMOO value between
6 chambers and selected monitoring sites.
7 When taken by themselves, the importance of these findings may be debatable because
8 the clean sites are not representative of loblolly growing regions and there is no
9 substantiating evidence of differing effects at these levels. However, the coupling of the air
10 quality considerations, as described by Lefohn et al. (1989, 1992a), with the biological
December 1993
5-148
DRAFT-DO NOT QUOTE OR CITE
-------
1
£,0.8
CM
fo.e
|°-4
< 0.2
°(
a)
o 2*0.8
CM
T3
o £°'6
o ®
x w 0.4
X 3
x < 0.2
c)
o
o
o
X
X
X
) 100 200 300 400 500 600 700 °0 100 200 300 400 500 600 70
SumO (ppm-h)
Sum06 (ppm-h)
1
~0.8:
CM
•o
C0.6'
^J
§0.4
3
xT
<0.2
f\
b)
£-°-8>
CM
s, ,
^0.6'
o ~~
•o
|0.4
x 3
^
<0.2
d)
o
o
i
X
X
'
'0 100 200 300 400 500 600 700
w126 (ppm-h)
'0 100 200 300 400 500 600 700
SumOS (ppm-h)
Figure 5-20. Reduction in volume production of loblolly pine seedlings (family 91) in
relation to four exposure indices (a-d) (Lefohn et al., 1992a).
1 findings reported by Musselman et al. (1983) and Hogsett et al. (1985a), builds a consistent
2 picture that the SUMOO index does not properly describe the occurrence of the higher hourly
3 average concentrations.
December 1993
5-149
DRAFT-DO NOT QUOTE OR CITE
-------
^
0)
1
«v
0)
3s
3
E
=»
O
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
n n
\j.\j
I l t I I I I I I I I I
A W126 0 • A A A
* Both * A
• A A
A
-
A
•
A
~~ •
A
^
"A
•
A
• A
• A
A
I 1 I I I I I I I I I I
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12
Level (ppm)
Figure 5-21. A comparison between the resulting cumulative frequencies for the
exposure parameters (a) sum of all hourly average concentrations (SUMOO)
and (b) the sigmoidally weighted integrated exposure index, W126. The
ozone data were collected in 1981 at a site located in the Mark Twain
National Forest, Missouri. The U.S. EPA AIRS site number is 291230001
(Lefohn et al., 1989).
1 The selection of the weighting scheme of hourly O3 concentrations for use in
2 characterizing plant exposure and its relationship to plant response is difficult because of
3 limitations in the data. The results of studies based on analysis of experimental and air
4 quality data substantiate O3 uptake as the determining factor of response. Stomata controls
5 the rate of O3 uptake. Stem et al. resistance is influenced by ambient conditions surrounding
6 the plant, as well as other factors. Concentration has been identified as important in
7 predicting response and concentrations as low as 50 ppb may contribute to response. Several
8 lines of evidence suggest that the peak-weighted, cumulative indices yield better predictions
December 1993
5-150
DRAFT-DO NOT QUOTE OR CITE
-------
1 of yield loss than the mean and peak indices. The optimal weighting function to describe
2 plant exposure has not been determined but should give greater weight to the peak
3 concentrations. Because the NCLAN studies used ambient-added exposures that have
4 considerably more concentrations above 0.10 ppm than ambient conditions in polluted
5 monitoring sites, one must be careful in the selection of an exposure index based on
6 statistical fit.
7
8 5.5.3 Summary
9 The effects of O3 on individual plants and factors that modify plant response to O3 are
10 complex and vary with factors, such as species, environmental conditions, and soil and
11 nutrient conditions. Because of the complex effect of O3 and its interactions with physical
12 and genetic factors that influence response, it is difficult to develop a measure of exposure
13 that relates well with plant response based on experimental data. At best, experimental
14 evidence of the impact of O3 on biomass production can indicate important factors of
15 O3 exposure that modify plant response, which become considerations in developing an
16 exposure index.
17 Considerable evidence of the primary mode of action of O$ on plants (injury to proteins
18 and membranes, reduction in photosynthesis, changes in allocation of carbohydrate, and early
19 senescence), which ultimately lead to reductions in biomass production, identifies O3 uptake
20 as an important factor (see Section 5.2). Ozone uptake is controlled by canopy conductance,
21 stomatal conductance, and O3 concentration outside the leaf (see Figure 5-2). Any factor
22 that will affect stomatal conductance (e.g., light, temperature, humidity, soil and atmospheric
23 chemistry and nutrients, time of day, phenology, and biological agents) will affect Oj uptake
24 and, consequently plant response.
25 Evidence from studies that applied two or more different exposure regimes substantiate
26 the importance of daytime peak concentrations, respite time, and phenology in eliciting a
27 response. Ozone effects on plants exposed to two (or more) regimes having equal total
28 exposure were greater for exposures experiencing the higher peak concentrations, respite
29 time of 2 to 6 days, or peak concentrations during period of maximum leaf expansion, This
30 conclusion is consistent with the mode of action of O3 on plants and with the conclusions in
31 the previous Environmental Protection Agency criteria document (U.S. Environmental
December 1993 5.151 DRAFT-DO NOT QUOTE OR CITE
-------
1 Protection Agency, 1986) and its Supplement (U.S. Environmental Protection Agency, 1992)
2 and is equally true for seedlings based on a recent seedling study (Hogsett and Tingey,
3 1990).
4 Further, the biochemical mechanisms, discussed in Section 5.2, describe the mode of
5 action of O3 on plants as the culmination of a series of physical, biochemical, and
6 physiological events leading to alterations in plant metabolism. Ozone-induced injury is
7 cumulative, resulting in net reductions in photosynthesis, changes in allocation of
8 carbohydrate, and early senescence, which lead to reductions in biomass production
9 (Section 5.2). Increasing O3 uptake will result in increasing reductions in biomass
10 production. Retrospective analyses comparing replicate studies of equal and varying
11 durations are in general agreement that the ultimate impact of 03 on plant response depends
12 upon the integration of repeated peak concentrations during the ecological period of growth.
13 The optimum exposure index that relates well with plant response should incorporate
14 the factors (directly or indirectly) described above; unfortunately, such an index has not yet
15 been identified. At this time, exposure indices that weight the hourly O3 concentrations
16 differentially appear to be the best candidates for relating exposure with predicted plant
17 response. Peak concentrations in ambient air occur primarily during daylight thus, these
18 indices, by providing preferential weight to the peak concentrations, give greater weight to
19 the daylight concentrations rather than nighttime concentrations (when stomatal conductance
20 is minimal).
21 Studies reported in the literature show that when O3 is the primary source of variation
22 in response, year-to-year variations in plant response are minimized by the peak-weighted,
23 cumulative exposure indices. The data suggest that the exposure indices that cumulate hourly
24 O3 concentrations over the growing season and give preferential weight to the peak
25 concentrations are biologically appropriate as characterizations and representations of plant
26 exposure. Generally, the peak-weighted, cumulative indices rekte well with plant response
27 and order the treatment means in monotonically decreasing fashion with increasing exposure,
28 based on studies that apply two or more types of exposure regimes and when combining data
29 from replicate studies of the same species. No studies have been designed specifically to
30 evaluate the adequacy of the peak-weighted, cumulative indices. Consequently, it is not
31 possible to discriminate among the various peak-weighted, cumulative indices based on
December 1993 5-152 DRAFT-DO NOT QUOTE OR CITE
-------
1 experimental data. Functional weighting approaches including an allometric, sigmoid, or
2 threshold weighting have been suggested and, in earlier retrospective studies compared, but
3 there is no evidence to favor one approach over the other on the basis of statistical fits to the
4 data.
5 The study of Fuhrer et al. (1992) illustrates some of the limitations in applying
6 exposure indices. The study is significant for its use of the mean O3 flux in minimizing the
7 year-to-year variation in response when combining replicate studies, indicating the
8 importance of environmental conditions in quantifying the relationship between O3 exposure
9 and plant response.
10 The factors such as respite time, temporal variation, phenology, canopy structure,
11 physiological processes, environmental conditions, and soil and nutrient conditions, are
12 important in determining the impact of O3 on crops and trees but are not well understood and
13 interact with concentration and duration in different fashions depending upon species. Ozone
14 uptake integrates these factors with atmospheric conditions, relates well with plant response,
15 but is difficult to measure. Empirical functions to predict stomatal conductance have been
16 developed for particular species (e.g., Losch and Tenhunen, 1981) but have not been used to
17 estimate O3 uptake or used in development of exposure indices.
18 The exposure-response studies in the published literature support the conclusion that
19 O3 effects are cumulative and peak concentrations are more important than lower
20 concentrations in determining the magnitude of plant response. The peak-weighted,
21 cumulative indices appear to have major advantages over the mean (e.g., 7-h seasonal mean),
22 peak indices (e.g., 2ndHDM), and the index that cumulates all hourly average concentrations
23 (i.e., SUMOO). Crop yield loss and biomass reduction are better estimated using the peak-
24 weighted, cumulative indices than the 2ndHDM index; when duration of exposure is taken
25 into consideration, peak-weighted cumulative indices perform better than the seasonal mean
26 indices. In addition, results have been published to indicate that the SUMOO index does not
27 adequately relate exposure with biological effects because the index focuses on the lower
28 hourly average concentrations.
29 For predicting the effects of O3 on vegetation under ambient conditions using
30 experimental exposure-response models, the types of exposure regimes used in the
31 experiments should be taken into consideration. For example, NCLAN experiments
December 1993 5.153 DRAFT-DO NOT QUOTE OR CITE
-------
1 contained peak hourly average concentrations in their regimes. Any exposure index, based
2 on the NCLAN experiments, should take into consideration the presence of these peak
3 concentrations. By doing so, the situation may be avoided where two sites, which experience
4 two distinct distributions of high hourly average concentrations, but have the same value of
5 cumulation (e.g., same SUM06 or W126 value), exhibit differing biological effects.
6 Future research needs to focus on: (1) validation of the application of exposure-
7 response relationships based on chambered studies to ambient conditions; (2) evaluation of
8 peak-weighted, cumulative indices for quantifying the relationship of O3 exposure to ensuing
9 plant response; (3) a better understanding of the effects of O3 and its interactions with
10 environmental conditions, soil chemistry, and nutrient conditions; (4) development of
11 mechanistic models of stomatal conductance for use in weighting O3 concentrations in
12 ambient air; and (5) more multi-year exposure studies on trees to predict the long-term
13 effects on O3 of biomass production.
14
15
16 5.6 EXPOSURE-RESPONSE OF PLANT SPECIES
17 5.6,1 Introduction
18 Determining the response of plants to O3 exposures continues to be a major challenge.
19 The effects of vegetational exposure are usually evaluated by exposing various plant species
20 under controlled experimental conditions such as those discussed in Section 5.2. Plant
21 responses are influenced not only by the biochemical and physiological changes that may
22 occur within the plant after O3 entry (Section 5.3, Mode of Action, see also Figure 5-5) but
23 also by the many factors that modify plant response (Section 5.4). Of the factors discussed
24 in Section 5.4, those that are most likely to apply under controlled experimental conditions
25 are the genetic makeup and age of the plant. Compensatory responses (Section 5.3.4.2) will
26 also influence plant response. This section will analyze, summarize, and evaluate what is
27 known about the response of various plant species or cultivars, either as an individuals or in
28 populations to O3 exposure. Species as populations will only be considered in the case of
29 pasture grasses, or forage mixes, which commonly occur as mixed stands. Emphasis will be
30 placed on those studies conducted since the publication of the previous criteria document
31 1986 (U.S. Environmental Protection Agency, 1986). Much of the discussion of vegetational
December 1993 5-154 DRAFT-DO NOT QUOTE OR CITE
-------
1 response to O3 exposure in the current document is based on the conclusions of both the
2 1978 and 1986 documents (U.S. Environmental Protection Agency, 1978, 1986). Therefore,
3 to provide a basis for understanding the effects presented below, the conclusions of the two
4 documents are summarized.
5 Finally, the results of O3 exposure-response presented in this section must be related to
6 one or more assessment endpoints. Historically, the dollar value of lost production was the
7 endpoint of interest, however, other endpoints (e.g., biodiversity, habitat, aesthetics,
8 recreation) must be considered now, particularly as the impacts of O3 on long-lived species
9 of ecological importance are evaluated (Tingey et al., 1990).
10
11 5.6.2 Summary of Conclusions from the Previous Criteria Documents
12 The experimental data presented in the 1978 and 1986 criteria documents dealt with the
13 effects of O3 primarily on agricultural crops species (U.S. Environmental Protection Agency,
14 1978, 1986). The chapter on vegetational effects in the 1978 criteria document (U.S.
15 Environmental Protection Agency, 1978) emphasized visible injury and growth effects;
16 however, the growth effects were not those that affected yield. This emphasis was dictated
17 by the kind of data available at the time. The document also presented data dealing with the
18 response of the San Bernardino ecosystem to O3. This information was also discussed in the
19 1986 document (U.S. Environmental Protection Agency, 1986). It remains the best study of
20 ecosystem responses to O3 stresses (see Section 5.7).
21 The 1986 document emphasized the fact that though foliar injury on vegetation is one
22 of the earliest and most obvious manifestations of O3 exposure, the effects of exposure are
23 not limited to visible injury. Plant foliage is only the primary site of plant response to
24 O3 exposures. Significant secondary effects, include reduced growth, both in foliage and
25 roots. Impacts range from reduced plant growth and decreased yield to changes in crop
26 quality and alterations in plant susceptibility to biotic and abiotic stresses. Also, the 1986
27 document noted that O3 exerts a phytotoxic effect only if a sufficient amount reaches
28 sensitive sites within the leaf (see Section 5.3). Ozone injury will not occur if (1) the rate of
29 uptake is low enough that the plant can detoxify or metabolize O3 or its metabolites; or
30 (2) the plant is able to repair or compensate for the effects (Tingey and Taylor, 1992; U.$.
31 Environmental Protection Agency, 1986). Cellular disturbances that are not repaired or
December 1993 5-155 DRAFT-DO NOT QUOTE OR CITE
-------
1 compensated are ultimately expressed as visible injury to the leaf or as secondary effects that
2 can be expressed as reduced root growth, or reduced yield of fruits or seeds, or both. Ozone
3 would be expected to reduce plant growth or yield if (1) it directly impacts the plant process
4 e.g., photosynthesis that was limiting plant growth; or (2) it impacts another step sufficiently
5 so that it becomes the step limiting plant growth (U.S. Environmental Protection Agency,
6 1986; Tingey, 1977). Conversely, if the process impacted is not or does not become rate-
7 limiting, O3 will not limit plant growth. These conditions also suggest that there are
8 combinations of O3 concentration and exposure duration that a plant can experience that will
9 not result in visible injury or reduced plant growth and yield. Indeed, numerous studies have
10 demonstrated this fact. This information is still pertinent today (Section 5.3)
11 Ozone can induce a diverse range of effects beginning with individual plants and then
12 proceeding to plant populations and ultimately communities. The effects may be classified as
13 either (1) injury or (2) damage. Injury encompasses all plant reactions such as reversible
14 changes in plant metabolism (e.g., altered photosynthesis), leaf necrosis, altered plant
15 quality, or reduced growth that does not impair yield or the intended use or value of the
16 plant (Guderian, 1977). Thus, for example, visible foliar injury to ornamental plants,
17 detrimental responses in native species, while reductions in fruit and grain production in
18 cultivated plants are all considered damage or yield loss. Although foliar injury is not
19 always classified as damage, its occurrence indicates that phytotoxic concentrations of O3 are
20 present and, therefore, studies should be conducted to assess the risk to vegetation.
21 The concept of limiting values used to summarize visible foliar injury in the 1978
22 document was also considered valid in the 1986 document (U.S. Environmental Protection
23 Agency, 1978, 1986). Jacobson, (1977) developed limiting values by reviewing the scientific
24 literature and identifying the lowest concentration and exposure duration reported to cause
25 visible injury to a variety of plant species. Expressed in another way, limiting values were
26 concentrations and durations of exposure below which visible injury did not occur.
27 A graphical analysis presented in both of the previous documents indicated the limit for
28 reduced plant performance was an exposure to 0.05 ppm for several hours per day for
29 greater than 16 days. Decreasing the exposure period to 10 days increased the concentration
30 required to cause injury to 0.1 ppm, and a short, 6 day exposure further increased the
31 concentration to cause injury to 0.30 ppm.
December 1993 5-156 DRAFT-DO NOT QUOTE OR CITE
-------
1 By 1986, a great deal of new information concerning the effects of O3 on the yield of
2 crops plants had become available, both through the EPA National Crop Loss Assessment
3 Network (NCLAN) and the results of research funded by other agencies. The NCLAN
4 project was initiated by the Environmental Protection Agency in 1980 primarily to improve
5 estimates of yield loss in the field and to estimate the magnitude of crop losses caused by
6 O3 (Heck et al., 1982). The primary objectives were:
7
8 1. To define the relationships between yields of major agricultural crops and
9 O3 exposure as required to provide data necessary for economic assessments and the
10 development of National Ambient Air Quality Standards;
11
12 2. To assess the national economic consequences resulting from the exposure of major
13 agricultural crops to O3;
14
15 3. To advance understanding of the cause and effect relationships that determine crop
16 responses to pullutant exposures.
17
18 The cultural conditions used in the NCLAN studies approximated typical agronomic
19 practices. The methodology used in these studies is described in Section 5.2.
20 Yield loss in the 1986 document was defined as an impairment in the intended use of
21 the plant. This concept included reductions in aesthetic values, the occurrence of foliar
22 injury (changes in plant appearance), and losses in terms of weight, number, or size of the
23 plant part that is harvested. Yield loss may also include changes in physical appearance,
24 chemical composition, or the ability to withstand quality storage; which collectively are
25 termed crop quality. Losses in aesthetic values are difficult to quantify. Foliar injury
26 symptoms can substantially reduce the marketability of ornamental plants or crops in which
27 the foliage is the plant part (e.g., spinach, lettuce, cabbage) and constitute yield loss with or
28 without concomitant growth reductions. At that time (1986) most studies of the relationship
29 between yield loss and O3 concentration focused on yields as measured by weight of the
30 marketable organ of the plant.
31 Open-top field chamber studies conducted to estimate the impact of O3 on the yield of
32 various crops species, e.g., the NCLAN program, were grouped into two types, depending
33 on the experimental design and statistical methods used to a analyze the data: (1) studies that
34 developed predictive equations relating O3 exposure to plant response, and (2) studies that
December 1993 5_157 DRAFT-DO NOT QUOTE OR CITE
-------
1 compared discrete treatment level to a control. The advantage of the regression approach is
2 that exposure-response models can be used to interpolate results between treatment levels (see
3 Section 5.2.2).
4 Using NCLAN data as an example of plant response, the O3 concentrations that could
5 be predicted to cause 10 or 30% yield loss were estimated using the Weibull function
6 (Table 5-16; Table 6-19, U.S. Environmental Protection Agency, 1986). The data in
7 Table 5-16 are based on yield-response functions for 38 species or cultivars developed from
8 studies using open-top chambers. Review of that data indicated that ten percent yield
9 reductions could be predicted for 57% of the species or cultivars when 7-h seasonal mean
10 concentrations were below 0.05 ppm, for 35 % when seasonal mean concentrations were
11 between 0.04 and 0.05 ppm, but only 19% required a 7-h seasonal mean concentrations in
12 excess of 0.08 ppm to suffer a 10% loss in yield. Furthermore, approximately 11% of the
13 38 species or cultivars would be expected to have a yield reduction of 10% loss at 7-h
14 seasonal mean concentrations below 0.05 ppm, suggesting that these plants are very sensitive
15 to O3.
16 Grain crops were apparently less sensitive than the other crops. The data also
17 demonstrate that the sensitivity within species may be as great as difference between species.
18 For example, at 0.04 ppm O3, estimated yield losses ranged from 2 to 15% in soybean and
19 from 0 to 28% in wheat. Year to year variations in plant response were also observed
20 during the studies.
21 Discrete treatments were used to determine yield loss in some studies. These
22 experiments were designed to test whether specific O3 treatments were different from the
23 control rather than to develop exposure-response equations and the data were analyzed using
24 analyses of variance. When summarizing these studies using discrete treatment levels, as
25 opposed to the variable concentrations used in NCLAN, the lowest O3 concentration that
26 significantly reduced yield was determined from analyses done by the authors. Frequently,
27 the lowest concentration used in the study was the lowest concentration reported to reduce
28 yield; hence it was not always possible to estimate a no-effect exposure concentration.
29 In general, the data indicated that O3 concentrations of 0.10 ppm (frequently the lowest
30 concentration used in the studies) for a few hours per day for several days to several weeks
December 1993 5-158 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-16. ESTIMATES OF THE PARAMETERS FOR FITTING THE WEffiULL
MODEL USING THE 7-HOUR SEASONAL MEAN OZONE CONCENTRATIONS*'*
Parameters for Weibull Model
Concentration for
Predicted Yield
Losses of:
Crop
LEGUME CROPS
Soybean, Corsoy
Soybean, Davis (81)
Soybean, Davis (CA-82)"
Soybean, Davis (PA-82)C
Soybean, Essex (81)
Soybean, Forrest (82-1)
Soybean, Williams (81)
Soybean, Williams (82-1)
Soybean, Hodgson
Bean, Kidney (FP)f
Peanut, NC-6
GRAIN CROPS
Wheat, Abe (82)
Wheat, Arthur 71 (82)
Wheat Roland
Wheat, Vona
Wheat, Blueboy n (T)
Wheat, Coker 47-27 (T)
Wheat, Holly (T)
Wheat, Oasis (T)
Corn, PAG 397
Corn, Pioneer 3780
Corn, Coker 16 (T)
Sorghum, DeKalb-28
Barley, Poco
FIBER CROPS
Cotton, Acala SJ-2 (81-1)
Cotton, Acala SJ-2 (82-1)
Cotton, Stoneville
HORTICULTURAL
CROPS
Tomato, Murrieta (81)
Tomato, Murrieta (82)
Lettuce, Empire (T)
Spinach, America (T)
Spinach, Hybrid (T)
Spinach, Viroflay (T)
Spinach, Winter Bloom (T)
A
a
2785.00
5593.00
4931.00
4805.00
4562.00
4333.00
4992.00
5884.00
2590.00
2878.00
7485.00
5363.00
4684.00
5479.00
7857.00
5.88
5.19
4.95
4.48
13968.00
12533.00
240.00
8137.00
1.99
5546.00
5872.00
3686.00
32.90
32.30
1245.00
21.20
36.60
41.10
20.80
A
a
0.133
0.128
0.12/
0.103
0.187
0.171
0.211
0.162
0.138
0.120
0.111
0.143
0.148
0.113
0.053
0.175
0.171
0.156
0.186
0.160
0.155
0.221
0.296
0.205
0.199
0.088
0.112
0.142
0.082
0.098
0.142
0.139
0.129
0.127
A
c
1.952
0.872
2.144
4.077
1.543
2.752
1.100
1.577
1.000
1.171
2.249
2.423
2.154
1.633
1.000
3.220
2.060
4.950
3.200
4.280
3.091
4.460
2.217
4.278
1.228
2.100
2.577
3.807
3.050
1.220
1.650
2.680
1.990
2.070
CP
0.022
0.025
0.019
0.019
0.014
0.017
0.014
0.017
0.017
0.019
0.025
0.023
0.023
0.023
0.022
0.030
0.030
0.030
0.030
0.015
0.015
0.020
0.016
0.020
0.018
0.012
0.026
0.012
0.012
0.043
0.024
0.024
0.024
0.024
10%d
0.048
0.038
0.048
0.059
0.048
0.076
0.039
0.045
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
30%d
0.082
0.071
0.081
0.081
0,099
0.118
0.093
0.088
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
December 1993
5-159 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-16 (cont'd). ESTIMATES OF THE PARAMETERS FOR FITTING THE
WEIBULL MODEL USING THE 7-HOUR SEASONAL MEAN OZONE
CONCENTRATIONS^
Concentration for
Predicted Yield
Parameters for Weibull Model Losses of:
Crop
Turnip, Just Right (T)
Turnip, Pur Top W.G. (T)
Turnip, Shogoin (T)
Turnip, Tokyo Cross (T)
at
10.89
6.22
4.68
15.25
a
0.090
0.095
0.096
0.094
c
3.050
2.510
2.120
3.940
CP
0.014
0.014
0.014
0.014
10%"
0.043
0.040
0.036
0.053
30%"
0.064
0.064
0.060
0.072
Data are from Heck et al. (1984b) and are based on individual plot means unless the crop name is followed by
"(T)". The "T" indicates that the parameters were based on treatment means and the data are from Heck et al.
(1983a). The parameters given in Heck et al. (1983a, 1984b) also contain the standard errors of the
parameters.
All estimates of a are in ppm. The yield is expressed as kg/Ha for all crops except barley—see weight (g per
head); tomato (both years)—fresh weight jlcg per plot); cotton—lint + seed weight (kg/ha); peanut—pod weight
(kg/Ha). In cases where the estimated c parameter is exactly 1.0, it has been bounded from below to obtain
covergence in the nonlinear model fitting routine. Parameters were estimated from data not showing the
expected Weibull form. Caution should be used in interpreting these Weibull models. Other models might
better describe the behavior observed in these experiments. For those crops whose name is followed by "(T)"
the yield is expressed as g/plant.
°The 03 concentration in the charcoal filtered chambers expressed as a 7-h seasonal mean concentration.
The 7-h seasonal mean O3 concentration (ppm) that was predicted to cause a 10 or 30% yield loss
(compared to charcoal-filtered air).
eCA and PA refer to constant and proportional O3 addition.
Only the bean data from the full plots are shown. The partial plot data are given Heck et al. (1984b).
Source: U.S. Environmental Protection Agency (1986).
1 generally caused significant yield reductions. The concentrations derived from the
2 regressions studies were based on a 10% yield loss, while in the studies using the analysis of
3 variance the 0.10 concentration frequently induced mean yield losses of 10 to 50%.
4 A chemical protectant, ethylene diurea (EDU) was used to provide estimates of yield
5 loss. The impact of O3 on yield was determined by comparing the yield data from plots
6 treated with EDU with those that were not. Studies indicated that yields were reduced by
7 18 to 41 % when ambient 63 concentrations exceeded 0.08 ppm during the day for 5 to
8 18 days over the growing season.
9 In summary, the 1986 criteria document (U.S. Environmental Protection Agency, 1986)
10 states that the following several general conclusions can be drawn from the various
11 approaches used to estimate crop loss yield; (1) Based on the comparison of data obtained
December 1993 5-160 DRAFT-DO NOT QUOTE OR CITE
-------
1 from crop yield in charcoal-filtered and unfiltered (ambient) exposures clearly indicate that
2 O3 at ambient levels are sufficiently elevated in several parts of the country to impair the
3 growth and yield of plants. This conclusion is supported by data from the chemical
4 protectant studies and extends it to other plant species; (2) Both of the above mentioned
5 approaches indicate that effects occur with only a few O3 occurrences above 0.08 ppm;
6 (3) The growth and yield data cited in the 1978 criteria document (U.S. Environmental
7 Protection Agency, 1978) indicate that several plant species exhibited growth and yield
8 effects when the mean O3 concentration exceeded 0.05 ppm for 4 to 6 h/day for at least
9 2 weeks; and (4) The data obtained from regression studies conducted to develop exposure-
10 response functions for estimating yield loss indicated that at least 50% of the species/cultivars
11 tested were predicted to exhibit a 10% yield loss at 7-h season mean O3 concentrations of
12 0.05 ppm or less. Though most of the data from the discrete treatment studies did not use
13 concentrations low enough to support the values cited above, the magnitude of yield losses
14 reported at 0.10 ppm under a variety of exposure regimes indicate that to prevent O3 effects
15 a substantially lower concentration is required.
16 The limiting values established in the 1978 were still deemed appropriate in the 1986
17 criteria document for ornamentals and certain vegetable crops where visible injury was still
18 considered the response of interest because appearance is of importance (e.g., spinach,
19 lettuce, cabbage) (U.S. Environmental Protection Agency, 1986). This remains the case
20 today.
21
22 5.6.3 Information in the Published Literature Since 1986
23 The major question to be addressed in this section is whether the conclusions of the
24 1986 criteria document summarized in the previous section, remain valid given the results of
25 research published since 1988. In particular, whether the response of plants to experimental
26 treatments at or near concentrations of 0.05 ppm (7 h seasonal mean), characteristic of
27 ambient concentrations in many areas, can be compared to a control, or reduced
28 O3 treatment to establish a potential adverse effect.
29 The 1986 (U.S. Environmental Protection Agency, 1986) made the following statement:
30 "The characterization and representation of plant exposures to O3 has been and continues to
December 1993 5.151 DRAFT-DO NOT QUOTE OR CITE
-------
1 be a major problem because research has not yet clearly identified which components of the
2 pollutant exposure cause plant response". This is still true today (See Section 5.5).
3 The aim of most air pollution research experiments have been designed to quantify the
4 relationship between pollutant exposure an agricultural crop yield. The problem is the
5 incorporation of the concentration, duration, frequency, age, genetic composition and respite
6 time into an exposure statistic or index which may be used to predict yield loss. The correct
7 exposure representation is the amount of pollutant entering the plant, not the ambient
8 concentration to which it is exposed (Taylor et al., 1982a; Tingey and Taylor, 1982). Most
9 indices were not developed from a biological basis, nor were they developed using an
10 experimental approach specifically designed to address all key factors (Lee et al., 1990).
11 A number or exposure indices have been developed in an attempt for depicting plant response
12 to O3 exposure (See Section 5.5). Much of the data in this section is evaluated using these
13 indices. For this reason several different exposure statistics are used to determine the effect
14 of an exposure on plant response. It should be remembered that the SUM06, which is used
15 more than any of the other indices, is the seasonal sum of hourly concentrations at or above
16 0.06 ppm (See Section 5.5).
17 Exposure indices calculated for each of 10 years (1982 to 1991) and two exposure
18 periods, June through August (three month) and May through September (five month) are
19 presented in Table 5-17 (modified from Tingey, et al., 1991). The monitoring data,
20 collected at non-urban sites, show that ambient O3 is frequently at, or near, the 7 h seasonal
21 mean (M7) that would be expected to cause a yield loss in crops based on the conclusions of
22 the 1986 criteria document. This table may be used for comparison of ambient-O3
23 concentrations to those used in experiments. Thirty-five percent of the 38 species or
24 cultivars under consideration would be predicted to have a 10% yield loss at a 7 h mean
25 concentration of between 0.04 and 0.05 ppm, but only 19% required a 7 h mean
26 concentration of greater than 0.08 ppm to suffer a predicted 10% loss in yield. Furthermore,
27 11 % of the 38 species or cultivars would be expected to have a yield reduction of 10% at a
28 7 h mean or less than 0.028 to 0.035 ppm (Table 6-17, U.S. Environmental Protection
29 Agency, 1986). It was also concluded that grain crops (with the exception of a few very
30 sensitive cultivars) were generally less sensitive than others, but that within-species
31 variability in sensitivity may be as great or greater than between species. These results are
December 1993 5-162 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-17. SUMMARY OF OZONE EXPOSURE INDICES CALCULA1
3- OR 5-MONTH GROWING SEASONS FROM 1982 TO 1991
FOR
3 mo
Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
(June-August)
HDM2
No. ppm
Sites
99
102
104
117
123
121
139
171
188
199
Among Years
5 mo
Mean CV
0.114 23
0.125 24
0.117 24
0.117 24
0.115 21
0.119 22
0.129 21
0.105 23
0.105 21
0.106 22
0.113 11
.7%
.9%
.6%
.6%
.8%
.9%
.3%
.1%
.6%
.0%
.1%
M7
ppm
Mean
0.052
0.056
0.052
0.052
0.052
0.055
0.060
0.051
0.053
CV
18.7%
21.9%
18.2%
17.1%
19.1%
17.6%
17.8%
17.5%
18.3%
0.054 18.4%
0.054 10.0%
SUMOO
ppnrh
Mean
82.9
86.1
84.1
84.6
85.3
86.9
97.6
86.4
85.7
87.7
87.0
CV
19.1%
22.1%
19.9%
18.0%
18.0%
17.3%
19.6%
19.9%
21.0%
21.3%
9.9%
SUM06
ppm-h
Mean
26.8
34.5
27.7
27.4
27.7
31.2
45.2
24.8
25.8
28.3
29.5
CV
68.8%
58.1%
58.4%
59.6%
65.0%
56.4%
46.8%
78.7%
76.2%
74.2%
42.1%
SIGMOID
ppm-h
Mean
26.3
33.0
27.4
27.4
27.7
30.4
42.9
25.8
26.6
28.9
29.4
CV
56.7%
52.3%
47.9%
47.6%
51.8%
46.8%
42.4%
59.4%
59.2%
59.5%
31.0%
(May-September)
M7
Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
No.
Sites
88
87
95
114
118
116
134
158
172
190
Among Years
ppm
Mean
0.048
0.051
0.048
0.048
0.048
0.050
0.054
0.047
0.049
0.050
0.049
SUMOO
ppm
CV Mean
20
22
18
18
20
20
18
18
19
19
9
.6%
.1%
.0%
.4%
.3%
.3%
.7%
.6%
.8%
.8%
.8%
122.9
129.6
126.2
124.5
123.3
128.7
141.7
127.8
129.4
130.6
129.0
•h
CV
22.3%
24.4%
19.1%
19.4%
21.4%
20.4%
22.0%
22.5%
22.7%
23.6%
9.9%
SUM06
ppm-h
SIGMOID
ppm-
Mean CV Mean
37.
44.
36.
36.
34.
3 70.
4 61.
7 60.
2 63.
9 70.
42.2 62.
58.
0 50.
32.7 87.
34.6 82.
36.8 80.
38.7 42.
9%
9%
8%
8%
7%
0%
5%
8%
7%
7%
5%
37.1
43.8
37.6
37.0
35.6
41.8
55.6
35.2
37.0
38.8
39.6
h
CV
57.8%
52.7%
46.9%
50.3%
55.7%
50.3%
45.0%
64.1%
62.1%
62.9%
29.8%
aUpdated and additional years from data given in Table HI of Tingey et al (1991) where the spatial and
temporal variation in ambient O3 exposures is expressed in terms of several exposure indices.
No. sites indicates the number of separate monitoring sites included in the analysis. Fewer sites had 5 mo of
available data than 3 mo of available data. The 2nd HDM index is calculated for sites with at least 3 mo of
available data.
CSUMOO, SUM06, M7, SIGMOID, and 2nd HDM are the cumulative sum above 0.0 ppm, the cumulative sum
above 0.06 ppm, the 7 h seasonal mean, the sigmoid weighted summed concentration, and the second highest
daily maximum 1 h concentration, respectively.
CV = coefficient of variation.
Source: Tingey et al. (1991).
December 1993
5-163
DRAFT-DO NOT QUOTE OR CITE
-------
1 similar to those previously obtained from Table 6-19 in the 1986 document. Lee et al.
2 (1993a) have revised Table 6-19 (see Table 5-16) in U.S. Environmental Protection Agency
3 (1986) using re-calculated peak-weighted exposure indices (shown to be more appropriate
4 than long-term means for relating effects to ambient concentrations) for the 54 studies
5 (Tables 5-18 and 5-19).
6 In 1992, a Supplement to the Air Quality Criteria Document for Oxidants reviewed
7 effects of oxidant exposure on vegetation. Considerable emphasis was placed on the
8 appropriate exposure index for relating biological effects of O3 on plants (U.S.
9 Environmental Protection Agency, 1992). An analysis of the data at that time showed that a
10 seasonal mean concentration (e.g., 7 or 24-h) might not be the best expression of the
11 exposure since it did not weight high concentrations differently from low concentrations and
12 it did not account for the variable length of growing seasons or exposure durations.
13 Unfortunately, it is often impossible to calculate the different possible exposure indices
14 (means, cumulative peak-or threshold-weighted, or continuously weighted [sigmoid]
15 cumulative) from information given in published papers. Thus, difficulties remain in
16 comparing exposure-response studies that utilize different exposure indices. However,
17 reported responses and concentrations of O3 can be compared to those that occur at ambient
18 concentrations, and thence to other exposure indices (Table 5-17).
19
20 5.6.3.1 Effects of Ozone on Short-Lived (Less Than 1-Year) Species
21 Plant species can be characterized by their life-span, either short-lived annual species,
22 or longer-lived perennials and trees. Physiological processes may be related to life-span (for
23 instance, leaf gas exchange tends to be lower in longer-lived trees than in crop species) and
24 so, the response to O3 may be different (Reich, 1987). In addition, multiple year exposures
25 and carry-over effects may be of importance in long-lived species, but of no concern in
26 annuals. Accordingly, animals and perennials will be discussed separately. The response of
27 plants to O3 is also affected by interactions with other physical, chemical, and biological
28 factors. Those interactions are discussed elsewhere in this document (Section 5.3). In most
29 cases, the research analyzed here was conducted under near-optimal conditions of water and
30 nutrient availability. While deviations from these conditions may affect the magnitude of
December 1993 5-164 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-18. COMPARISON OF EXPOSURE-RESPONSE CURVES CALCULATED
USING THE 3-MONTH, 24-HOUR SUM06 VALUES FOR 54 NCLAN CASES
Species
Barley (Linear)
Barley (Linear)
Corn(L)
Corn (L)
Cotton (L)
Cotton (L)
Cotton (L)
Cotton (L)
Cotton (L, Linear)
Cotton (L, Linear)
Cotton
Cotton
Cotton
Kidney Bean
Kidney Bean (L)
Lettuce (T)
Peanut (L)
Potato
Potato
Sorghum
Soybean
Soybean
Weibull/Linear Model
Parameters
Cultivar Moisture* ABC
CM-72 DRY
CM-72 WET
PIO
PAG
ACALA DRY
ACALA WET
ACALA DRY
ACALA WET
ACALA DRY
ACALA WET
STONEVILLE
MCNAIR DRY
MCNAIR WET
CAL LT RED
CAL LT RED
EMPIRE
NC-6
NORCfflP
NORCfflP
DEKALB
CORSOY
CORSOY
7741.1
8776.6
9627.4
10730.1
6465.0
9808.0
7009.8
7858.8
5.693
5.883
3576.1
3698.8
4811.0
2488.2
2484.3
7196.6
6402.5
5900.7
5755.6
8046.2
2652.6
1891.7
-4.412
15.485
92.61
94.36
92.59
71.17
83.78
78.01
-0.0011
-0.0017
94.6
165.81
117.02
27.41
44.24
54.87
100.12
93.84
79.26
178.05
57.1
65.21
2.823
4.316
2.361
1.997
1.849
1.311
2.012
2.778
1.534
3.885
2.691
5.512
2.226
1.000
1.654
2.338
1.726
5.160
RMSE°
1215
1175
680
1248
1097
521
949
937
104
90
226
342
366
333
397
613
351
742
675
441
166
282
3 mo 24-h
SUM06e
Values for Yield
Losses of
R2 10% 30%
0.12
NA
0.93
0.80
0.45
0.96
0.80
0.85
0.06
0.20
0.91
0.46
0.89
0.72
0.71
0.74
0.97
0.63
0.49
0.48
0.91
0.63
175.5
250.0
41.7
56.0
35.7
23.1
24.8
14.0
94.9
60.3
30.9
73.8
27.0
15.4
19.2
36.5
36.4
9.9
20.3
68.0
15.5
42.2
526.4
250.0
64.3
74.3
59.8
42.5
48.0
35.5
321.3
204.0
56.7
114.4
59.7
21.0
30.2
45.5
63.0
33.5
42.5
114.6
31.4
53.4
December 1993
5-165
DRAFT-DO NOT QUOTE OR CITE
-------
TA&LE 5-18 (cont'd). COMPARISON OF EXPOSURE-RESPONSE
CURVES CALCULATED USING THE 3-MONTH, 24-HOUR SUM06 VALUES
FOR 54 NCLAN CASES
Species
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Tobacco (L)
Turnip (T)
Turnip (T)
Turnip (T)
Cultivar
AMSOY
PELLA
WILLIAMS
CORSOY
CORSOY
CORSOY
CORSOY
CORSOY
CORSOY
WILLIAMS
WILLIAMS
HODGSON
DAVIS
DAVIS
DAVIS
DAVIS
DAVIS
DAVIS
YOUNG
YOUNG
MCNAIR
JUST RIGHT
PURPLE TOP
SHOGOIN
Weibull/Linear Model
Parameters
Moisture* ABC
DRY
WET
DRY
WET
DRY
WET
DRY
WET
DRY
WET
DRY
WET
DRY
WET
1907.2
2619.9
2368.4
2229.8
2913.8
3528.1
4905.0
5676.1
5873.9
6305.2
7338.4
2052.4
3929.7
4815.5
2007.1
4568.0
5775.6
8082.7
5978.8
7045.0
5177.4
12.7
5.7
4.4
75.
174.
146.
92.
311.
91
13
37
0
04
103.83
117.98
97.46
65.73
99.18
78.71
79.97
131.57
85.71
542.36
158.57
90
113
183
145
172
25
29
29
.18
.89
.63
.63
.55
.68
.26
.18
2.739
1.000
1.000
9.593
1.527
15.709
3
1
1
1
1
1
1
1
1
1
3
1
1
1
1
1
.590
.000
.319
.456
.344
.000
.000
.734
.000
.539
.348
.442
.448
.277
.186
.806
1.437
1
.548
RMSEC
390
311
527
193
330
400
401
508
512
389
377
361
524
346
556
495
920
927
244
424
306
0.810
0.590
0.660
,d
R2
0.41
0.51
0.27
0.16
0.38
0.55
0.80
0.81
0.89
0.87
0.94
0.78
0.64
0.87
0.04
0.61
0.55
0.71
0.93
0.93
0.81
0.96
0.92
0.81
3 mo 24-h
SUM066
Values for Yield
Losses of
10% 30%
33.4
18.3
15.4
72.8
71.3
90.0
63.0
10.3
11.9
21.1
14.8
8.4
13.9
23.4
57.1
36.8
46.0
23.9
38.8
25.0
25.9
7.4
6.1
6.8
52.1
62.1
52.2
82.6
158.4
97.2
88.5
34.8
30.1
48.8
36.5
28.5
46.9
47.3
193.4
81.2
66.3
55.7
90.1
65.0
72.3
14.5
14.3
15.0
December 1993
5-166
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-18 (cont'd). COMPARISON OF EXPOSURE-RESPONSE
CURVES CALCULATED USING THE 3-MONTH, 24-HOUR SUM06 VALUES
FOR 54 NCLAN CASES
3 mo 24-h
SUM06e
Weibull/Linear Model
Parameters
Species
Turnip (T)
Wheat
Wheat
Wheat
Wheat
Wheat
Wheat
Wheat
Cultivar
TOKYO
CROSS
ABE
ARTHUR
ROLAND
ABE
ARTHUR
VONA
VONA
Moisture* A
11.7
5149.8
4455.8
5028.9
6043.1
5446.9
5384.0
4451.0
B
27.83
52.89
60.87
52.32
47.39
72.34
27.74
33.5
C
2.142
3.077
2.176
1.173
7.711
2.462
1.000
1.818
RMSE6
3.250
399
264
405
226
349
608
654
2d
R2
0.78
0.90
0.92
0.91
0.74
0.57
0.88
0.64
Values for Yield
Losses of
10%
9.7
25.5
21.6
7.7
35.4
29.0
2.9
9.7
30%
17.2(
37.8
37.9
21.7
41.5
47.6
9.9
19.0
aWet refers to experiments conducted under well-watered conditions while dry refers to experiment conducted
under some controlled level of drought stress.
bFor those studies whose species name is followed by "(Linear)" a linear model was fit.
A Weibull model was fit to all other studies and estimates of B parameter are in ppm-h. The
yield is expressed in kg/Ha for all crops except turnip (g/plant) and lettuce (g/m). In cases
where the estimated C parameter is exactly 1.0, the shape parameter has been bounded from
below to obtain convergence in the nonlinear model fitting routine. For those studies whose
species name is followed by "(L)" a log transformation was used to stabilize the variance. For
those crops whose name is followed by "(T)" the yield is expressed as either g/plant or g/m.
cThe root mean square error based on individual plot means.
dR2 or multiple correlation coefficient measures the proportion of total variation about the mean
response explained by the regression on individual plot means..
eThe 24-h SUM06 value (ppm-h) that was predicted to cause a 10 or 30% yield loss
(compared to zero SUM06).
Source: Lee et al. (1993b).
1 response, it is important to understand the potential of Oj exposure, and to understand the
2 consequences.
3 Several papers (Lee et al., 1988, 1991, 1993a,b; Lefohn et al., 1988; Lesser et al.,
4 1990; Tingey et al., 1991) present a re-analysis of NCLAN data and/or data from field
December 1993
5-167 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-19. COMPARISON OF EXPOSURE-RESPONSE CURVES CALCULATED
USING THE 24-HOUR W126 VALUES FOR 54 NCLAN CASES
Species
Cultivar Moisture* A
Barley
Barley
Corn (L)
Corn (L)
Cotton
Cotton
Cotton
Cotton
Cotton
Cotton
Cotton
Cotton
Cotton
(L)
(L)
(L)
(L)
(L)
(L)
Kidney Bean
Kidney Bean (L)
Lettuce (T)
Peanut
Potato
Potato
(L)
Sorghum
Soybean
Soybean
CM-72 DRY
CM-72 WET
PIO
PAG
ACALA DRY
ACALA WET
ACALA DRY
ACALA WET
ACALA DRY
ACALA WET
STONEVILLE
MCNAIR DRY
MCNAIR WET
CALLTRED
CAL LT RED
EMPIRE
NC-6
NORCfflP
NORCfflP
DEKALB
CORSOY
CORSOY
Weibullb
B
8133.2 1109.6
8927.2 57439.6
9605.0 92.9
10686.7 94.5
6482.8
9817.
7022.
7927.
310.
393.
3
7
1
1
2
3592.1
3700.9
4817.6
2484.7
2475.2
7197.4
6386.0
5867
.2
5777.9
8049.7
2660
1895
.3
.6
89.9
66.6
81.3
74.7
174.1
582.6
94.1
174.1
113.5
28.0
44.2
54.6
97.4
96.3
113.9
205.9
58.8
63.3
C RMSEC
1.000 1214
1.000 1175
2.594 650
4.1901253
1.949 1075
1.603
1.540
1.070
2.189
1.000
1.582
2.430
1.410
3.706
2.353
4.921
1.905
1.000
1.299
1.963
1.455
4.032
514
948
943
104
90
223
344
360
332
401
614
370
754
675
439
169
280
2d
R2
0.13
NA
0.93
0.80
0.47
0.96
0.80
0.85
0.06
0.20
0.91
0.45
0.89
0.72
0.70
0.74
0.96
0.62
0.48
0.48
0.91
0.63
24-h W126e
Values for Yield
Losses of
10% 30%
116.
6051.
39.
55.
28.
16.
18.
9.
62.
61.
22,
68.
23.
15.
17
34
9 395.8
9 20487.3
0 62.4
2 73.9
3
4
8
1
,3
,4
,7
9
.0
.3
.0
.6
29.9
10
20
65
12
36
.1
.1
.4
.5
.2
53.0
35.0
41.6
28.5
108.7
207.8
49.1
113.9
54.6
21.2
28.5
44.3
56.7
34.3
51.5
121.8
28.9
49.0
December 1993
5-168 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-19 (cont'd). COMPARISON OF EXPOSURE-RESPONSE CURVES
CALCULATED USING THE 24-HOUR W126 VALUES FOR 54 NCLAN CASES
Species
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Cultivar
AMSOY
PELLA
WILLIAMS
CORSOY
CORSOY
CORSOY
CORSOY
CORSOY
CORSOY
WILLIAMS
WILLIAMS
HODGSON
DAVIS
DAVIS
DAVIS
DAVIS
DAVIS
DAVIS
YOUNG
YOUNG
Moisture A
Weibullb
B
1926.1
2602.4
2341.8
DRY
WET
DRY
WET
DRY
WET
DRY
WET
2229.3
2929.7
3533.5
4909.5
5597.1
5884.8
6314
7352
2044
.1
.3
.6
3837.6
DRY
WET
DRY
WET
DRY
WET
4810
1992
4595
.8
.3
.4
5770.1
8101
.3
5994.2
7075.0
79.0
161.
138.
88.
5
6
2
470.2
113.
126.
2
5
95.7
65.6
106.3
80.7
76.2
130.3
87.5
537.6
170.9
90.6
118
199
.2
.8
149.7
C RMSEC
1.977
1.000
1.000
8.632
1.128
11.095
2.803
1.000
1.139
1.243
1.162
1.000
1.000
1.494
1.000
1.253
2.796
1.220
1.251
1.133
390
314
533
192
329
403
405
526
515
391
368
361
530
352
558
496
928
939
244
418
R2
0.41
0.50
0.25
0.16
0.39
0.54
0.80
0.80
0.88
0.87
0.95
0.78
0.63
0.86
0.03
0.61
0.54
0.70
0.93
0.93
24-h W126e
Values for Yield
Losses of
10% 30%
25.3
17.0
14.6
67.9
64.0
92.4
56.7
10.1
9.1
17.4
11.6
8.0
13.7
19.4
56.6
28.4
40.5
18.7
33.1
20.5
46.9
57.6
49.4
78.2
188.6
103.1
87.6
34.1
26.6
46.4
33.2
27.2
46.5
43.9
191.7
75.1
62.7
50.8
87.7
60.2
December 1993
5-169 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-19 (cont'd). COMPARISON OF EXPOSURE-RESPONSE CURVES
CALCULATED USING THE 24-HOUR W126 VALUES FOR 54 NCLAN CASES
Species
Tobacco (L)
Turnip (T)
Turnip (T)
Turnip (T)
Turnip (T)
Wheat
Wheat
Wheat
Wheat
Wheat
Wheat
Wheat
Weibullb
Cultivar Moisture* A B
MCNAIR
JUST RIGHT
PURPLE TOP
SHOGON
TOKYO
CROSS
ABE
ARTHUR
ROL
ABE
ARTHUR
VONA
VONA
5223.9
12.7
5.8
4.4
11.7
5138.1
4467.4
5074.4
6042.8
5440.0
5300.8
4462.7
179.8
24.1
28.2
28.2
26.8
53.3
63.8
51.2
48.5
76.1
25.0
32.3
C RMSEC
1.018 291
1.473 0.96
1.155 1
1.174 1
1.710 3
2.602 407
1.747 264
1.000 397
5.843 225
2.100 349
1.000 679
1.517 665
2d
R2
0.83
0.92
0.82
0.78
0.89
0.92
0.91
0.75
0.57
0.85
0.63
24-h W126e
Values for Yield
Losses of
10% 30%
19.7
5.21
4.0
4.1
7.2
22.4
17.6
5.4
33.0
26.1
2.6
7.3
65.3
2.0
11.6
11.7
14.7
35.8
35.4
18.3
40.6
46.6
8.9
16.4
Wet refers to experiments conducted under wll-watered conditions while dry refers to experiments conducted
under some controlled level of drought.
All estimates of B parameter are in ppm-h. The yield is expressed in kg/Ha for all crops except turnip
(g/plant) and lettuce (g/m). In cases where the estimated C parameter is exactly 1.0, the shape parameter has
been bounded from below to obtain convergence in the nonlinear model fitting routine. For those studies
whose species name is followed by "(L)" a log transformation was used to stabilize the variance. For those
crops whose name is followed by "(T)" the yield is expressed as either g/plant or g/m.
cThe root mean square error based on individual plot means.
R or multiple correlation coefficient measures the proportion of total variation about the mean response
explained by the regression on individual plot means.
"The 24-h W126 value (ppm-h) that was predicted to cause a 10 or 30% yield loss (compared to zero
W126).
Source: Lee et al. (1993b).
December 1993
5-170 DRAFT-DO NOT QUOTE OR CITE
-------
1 studies conducted on potato that were not part of the NCLAN project. Lee et al. (1988,
2 1991) examined a number of measures of O3 exposure in relation to response data collected
3 in the experiments. They were particularly interested in examining the ability of a seasonal
4 mean, a cumulative exposure index, and the second highest daily maximum concentration
5 (2HDM) to predict the biological response of the plant. They found that no particular index
6 of O3 concentration dominated as best in all studies, but that cumulative indexes that
7 weighted high concentrations at the "grain-filling" stage of the life cycle were better than a
8 seasonal mean. Seasonal means did work well within a given experiment where treatments
9 were highly correlated. The 2HDM was consistently a poor predictor of plant response.
10 In a re-analysis of NCLAN data, Lesser et al. (1990) presented composite exposure-
11 response functions for a number of crop species, or groups of species. Predicted yield losses
12 (compared to yield at an assumed background concentration of 0.025 ppm) of up to 20%
13 occurred at a 12 h seasonal mean of 0.06 ppm, with a loss of 10% at a 12 h mean
14 concentration of about 0.045 ppm.
15 Tingey et al. (1991) and Lee et al. (1993) went on to re-analyze the crop response data
16 using three measures of exposure: (1) the SUM06 (the cumulative sum of all hourly
17 O3 concentrations greater than 0.06 ppm), (2) the 7 h seasonal mean, and (3) the 2HDM.
18 Their analysis included crops that account for 70% of all crop land in the United States and
19 73% of the agricultural receipts. The analysis included 31 field experiments with 12 crop
20 species, conducted in open-top chambers and resulted in composite exposure-response
21 functions. The results of their studies and additional re-analyses done since then are
22 summarized in Tables 5-20 and 5-21. They concluded that to limit yield loss to 10% or less
23 in 50% of the cases (all experiments and crops), a SUM06 of 24.4 ppnvh (or 26.4 ppm-h
24 based on 24 h), a 7 h seasonal mean of 0.049 ppm, or a 2HDM of 0.094 ppm would be
25 required. A SUM06 of about 37 ppnvh should limit yield losses to 20% in 50% of the
26 cases. If one standard error were added to or subtracted account for the variability, the
27 metrics would be reduced to 21, 0.046, and 0.088 or increased to 27.8 ppm-h, 0.049 ppm
28 and 0.10 ppm respectively. To limit the loss to 10% or less in 75% of the cases would
29 require 14.2 ppm-h, 0.040 ppm and 0.051 ppm respectively (Table 5-20). These values are
30 based on studies of both well-watered and drought stressed plants.
December 1993 5.171 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-20. THE EXPOSURE LEVELS (USING VARIOUS INDICES)
ESTIMATED TO CAUSE AT LEAST 10% CROP LOSS IN
50 AND 75% OF EXPERIMENTAL CASES
50th Percentile No.
NCLANData(N=49;
NCLAN Data (N=39;
NCLAN Data (N=54;
NCLAN Data (N = 42;
NCLAN Data (N= 10;
NCLAN Data (N= 10;
Cotton Data (N=5)
Soybean Data (N=13)
Wheat Data(N=6)
Cotton Data(N=5) **
Soybean Data (N= 15)
Wheat Data (N=7) **
75th PERCENTILE #
NCLAN Data (N = 49;
NCLAN Data (N= 39;
NCLAN Data (N =54;
NCLAN Data (N = 42;
NCLAN Data (N= 10;
NCLAN Data (N= 10;
Cotton Data (N=5)
Soybean Data (N = 13)
Wheat Data (N=6)
Cotton Data (N=5) **
Soybean Data (N=15)
Wheat Data(N = 7) **
SUM06
Wet and Dry) (?)
Wet Only)
Wet and Dry) **
Wet Only) **
Wet)
Dry)
**
Wet and Dry)
Wet Only)
Wet and Dry) **
Wet Only) **
Wet)
Dry)
**
24.4
22.3
26.4
23.4
25.9
45.7
23.6
26.2
21.3
30.0
23.9
25.9
14.2
14.3
16.5
17.2
16.4
24.0
21.8
14.2
11.7
21.1
15.3
5.1
SE* S1GMOID
3.
1.
3.
3.
4.
23.
4
0
2
1
5
3
2.3
5.
4
15.2
12.7
6.5
10.5
4,
2,
4
3
3
0
5
0
2
6
4
2
.2
.7
.3
.0
.7
.8
.0
.1
.5
.0
.1
.6
21.
19.
23.
22.
23.
40.
19.
22.
5
4
5
9
4
6
3
6
19.3
27.2
22.0
21.4
11.9
12.6
14.5
14.7
13
22
17
12
10
16
13
8
.7
.3
.5
.4
.9
.7
.4
.5
SE
2.0
2.3
2.4
4.7
3.2
0.1
2.3
3.6
12.7
12.8
8.0
9.4
5.6
2.3
3.2
2.4
3.2
0.1
2.8
0.1
2.4
5.7
4.1
3.4
M7
0.049
0.046
NA
NA
0.041
0.059
0.041
0.044
0.061
NA
NA
NA
0.040
0.039
NA
NA
0.040
0.053
0.041
0.041
0.054
NA
NA
NA
SE 2ndHDM
0.003
0.003
NA
NA
0.001
0.014
0.001
0,005
0.018
NA
NA
NA
0.007
0.005
NA
NA
0.001
0.022
0.001
0.006
0.032
NA
NA
NA
0.094
0.090
0.099
0.089
0.110
0.119
0.066
0.085
0.098
0.075
0.088
0.097
0.051
0.056
0.073
0.070
0.080
0.093
0.065
0.069
0.062
0.070
0.078
0.054
SE
0.006
0.010
0.011
0.008
0.042
0.017
0.032
0.013
0.059
0.012
0.008
0.028
0.010
0.006
0.006
0.006
0.032
0.003
0.014
0.004
0.035
0.034
0.007
0.027
# The numbers in parentheses are the number of cases used in deriving the various exposure levels.
* Standard error (SE).
? NCLAN data refers to studies conducted as part of the NCLAN project. Wet and dry refer to watery regimes
used in the studies, wet being well-watered, and dry meaning some level of drought stress was imposed.
** 24-h exposure statistics reported in Lee et al. (1993). Relative yield loss for 2ndHDM is relative to yield at
0.04 ppm rather than 0.00 ppm as was used in Tingey et al. (1991).
Modified from: Tingey et al. (1991).
1 A further analysis by Lee et al. (1993b) provides composite exposure-response
2 functions for all NCLAN studies, as well as for soybean and wheat experiments
3 (Table 5-21). In the analysis, they calculated the SUM06 based on 24 h per day
December 1993
5-172
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-21. SUM06 LEVELS ASSOCIATED WITH 10 AND 20% YIELD LOSS
FOR 50 AND 75% OF THE NCLAN CROP STUDIES
Weibull equations (all 54 NCLAN studies):
50th Percentilea PRYL = 1 - exp(-[SUM06/89.497]**1.84461)
75th Percentile PRYL = 1 - exp(-[SUM06/60.901]**l.72020)
Weibull equations (all 22 NCLAN soybean studies; 15 well-watered, 7 water-stress):
50th Percentile PRYL = 1 - exp(-[SUM06/l 17.68]**!.46509)
75th Percentile PRYL = 1 - exp(-[SUM06/88.99]**1.47115)
Weibull equations (15 NCLAN well-watered soybean studies):
50th Percentile PRYL = 1 - exp(-[SUM06/l 12.75]**!.46150)
75th Percentile PRYL = 1 - exp(-[SUM06/79.62]** 1.36037)
Weibull equations (7 NCLAN wheat studies):
50th Percentile PRYL = 1 - exp(-[SUM06/49.02]**3.52788)
75th Percentile PRYL = 1 - exp(-[SUM06/29.56]**l .29923)
SUM06 levels associated with 10% and 20% yield loss for 50% and 75% of the crops.
All 54 NCLAN cases
Relative 10%
Yield Loss 20%
Percent of crops
50% 75%
26.4 16.5
39.7 25.5
All 22 NCLAN soybean cases
15 well-watered soybean cases
All 7 NCLAN wheat cases
Relative
Yield Loss
Relative
Yield Loss
10%
20%
10%
20%
Relative
Yield Loss 20%
Percent of crops
50% 75%
25.3 19.3
42.3 32.1
Percent of crops
50% 75%
24.2 15.2
40.4 26.4
Percent of crops
50% 75%
25.9 5.2
32.0 9.3
50th and 75th percentile refer to the percentage of studies analyzed in which loss of the stated magnitude would
have been prevented.
Source: Lee et al. (1993a,b).
December 1993
5-173
DRAFT-DO NOT QUOTE OR CITE
-------
1 03 concentrations and the resulting exposure to prevent crops from yield loss is slightly
2 higher than they previously calculated (26.4 ppnvh versus 24.4 ppnvh, Table 5-20).
3 A composite exposure-response function based on uncited data using GIS, TRIGRO and
4 ZELIG models is illustrated in Figure 5-22.
5 Research since 1986 has largely focused on understanding the response of trees and
6 other perennials to O3 (covered in the next section) and of five crop species: cotton, wheat,
7 spring rape, Phaseolus bean, and soybean. A number of the studies were conducted as part
8 of NCLAN, but many were also the result of research activity in Europe. Results of these
9 studies, as well as those species studied less intensively, are summarized in Table 5-22.
10 Yield losses in cotton of 13 to 19% have been reported at 12 h mean concentrations of
11 0.050 or 0.044 ppm by Heagle et al. (1988), Miller et al. (1988), and Temple et al. (1988b)
12 (Table 5-22). These are typical ambient concentrations as listed under M7 (Table 5-17).
13 The same experiments showed that drought stress reduced the predicted yield loss due to O3,
14 but did not eliminate it.
15 Wheat yields have been reduced by 0 to 29%, depending on the cultivar and exposure
16 conditions (Adaros et al., 1991a; Fuhrer et al., 1989; Grandjean and Fuhrer, 1989; Kohut
17 et al., 1987; Pleijel et al., 1991) (Table 5-22). In no case was a 7 h average of greater than
18 0.062 ppm required to cause the reported loss, but Slaughter et al. (1989) suggest that hourly
19 concentrations above 0.06 ppm during the period following anthesis may be particularly
20 effective in reducing yield.
21 Studies with spring rape in Europe have documented yield losses of 9.5 to 26.9% at
22 8 h growing season average concentrations ranging from 0.03 to 0.06 ppm (Adaros et al.,
23 1991b,c) (Table 5-22).
24 The yield of Phaseolus beans (fresh pods) was reduced by 17% at a 7 h average of
25 0.045 ppm (Schenone et al., 1992) or 20% at an 8 h growing season average of 0.080 ppm
26 (Bender et al., 1990). In a similar study, Heck et al. (1988) the predicted yield of sensitive
27 cultivars was reduced an average of 17.3% by exposure to a 7 h growing season mean of
28 0.05 ppm, but resistant cultivars suffered only a 1.6% loss. Temple (1991) reported
29 reductions in dry bean yield of 44 to 73% in three cultivars grown in California and exposed
30 to a 12 h seasonal mean of 0.072 ppm. One other cultivar increased in yield hi non-filtered
December 1993 5-174 DRAFT-DO NOT QUOTE OR CITE
-------
75th percontile
50th peroentite
25th percentfte
~r
10
"i
20
100%-]
90%
80%
70%-:
8 60%^
S 50%-;
£ '•
•M 40%-
•s :
& 30%-
20%-
10%-
0%
B. Tree Seedlings
30 40
24-h Sum06 (ppm-h)
T
50
"1"
60
^s^^ «H« ^^ _ »•»•" _— _
~i "T" "i"" "r r"
10 20 30 40 50
24-h Sum06 (ppm-h) (adjusted to 92 days)
75th percentile
; 50th percentile
' 25th percentile
"\~"
60
Figure 5-22. Box-plot distribution of biomass loss predictions from Weibull and linear
exposure-response models that relate biomass and ozone exposure as
characterized by the 24-h SUM06 statistic using data from (A) 31 crop
studies from the National Crop Loss Assessment Network (NCLAN)
program and (B) 26 tree seedling studies conducted at the Environmental
Research Laboratory in Corvallis, Oregon; Smoky Mountain National
Park, Tennessee; Michigan; Ohio; and Alabama. Separate regressions
were calculated for studies with multiple harvests and/or cultivars
resulting in a total of 54 individuals equations from the 31 NCLAN studies
and 56 equations from the 26 seedling studies. Each equation was used to
calculate the predicted relative yield or biomass loss at 10, 20, 30, 40, 50,
and 60 ppnrh and the distributions of the resulting losses plotted. The
solid line is the calculated Weibull fit at the 50th percentile. From Hogsett
et al. (1993).
December 1993
5-175
DRAFT-DO NOT QUOTE OR CITE
-------
O
o
1
1
1
ON
O
^
H
6
o
2j
9
0
C
H
O
n
TABLE 5-22. A SUMMARY OF STUDIES REPORTING THE EFFECTS OF OZONE
ON THE GROWTH, PRODUCTIVITY, OR YIELD OF ANNUAL PLANTS PUBLISHED SINCE
U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
Species
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
E/D/
PC/Ca Concentration15
C 18 or 24 ppb vs. 59 or
72 ppb, 9 h mean
C 23, 40, and 66 ppb 7 h
mean
C 97 ppb vs. 38, 23, 16,
23 ppb 7 h mean
C 17 to 122 ppb 7 h mean
C 25 and 50 ppb 7 h mean
C 20 and 50 ppb 12 h mean
C 25 and 55 ppb 7 h mean
C 27 and 54 ppb 7 h means
Duration
13 weeks,
2 growing
seasons
84 days
4 31 day
periods,
1 growing
season
69 days
about 90 days
107 days
64, 70, and 62
days,
3 growing
seasons
about 109 and
103 days,
2 growing
seasons
Facility0 Variable41 Effect6
OTC seed yield 12.5% reduction over
filtered air averaged over
cultivars. Between caltivar
differences as great as ozone
effect.
OTC seed yield 15.8 and 29% reduction
over 23 ppb.
OTC seed yield 30 to 56 % reduction over
in pots control, most loss in mid to
late growth stage.
OTC seed yield 8% at 35 ppb to 41 % at 122
ppb.
OTC seed yield Predicted loss of 10%.
OTC seed yield Predicted loss of 13%.
OTC seed yield Predicted loss of 15%.
OTC seed yield Predicted loss of 12 and
14%.
Reference
Mulchi et al.
(1988)
Mulchi et al.
(1992)
Heagle et al.
(1991)
Kohut et al.
(1986)
Heagle et al.
(1986b)
Miller et al.
(1989)
Heggestad and
Lesser (1990)
Heagle et al,
(1987)
-------
TABLE 5-22 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
CD
1
J2 Species
Soybean
Soybean
Soybean
± Cotton
-j
U Cotton
fr
O Cotton
§
^? Cotton
§
H Cotton
o
i
o
UZAJrNJk UIN lilt LrKUWlH, itCULIUU 11V11 Y , UK YLfcLJJ Uf AN IN U AT, fJUAINlS
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
E/D/
PC/Ca Concentration6 Duration
C filtered and non-filtered about
air-concentration not 125 days
reported 2 growing
seasons
C 10 to 130 ppb 8 weeks,
6.8 h/day
C 200 ppb 12 h, up to
4 times
C 15 to 111 ppb 12 h mean 123 days
C 10 to 90 ppb 12 h mean 102 days
C 25 to 74 ppb 12 h mean 123 days
C 22 to 44 ppb 12 h mean 124 days
C 26 to 104 ppb 7 h mean 1 19 days
c d
Facility Variable
OTC seed yield
GC biomass
GC shoot and
root
weight
OTC leaf,
stem, and
root
weight
OTC lint
weight
OTC lint
weight
OTC lint
weight
OTC lint
weight
Effect6
No difference.
Predicted reduction of 16 or
33 % at 60 and 100 ppb vs
25 ppb.
No effect at maturity.
Up to 42 % reduction in leaf
and stem, and 6 1 %
reduction in root weights.
40 to 71 % reduction at
highest concentration
determinant cultivars more
susceptible.
Predicted loss of 26.2% at
74 ppb.
Predicted loss of 19% at
44 ppb.
Predicted loss of 1 1 % at
53 ppb.
Reference
Johnston and
Shriner (1986)
Amundson et al.
(1986)
Smith et al.
(1990)
Temple et al.
(1988c)
Temple (1990b)
Temple et al.
(1988b)
Heagle et al.
(1988)
Heagle et al.
(1986a)
-------
TABLE 5-22 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
1
^_ n
^O
i
!!j
oo
O
3
O
O
g
3
10% in 8 lines at
63 ppb 7 h mearl.
15.5% reduction at 45 ppb
(3S» ppnvh).
3.5 to 26% reduction in
resistant and sensitive
cultivars at 55-60 ppb.
20% reduction at 80 ppb.
55 to 75 % reduction at
72 ppb 12 h mean,
198 highest hour.
26-42% reduction at 38 to
50 ppb.
growth response detected if
exposure separated by 3 to
5 days.
13% reduction at 40 ppb.
Reference
Eason and
Reinert (1991)
Schenone et al.
(1992)
Heck et al.
(1988)
Bender et al.
(1990)
Temple (1991)
Sanders et al.
(1992)
McCool et al.
(1988)
Fuhrer et al.
(1989)
-------
TABLE 5-22 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
OZONE ON THE GROWTH, PRODUCTIVITY, OR YIELD OF ANNUAL PLANTS
1
OJ
1
^3
o
1
o
1
o
§
o
Species
Wheat-Spring
Wheat-Spring
Wheat-Spring
Wheat-Spring
Wheat-Spring
Wheat-Spring
Wheat-Spring
Wheat-Spring
PUBLISHED SINCE U.S.
E/D/
PC/Ca Concentration0
C 21.6 to 80 and 24.6 to
93.5 ppm-h
C 3 to 56 ppb 7 h mean
C 8 to 101 and 20-221 ppb
8 h mean
C 0 to 38 ppb 8 h mean
C 17 to 77 ppb 7 h mean
C 25 to 75 ppb 8 h mean
C 6 to 10 ppb 6 h/day
C 10 to 125 ppb 6 h/day
ENVIRONMENTAL PROTECTION AGENCY (1986)
Duration
82 and
88 days in
2 growing
seasons
61 and
55 days in
2 growing
seasons
118 and
98 days in
2 growing
seasons
entire
growing
season
90 and
87 days hi
2 growing
seasons
40 days
21 days
21 days and
17 days
Facility0 Variable
OTC seed
weight
OTC seed
weight
OTC seed
weight
OTC seed
weight
OTC seed
weight
OTC total
weight
GC shoot dry
weight
GC top dry
weight
Effect6
48 to 54% reduction at
80 and 93.5 ppm-h.
7% reduction at 15 and
22 ppb.
10% reduction at 17 to
23 ppb.
5 % reduction at 38 ppb.
9.5 to 11.6 reduction at
37 and 45 ppb.
Reductions at 75 ppb.
Decreased 35-60% at
101 ppb in low and high
light.
Reduced by up to 35%.
Reference
Grandjean and
Fuhrer (1989)
Pleijel et al.
(1991)
Adaros et al.
(1991a)
DeTemmerman
et al. (1992)
Fuhrer et al.
(1992)
Johnsen et al.
(1992)
Mortensen
(1990b)
Mortensen
(1990c)
-------
December 1993
Y1
i
d
F
H
8
O
i
H
8
n
TABLE 5-22 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
OZONE ON THE GROWTH, PRODUCTIVITY, OR YIELD OF ANNUAL PLANTS
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
Species
Wheat-Winter
Wheat-Winter
Wheat-Winter
Wheat-Winter
Wheat-Winter
Barley-Spring
Barley-Spring
Barley-Spring
E/D/
PC/C* Concentration Duration
C 1 1 to 42 ppb 14 week 109 days
mean
C 30 to 93 ppb 4 h mean 39 and
40 days in
2 growing
5 day /week
4h/day
C 27 to 96 ppb 7 h mean 36 days
C 22 to 96 ppb 7 h mean 65 days and
36 days in
2 growing
seasons
C 23 to 123 ppb 4 h/day 5 days at
anthesis
C 6 to 45 ppb 7 h mean 96 days
C 0.6-27 ppb monthly mean growing
season
C 0.8-83 ppb 8 h mean 97, 108 and
98 days in
3 growing
seasons
Facility0
OTC
OTC
OTC
OTC
OTC
OTC
OTC
OTC in
pots
Variable*1 Effect6
seed No effect.
weight
seed Exposures > 60 ppb during
weight anthesis reduce yield.
seed 50% reduction at 96 ppb.
weight/
head
seed 33% and 22% reduction at
weight 42 and 54 ppb.
seed Up to 28% reduction.
weight
seed No effect.
weight
seed No effect.
weight
seed 0-13% reduction at highest.
weight
Reference
Olszyk et al.
Slaughter et al.
(1989)
Amundson et al.
(1987)
Kohut et al.
(1987)
Mulchi et al.
(1986)
Pleijel et al.
(1992)
Weigel et al.
(1987)
Adaros et al.
(1991b)
-------
TABLE 5-22 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
cr
u>
i
oo
o
§>
3
6
o
s
H
O
d
§
Species
Rape-Spring
Rape-Spring
Rape-Spring
Tomato
Tomato
Tomato
Silene acaulis
Plantago
lanceolata
16 other species
PUBLISHED SINCE U.S
E/D/
PC/CT Concentration
C 25 to 75 ppb 8 h mean
C 0.8-83 ppb 8 h mean
C 43 to 60 ppb 8 h mean
C 13 to 109 12 h mean
79.5 ppm h
C 10 to 85 ppb 6 h/day
C 18 to 66 ppb 12 h mean
5 to 80 ppb 8 h/day
5 to 80 ppb 8 h/day
5 to 80 ppb 8 h/day
••••"•J •m~ "m^r m-r ^s -^f -m. ,». • .M. -•. .«. • V^ A.^. .M. *.M^1M.*M,r x_*- .*. 4 ,•_*. 1 i ^ ^/ i » • -* JL M~*4. »' ^ i. fc_?
. ENVIRONMENTAL PROTECTION AGENCY (1986)
Duration
31 days
89, 113 and
84 days in
3 growing
seasons
89, 113 and
84 days in
3 growing
seasons
75 days
12-21 days
11 weeks
up to 90 days
up to 90 days
up to 90 days
Facility
OTC
OTCin
pots
OTCin
pots
OTC
GC
OTC
GC
GC
GC
Variable
premature
senescence
seed weight
seed weight
fresh
weight
shoot dry
weight
fresh fruit
weight
dry weight
dry weight
dry weight
Effect6
Increased at 75 ppb.
9.4-16% reduction at 30 or
51 ppb.
12-27% reduction.
17 to 54% reduction at 109,
no reduction at ambient.
35 to 62% reduction.
No effect.
25% reduction at 80 ppb.
14% reduction at 50 ppb.
No effect.
Reference
Johnsen et al.
(1992)
Adaros et al.
(1991b)
Adaros et al.
(1991c)
Temple (1990a)
Mortensen
(1992b)
Takemoto et al.
(1988c)
Mortensen and
Nilsen (1992)
Mortensen and
Nilsen (1992)
Mortensen and
Nilsen (1992)
-------
TABLE 5-22 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
a
B
O"
•i
MD
i
oo
NJ
o
£
i
§
5*
Q
H
O
c|
o
a
in
90
\JZjU in KJ UIN 1±1H, IjKUW 1
PUBLISHED SINCE U.S
Species
Radish
Lettuce
Lettuce
Faba bean
Fenugreek
Chickpea
Black-gram
Rice
Rice
Watermelon
Pea
E/D/
PC/Ca
C
c
c
c
c
c
c
c
c
c
c
Concentration
20 or 70 ppb 24 h mean
21 - 123 ppb 7 h mean
10 to 34 ppb 7 week
mean
6 or 15 ppb 24 h mean
120 ppb 7 h/day
120 ppb 7 h/day
120 ppb 7 h/day
0 to 200 ppb 5 h/day
50 ppb 24 h mean
15 to 27 ppb 7 h mean
10 to 35 ppb 12 h mean
±1, i-HUJLJLJ^HVH I, UK YlfcJLJJ U* AIN1NUA1-, JriiATVlS
. ENVIRONMENTAL PROTECTION AGENCY (1986)
Duration
27 days
52 days
64 days
134 days
4 weeks
4 weeks
4 weeks
5 days/week
15 weeks
8 weeks
81 days
58 and
52 days in
2 growing
seasons
Facility0
GC
OTC
OTC
ofc
CC
CC
CC
OTC
GC
OTC
OF
Variabled
shoot and
root growth
head weight
fresh weight
seed weight
dry weight
dry weight
dry weight
seed weight
dry weight
fresh weight
and number
(marketable)
fresh weight
Effect6
36 and 45% reduction al
70 ppb.
Significant reduction at
83 ppb, 35% at 128 ppb.
No effect.
No effect.
No significant effect.
No significant effect.
No significant effect.
12 to 21% reduction at
200 ppb.
No effect at 50 ppb.
20.8 and 21. 5%
reduction at 27 ppb.
linear decrease in yield
with increasing O3
Reference
Barnes and
Pfirrman (1992)
Temple et al.
(1986)
Olszyk et al.
(1986)
Sanders et al.
(1990)
Kasana (1991)
Kasana (1991)
Kasana (1991)
Kats et al. (1985)
Nouchi et al.
(1991)
Snyder et al.
(1991)
Runeckles et al.
(1990)
-------
TABLE 5-22 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
I PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
£ E/D/
g Species PC/Ca Concentration15 Duration
Green pepper C 19 to 66 ppb 12 h mean 77 days
Green pepper C 18 to 66 ppb 12 h mean 11 weeks
Celery C 18 to 66 ppb 12 h mean 11 weeks
Facility Variable Effect
OTC fresh fruit 12% reduction at 66 ppb.
weight
OTC fresh fruit 13% reduction in fruit
weight weight at 66 ppb.
OTC shoot dy 12% reduction at 66 ppb.
weight
Reference
Takemoto
(1988c)
Takemoto
(1988c)
Takemoto
(1988c)
et al.
et al.
et al.
E = evergreen, D = deciduous, C = crop. PC = perennial crop.
Means are seasonal means unless specified. Maximums are 1 h seasonal maxima unless otherwise specified. Cumulative exposures are SUMOO unless
otherwise specified, accumulation based on 24 h/day unless otherwise noted.
^ OTC = open-top chamber with plants in ground unless specified in pots; CC = closed chamber, outside; GC = controlled environment growth chamber or
oo CSTR; GH = greenhouse; F = field; OF = open-field fumigation.
The effect reported in the study that is a measure of growth, yield, or productivity.
effect measured at specified ozone concentration, over the range specified under concentration, or predicted (if specified) to occur based on relationships developed
in the experiment.
-------
1 chambers, but was severely affected in higher concentration O3 treatments. Sanders et al.
2 (1992) also observed yield stimulation at an 7 h growing season mean of 25 ppb (often
3 considered background in the U.S.), however, significant yield reductions were measured as
4 O3 concentrations increased to 50 ppb (7 h seasonal mean).
5 A number of studies have shown soybean yields to be reduced by 10 to 15 % at 7 or
6 12 h seasonal mean concentrations of 0.05 to 0.055 (Heagle et al., 1986b, 1987; Heggestad
7 and Lesser, 1990; Miller et al., 1989).
8 A number of the studies cited above, and some of those in Table 5-22, were conducted
9 as part of NCLAN, are considered in the discussions of Tingey et al. (1991), Lee et al.
10 (1993a.b). and Lesser et al. (1990), but many of the experiments (primarily those not part of
11 NCLAN) were not included in their analyses. While the range of variability in species
12 response to O3 is apparent, these studies support, for the most part, the conclusions of U.S.
13 Environmental Protection Agency (1986), Tingey et al. (1991), and Lesser et al. (1990).
14 Table 5-22 summarizes the studies reporting the response of annual plants, particularly crops,
15 as growth, dry weight, or yield to O3 exposures (concentrations x time) under experimental
16 conditions since the previous criteria document (U.S. Environmental Protection Agency,
17 1986). Based on the results of the studies reviewed in this section, including the reanalysis
18 of NCLAN, exposures for a three month period to O3 concentrations currently occurring hi
19 the ambient air (0.048 to 0.06 ppm, 7 h seasonal mean, See M7, Table 5-17) have been
20 shown to cause losses of 10% or more in the yield of the majority of major crop plants
21 grown in the country. A number of crop species are more sensitive, and greater losses could
22 be expected (Tables 5-18 through 5-22). It should be noted that a variety of methodologies
23 have been used to generate these data. Generally speaking, data obtained through growth
24 chamber experiments, and experiments conducted using potted plants, may be less reliable
25 when assessing the effects of O3 than results from field growth plants.
26
27 5.6.4 Effects of Ozone on Long-Lived Plants
28 Quantifying exposure-response in the case of perennial plants (agricultural crops such as
29 pastures, alfalfa and shrubs and trees) is complicated by they fact that they can receive multi-
30 year exposures and because the results of exposures in a previous year, or over a number of
31 years, may be cumulative. Reduction in growth and productivity, a result of altered carbon
December 1993 5-184 DRAFT-DO NOT QUOTE OR CITE
-------
1 (sugar) allocation, may appear only after a number of years or when carbohydrate reserves
2 are depleted (U.S. Environmental Protection Agency, 1986; Lawrence et al., 1993b; Garner,
3 1991; Garner ct al., 1989). A further complication is that in the case of evergreen plants,
4 the life span of a leaf exceeds one year, and may be on the plant for several years. In such
5 cases, loss of a leaf or a reduction in photosynthetic capacity may have a large effect on a
6 plant's ability to survive and grow. Physiological differences among species (rates of gas
7 exchange, for instance) may have a tendency to equalize exposure over a number of years,
8 however, as shown in Reich's (1987) analysis of crops, hardwoods and conifers, and Pye's
9 analysis of tree species (1988). Unfortunately, there is little experimental data regarding the
10 effects of long-term O3 exposure on perennial plants, as only a few experimental studies have
11 extended exposures beyond a single growing season. Most of what is known regarding the
12 effects of O3 on mature trees is from field observations. Some studies that have extended
13 observation of growth alterations into the season following exposures and thus observed
14 "carry-over effects" in several species. Hogsett et al. (1989) reported altered bud elongation
15 in ponderosa pine, lodgepole pine and western hemlock following a season of 03. Altered
16 root regrowth in ponderosa pine in the season following exposure that was correlated with
17 root storage carbohydrate was observed by Andersen et al. (1991). Most studies have used
18 seedlings due to the difficulty of exposing large trees. The extrapolation from seedlings to
19 large trees and to forest stands is not straight-forward, and will, most likely, depend on the
20 use of models (Hogsett et al., 1993; Laurence et al., 1993a,b; Taylor and Hanson, 1992).
21 Correlative studies, such as those conducted in the San Bernadino mountains of California,
22 indicate potentially large impacts on ecosystems (U.S. Environmental Protection Agency,
23 1986). This section will address three distinct types of long-lived plants: multiple-year
24 agricultural crops, deciduous shrubs and trees, and evergreen coniferous trees.
25
26 5.6.4.1 Perennial Agricultural Crops
27 Cooley and Manning (1988) conducted a greenhouse study of the response of alfalfa to
28 O3 applied at 0.06 to 0.08 ppm for 6 h per day, 5 days a week for 8 weeks during two
29 different years (to different plants). Ozone treatment reduced the growth and relative growth
30 rate (by about 15 to 20% for tops and 20 to 40% for roots) of plants before cutting, when
31 compared to a filtered-air control. The growth of roots was more affected than the growth of
December 1993 5-185 DRAFT-DO NOT QUOTE OR CITE
-------
1 tops, with a shift in the allocation pattern. In the second year of the study, 03 exposure was
2 continued after the plants were harvested and the impact of exposure on regrowth was
3 determined. In this case, they found that the relative growth rate in 63 exposed plants was
4 higher, perhaps because of an increased demand for carbon by the root systems of the
5 Orstressed plants. It is unclear whether these plants would sustain their increased growth,
6 and in fact, the authors speculate that the increased growth, in lieu of partitioning carbon to
7 other compounds, might alter the cold hardiness of the plants.
8 Ozone has been demonstrated to affect the growth of field grown alfalfa. Temple et al.
9 (1988a) reported a two-year study of alfalfa in which O3 at ambient concentrations (0.049 in
10 1984 and 0.042 ppm in 1985 for the seasonal 12-h means April to October) did not affect the
11 growth and yield of the plants, but at 12 h seasonal means of 0.063 and 0.078 ppm, yield
12 was reduced by about 15 and 19%. The exposure-response functions for the two years were
13 homogeneous; there was no indication of cumulative effect of 03 exposure, however, crown
14 weight, an indicator of health and vigor, of exposed plants was significantly reduced.
15 In a different field experiment conducted to determine the interactive effects of O3 and
16 simulated acid fog on stomatal conductance, photosynthesis, foliar injury, and yield of an
17 established stand of alfalfa, plants were exposed 12 h daily for 4 weeks (Temple et al.,
18 1987). Ozone was added in proportion to its concentration in the ambient air. Ambient
19 O3 concentrations during the experiment were 0.043 ppm. Ozone injury symptoms appeared
20 on the alfalfa exposed to 0.098 ppm (NF x 2.0), one week after the start of the regrowth
21 period. When exposures were at 0.081 and 0.066 ppm (NF x 1.7 and NF x 1.3), more
22 than a week was required for injury to appear. A one month exposure of the plants at the
23 end of the growing season resulted in a reduction of about 2.5 % in aboveground yield at a
24 12 h seasonal mean concentration of 0.04 ppm. At a concentration of 0.066 ppm, the
25 exposure resulted in a reduction in yield of approximately 18%. It should be noted that the
26 whole plant was exposed to ambient O3 for the growing season, only new leaves that had
27 developed after harvest received the one month exposure. Ozone exposures could shorten
28 the productive life of alfalfa stands in addition to its effect on yield.
29 The sensitivity of alfalfa to O3 had been demonstrated in an earlier study using filtered
30 air in field chambers (Neely et al., 1977; U.S. Environmental Protection Agency, 1986).
31 There was a 49% decrease in top dry weight when plants were exposed to 0.05 ppm,
December 1993 5-186 DRAFT-DO NOT QUOTE OR CITE
-------
1 7-h/day for 68 days. Increasing the concentration to 0.1 ppm produced a 51 % loss in dry
2 weight at harvest.
3 Kohul el ;il. (1988) and Hcaglc el al. (1989) experimented with forage mixtures
4 characteristic of the northeast and southeast, respectively. In both cases, exposure to
5 O3 resulted in a reduction in total forage yield of about 10 to 20% at 12 h seasonal mean
6 O3 concentrations of 0.045 to 0.05 ppm. In both cases, the clover component of the mix
7 was more sensitive than the grass, and was reduced in prevalence in the stand. The
8 relevance of these studies to competition and species composition is discussed in the
9 ecosystem response section (Section 5.7).
10 Results of studies or perennial plants conducted since 1986 are summarized in
11 Table 5-23. As with single-season agricultural crops, yields of multiple-year forage crops
12 are reduced at concentrations at, or near ambient (0.04 to 0.06 ppm) in many parts of the
13 country (Table 5-17).
14
15 5.6.4.2 Effects of Ozone on Deciduous Shrubs and Trees
16 Most of the information concerning the response of deciduous shrubs and trees to
17 episodes, season-long, or multiple-year exposures to O3 is based on field observations. The
18 longevity of perennial plants, and the size in the case of trees, makes experimental their
19 study under experimental conditions difficult. For this reason, there is little experimental
20 data concerning the response of deciduous shrubs and trees.
21 Most of the hardwood experiments included in Reich's analysis (1987), for example,
22 were exposed under laboratory or greenhouse conditions to relatively high concentrations for
23 short periods of time. Although exposures durations of weeks were conducted, square-wave
24 exposure regimes that do not capture important characteristics of ambient exposure were
25 used. In addition, in Pye (1988), the majority of the studies were conducted in laboratory or
26 greenhouse. The results of a few open-top chamber studies are cited, however, the majority
27 of these studies used O3 concentrations 0.10 ppm or higher. While the studies reported in
28 Table 6-23 of U.S. Environmental Protection Agency (1986) (see Table 5-16) document the
29 sensitivity of the seedlings of some species grown in chambers, little information of value
30 with regard to tree growth or biomass production in the long-term can be extrapolated from
December 1993 5.187 DRAFT-DO NOT QUOTE OR CITE
-------
December 1993
(j\
i
oo
oo
3
6
o
§
2
M
CJ
i
tn
8
O
TABLE 5-23. A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
OZONE ON THE GROWTH, PRODUCTIVITY, OR YIELD OF PERENNIAL CROP PLANTS
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
Species
Strawberry
Phleum pratense
Dactylis glomerata
Poa pratensis
Festuca rubra
Festuca pratensis
Agrostis tenuis
Lolium perenne
Trifolium pratense
Plantago major
Red clover
E/D/
PC/Ca
PC
PC
PC
PC
PC
PC
PC
PC
PC
PC
PC
Concentration
18 to 66 ppb 12 h mean
10 to 55 ppb 7 h mean
10 to 55 ppb 7 h mean
10 to 55 ppb 7 h mean
10 to 55 ppb 7 h mean
10 to 55 ppb 7 h mean
10 to 55 ppb 7 h mean
62 ppb 7 h mean
6 to 59 ppb 7 h mean
70 ppb 7 h mean
19 to 62 ppb 12 h mean
Duration
1 1 weeks
5 weeks
5 weeks
5 weeks
5 weeks
5 weeks
5 weeks
5 weeks
5 weeks
8 weeks
83 and
91 days in
2 growing
seasons
Facility0
OTC
GC
GC
GC
GC
GC
GC
GC
GC
GC
OTC
Variable
fresh fruit
weight
shoot dry
weight
shoot dry
weight
shoot dry
weight
shoot dry
weight
shoot dry
weight
shoot dry
weight
shoot dry
weight
shoot dry
weight
total dry
weight
dry
weight
Effect6
20% increase in fruit
weight at 66 ppb.
45% reduction at 55 ppb.
28% reduction at 55 ppb.
28% reduction at 55 ppb.
23% reduction at 55 ppb.
16% reduction at 55 ppb.
No effect.
No effect.
30% reduction at 59 ppb.
Reduced up to 36 %
depending on growth stage.
1 1 % reduction at 62 ppb.
Reference
Takemoto et al.
(1988c)
Mortensen
(1992a)
Mortensen
(1992a)
Mortensen
(1992a)
Mortensen
(1992a)
Mortensen
(1992a)
Mortensen
(1992a)
Mortensen
(1992a)
Mortensen
(1992a)
Reiling and
Davison (1992)
Kohut et al.
(1988)
-------
CD
1
u>
Species
TABLE 5-23 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
OZONE ON THE GROWTH, OR PRODUCTIVITY OF PERENNIAL CROP PLANTS
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
E/D/
PC/C"
Concentration
Duration
Facility0 Variable
Effect6
Reference
Timothy
Ladino clover tall
fescue pasture
Ladino clover
± Alfalfa
-------
O
S.
I
W
Variable
Duration
§
o
§
U
D.
c/3
oo
tv,
O
2-
y
growing
asons
o
u
I*
O
8
-2
u
"
8 -a 8 "8
o «a
. 43
S C
December 1993
5-190 DRAFT-DO NOT QUOTE OR CITE
-------
1 the experiments. Trees, because of their size, are difficult to study under controlled
2 conditions, therefore, most experiments have used seedlings in pots or in open top chambers.
3 Since 1986, a number of studies have been conducted documenting the sensitivity of
4 hardwoods to O3 (Table 5-24). Some species, such as black cherry, are very sensitive
5 (Davis and Skelly, 1992, 1992b; Simini et al., 1992), with SUM06 exposures as low as
6 12.9 ppnvh over 92 days (concentrations not given) predicted to cause a 10% yield loss
7 (Hogsett et al., 1993; Table 5-25). Based on studies previously reviewed, the growth of
8 some hardwood species, particularly those of the genus Populus, may be affected by ambient
9 concentrations of O3 (U.S. Environmental Protection Agency, 1978, 1986).
10 In a studies of the response of aspen clones to O3 at two field sites in Michigan,
11 Karnosky et al. (1992a,b) documented reductions in stem weight of up to 46% in sensitive
12 aspen clones after 70 days of exposure in open-top chambers to 0.08 ppm for 6 h per day,
13 3 days per week.
14 Tjoelker and Luxmoore (1991) found leaf abscission on tulip poplar seedlings to be
15 increased by exposure to a 7 h seasonal mean concentration of 0.108 ppm, resulting in a
16 doubling of the leaf turnover rate, but this was not translated into an effect on growth,
17 perhaps due to the indeterminate growth habit of the plant. In such plants, leaf production
18 continues throughout the growing season, which may permit the tree to maintain an optimal
19 leaf area. However, continued leaf growth could deplete carbon or nitrogen reserves.
20 Samuelson and Edwards (1993) found canopy weight of 30 year old northen red oak,
21 exposed in large open-top chambers, to be reduced by 41 % after exposure for 177 days at a
22 7-h seasonal mean of 0.069 ppm (87 ppnvh SUMOO) compared to a sub-ambient treatment at
23 a 7-h seasonal mean of 0.015 ppm (18 ppnvh SUMOO) a concentration found nowhere in the
24 world. Two-year old seedlings were not affected by similar exposures.
25 Hogsett et al. (1993) have developed exposure-response functions using uncited data for
26 aspen, red alder, black cherry, red maple, and tulip poplar (Table 5-25), as well as
27 composite functions for deciduous tree seedlings (Table 5-26). Their results show that, for
28 28 deciduous seedling cases, a SUM06 exposure of 31.5 ppm-h over 92 days a mean
29 concentration of approximately 0.055 ppm) should result in less than a 10% yield (biomass)
30 reduction in 50% of the cases. A 20% reduction in yield should result from a SUM06
December 1993 5-191 DRAFT-DO NOT QUOTE OR CITE
-------
Almond
u. Plum
TABLE 5-24. A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
OZONE ON THE GROWTH OR PRODUCTIVITY OF DECIDUOUS SHRUBS AND TREES
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
SJ
1— »
£
W
Species
Almond
Almond
E/D/
PC/Ca
D/PC
D/PC
Concentration
38 to 112 ppb 12 h mean
30 to 1 17 ppb 12 h mean
Duration
153 days
3.5 mo
Facility
OTC
OTC
Variable
total dry
weight
cross
sectional
Effect6
Linear reduction in
2 cultivars, no effect in
three.
6% reduction at 51 ppb.
Reference
Retzlaff et al.
(1992a)
Retzlaff et al.
(1991a)
Plum
Pear
Apricot
Rhus trilobata
area
16 weeks in
each of
2 growing
D/PC 250 ppb 4 n/wk
D/PC 44 to 111 ppb 12 h mean 191 and
213 days
D/PC 30 to 117 ppb 12 h mean 3.5 mo
D/PC 30 to 117 ppb 12 h mean 3.5 mo
D/PC 30 to 117 ppb 12 h mean 3.5 mo
D 10 to 75 ppb 12 h mean 3 mo
CC net growth 28 and 36% reduction in
year 1 and 2.
OTC number of 29% fewer fruit at ambient
fruit per and above.
tree
OTC cross 19% reduction at 51 ppb.
sectional
area
OTC cross 8% reduction at 51 ppb.
sectional
area
OTC cross 53% reduction at 117 ppb.
sectional
area
OTC in growth Increase in leaf weight in
pots ambient air. No other
effect.
McCool and
Musselman (1990)
Retzlaff et al.
(1992b)
Retzlaff et al.
(1991b)
Retzlaff et al.
(1991b)
Retzlaff et al.
(1991b)
Temple (1989)
-------
TABLE 5-24 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
1
I— >
VO
t— '
1
1
O
1
O
<-!
OZONE ON THE GROWTH OR PRODUCTIVITY OF DECIDUOUS SHRUBS AND TREES
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
Species
Black Cherry
Black Cherry
Red Oak
Red Oak
Red Oak
Red Maple
Red Maple
Tulip Poplar
E/D/
PC/Ca
D
D
D
D
D
D
D
D
Concentration
16 to 67 ppb 12 h mean
40 or 80 ppb 7 h/day
5 days/week
18 to 87 ppnvh 15 to
69 ppb 7-h mean
16 to 67 ppb 12 h mean
40 or 80 ppb 7 h/day
5 days/week
16 to 67 ppb 12 h mean
40 or 80 ppb 7 h/day
5 days/week
16 to 67 ppb 12 h mean
Duration
3 growing
seasons
8 or 12 weeks
177 days
3 growing
seasons
8 or 12 weeks
3 growing
seasons
8 or 12 weeks
3 growing
seasons
Facility0 Variable*1
OTC growth
and leaf
dynamics
GC growth
OTC tree
canopy
OTC growth
and leaf
dynamics
GC growth
OTC growth
and leaf
dynamics
GC growth
OTC growth
and leaf
dynamics
Effect6
Leaf abscission increased
with increasing ozone.
Reduced leaf, stem, and
root dry weight, and height
at 80 ppb.
Reduced 41 % at 82 ppnvh
or 69 ppb 7-h mean.
No effect.
Reduced root dry weight at
80 ppb.
No effect.
Reduced stem diameter and
dry weight at 80 ppb.
Leaf abscission increased
with increasing ozone.
Reference
Simini et al.
(1992)
Davis and Skelly
(1992)
Samuelson and
Edwards (1993)
Simini et al.
(1992)
Davis and Skelly
(1992)
Simini et al.
(1992)
Davis and Skelly
(1992)
Simini et al.
(1992)
-------
TABLE 5-24 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
3
i-»
u>
OZONE ON THE GROWTH OR PRODUCTIVITY OF DECIDUOUS SHRUBS AND TREES
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
Species
Yellow Poplar
European Beech
Aspen
Aspen
Aspen
Yellow Poplar
Paper Birch
E/D/
PC/C* Concentration11
D 40 or 80 ppb 7 h/day
5 days/week
D 10 to 90 ppb weekly
mean
D 80 ppb 6 h/day
3 days/week
D filtered air or 80 ppb
6 h/day 3 days/week
D ambient* 27 51, or
102 ppb exposure period
mean
D 0 to 200 ppb 8 h/day
3 days/week
D 60 to 80 ppb 7 h/day
5 days/week
Duration
8 or 12 weeks
5 years
70 and
92 days in
2 growing
seasons
93 days at
2 sites in
Michigan
10S days
4.5 mo
12 weeks
Facility0 Variable*1
GC growth
OTC growth
OTC stem
weight
OTC growth
CC dry
weight
GC growth
GH dry
weight
Effect6
Reduced leaf dry weight
and stem diameter at
80 ppb.
Reduced shoot growth and
leaf area.
No effect on tolerant clones
46% reduction for sensitive
clones in one year 5%
(tolerant) and 74%
(sensitive) reductions in the
second year.
18 to 26% reduction in
diameter growth.
40% reduction— 44%
reduction in early growth
the following year.
Up to a 24% reduction at
200 ppb, but moderated by
pH treatment.
Decreased shoot and root
weight and leaf area.
Reference
Davis and Skelly
(1992)
Billen et al.
(1990)
Karnosky et al.
(1992b)
Karnosky et al.
(1992a)
Keller (1988)
Jensen and Patton
(1990)
Keane and
Manning (1988)
-------
TABLE 5-24 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
I
VO
VO
OZONE ON THE GROWTH OR PRODUCTIVITY OF DECIDUOUS SHRUBS AND TREES
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
Species
Betula pubescens
Betula pubescens
Alnus incana
E/D/
PC/Ca
D
D
D
Concentration
25 to 82 ppb 7 h/day
25 to 82 ppb 7 h/day
25 to 82 ppb 7 h/day
Duration Facility0 Variable
50 days GC dry
weight
50 days GC dry
weight
50 days GC dry
weight
Effect*
Shoot and root dry weight
decreased linearly with
ozone.
Shoot and root dry weight
decreased linearly with
ozone.
Shoot and root dry weight
decreased linearly with
ozone.
Reference
Mortensen and
Skre (1990)
Mortensen and
Skre (1990)
Mortensen and
Skre (1990)
V aE = evergreen, D = deciduous, C = crop, PC = perennial crop.
vo bMeans are seasonal means unless specified. Maximums are 1 h seasonal maxima unless otherwise specified. Cumulative exposures are SUMOO unless
otherwise specified, accumulation based on 24 h/day unless otherwise noted.
°OTC = open-top chamber with plants in ground unless specified in pots; CC = closed chamber, outside; GC = controlled environment growth chamber or
CSTR; GH = greenhouse; F = field; OF = open-field fumigation.
O ^Ths effect reported in the study mat is a measure of growth, yield, or productivity.
5 Effect measured at specified ozone concentration, over the range specified under concentration, or predicted (if specified) to occur based on relationships developed
2 in the experiment.
-------
TABLE 5-25. EXPOSURE-RESPONSE EQUATIONS THAT RELATE TOTAL
BIOMASS (FOLIAGE, STEM, AND ROOT) TO 24-HOUR SUM06 EXPOSURES (C)
ADJUSTED TO 92 DAYS (ppm-h/year)
Rate of
Growth
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
SLOW
FAST
FAST
FAST
Habit Study Species
D
D
D
D
D
D
D
D
D
D
D
D
D
D
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
D
D
D
D
D
D
D
D
D
D
D
D
1
1
2
2
3
3
4
4
4
4
5
5
5
6
7
7
7
7
8
8
8
9
9
10
10
10
10
11
11
11
12
13
14
15
15
16
16
17
18
19
20
21
21
22
Aspen - wild
Aspen - wild
Aspen - wild
Aspen - wild
Aspen - wild
Aspen - wild
Aspen 21 6
Aspen 253
Aspen 259
Aspen 271
Aspen 216
Aspen 259
Aspen 271
Aspen - wild
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Red Alder
Red Alder
Red Alder
Red Alder
Red Alder
Red Alder
Black Cherry
Black Cherry
Red Maple
Tulip Poplar
Tulip Poplar
Tulip Poplar
SUM06 for Loss
Location Exposure1* Weibull Parameters of
(State) Days Year Harvests* ABC 10% 30%
OR
OR
OR
OR
OR
OR
MI
MI
MI
MI
MI
MI
MI
MI
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
OR
TN
TN
TN
TN
TN
TN
84
84
118
118
112
112
82
82
82
82
98
98
98
98
113
113
234
234
118
118
230
111
111
113
113
234
234
118
118
230
140
84
121
113
113
118
118
112
76
140
55
75
184
81
1989
1989
1991
1991
1990
1990
1990
1990
1990
1990
1991
1991
1991
1991
1989-90
1989-90
1989-90
1989-90
1991-92
1991-92
1991-92
1989
1989
1989-90
1989-90
1989-90
1989-90
1991-92
1991-92
1991-92
1992
1991
1990
1989
1989
1991
1991
1992
1989
1992
1988
1990-91
1990-91
1992
1
2
1
2
1
2
1
1
1
1
1
1
1
1
1
2
3
4
1
2
3
1
2
1
2
3
4
1
2
3
1
1
1
1
2
1
2
1
1
1
1
1
3
1
9.9
17.7
31.0
75.6
67.8
96.9
54.5
73.1
79.1
91.3
37.4
35.2
35.7
19.0
16.8
27.9
33.3
83.5
26.7
85.9
119.1
12.8
25.8
12.9
25.7
32.1
90.1
20.2
47.1
44.5
134.6
136.0
42.4
84.4
206.8
63.5
248.8
54.1
53.7
37.1
28.5
45.8
334.1
150.1
96.3
165.2
130.0
124.9
111.0
142.1
121.1
265.5
92.7
44.9
128.6
95.9
73.1
263.1
462.7
3.8E+17
438.9
2887.0
109.5
-0.0058
218.7
246.9
365.2
233.7
358.8
327.8
634.3
266.4
206.5
458.5
235.8
442.8
217.0
253.0
179.9
501.7
2.0E+13
274.4
79.1
176.6
387.1
46.4
623.5
50.8
1.316 19.09
1.000 19.06
3.062 48.62
5.529 64.80
6.532 64.60
1.257 19.48
1.609 33.56
1.000 31.38
1.000 10.96
8.964 39.20
1.000 12.72
1.000 9.49
4.012 39.16
1.000 26.02
1.844 111.17
1.000 250.00
5.383 113.61
1.000 119.61
57.655 82.13
(lin) 250.00
12.254 72.80
1.000 21.56
1.000 31.89
1.000 20.05
1.000 30.77
1.000 13.58
1.000 26.27
1.000 21.88
1.000 16.96
1.257 30.61
2.570 64.56
1.000 51.10
1.427 34.08
1.000 21.70
5.294 95.76
1.000 41.21
1.000 250.00
1.107 29.50
1.123 12.91
1.168 16.90
1.537 149.75
4.518 34.56
1.000 32.85
1.852 17.12
48.21
64.54
72.41
80.79
77.86
51.40
71.60
106.23
37.10
44.91
43.06
32.11
53.07
88.08
215.37
250.00
142.49
404.91
83.88
250.00
80.42
73.00
107.95
67.87
104.18
45.97
88.94
74.09
57.42
80.77
103.76
172.98
80.10
73.46
120.57
139.51
250.00
88.79
38.23
48.00
331.07
45.27
111.19
33.07
December 1993
5-196 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-25 (cont'd). EXPOSURE-RESPONSE EQUATIONS THAT RELATE
TOTAL BIOMASS (FOLIAGE, STEM, AND ROOT) TO 24-HOUR SUM06
EXPOSURES (C) ADJUSTED TO 92 DAYS (ppm-h/year)
Rate of
Growth
FAST
FAST
SLOW
SLOW
SLOW
SLOW
SLOW
Habit Study Species
E
E
D
D
E
E
E
23
23
24
24
25
25
26
Loblolly GAKR 15-91
Loblolly GAKR 15-23
Sugar Maple
Sugar Maple
E. White Pine
E. White Pine
Virginia Pine
SUM06 for Loss
Location Exposure1" Weibu» Parameters of6
(State) Days Year Harvests* A B C 10% 30%
AL
AL
MI
MI
MI
MI
MI
555
555
83
180
83
180
98
1988-89
1988-89
1990-91
1990-91
1990-91
1990-91
1992
3
3
1
3
1
3
1
22.7
20.4
4.12
24.63
0.35
1.21
78.3
4,402.5
13,125.4
100.0
110.2
63.1
719.5
3,045.1
1.000 76.89
1.000 229.24
40.069 104.79
5.987 38.68
4.191 40.90
1.000 38.74
1.000250.00
260.30
250.00
108.03
47.42
54.72
131.16
250.00
Harvest 1 occurs immediately following end of first year of exposure. Harvest 2 occurs in spring following first year of exposure.
Harvest 3 occurs immediately following end of second year of exposure. Harvest 4 occurs in spring following second year of exposure.
Duration corresponds to the length in days of the first year of exposure for Harvests 1 and 2 and to the total length of the first and
second years of exposure for Harvests 3 and 4.
"To compare the results from seedling studies of varying exposure duration," the SUM06 value is calculated for an exposure of fixed
period of 92 days per year. "For example, Study 1 Harvest 1 has an exposure duration of 84 days and a" SUM06 value of 19.09 ppm-h
over 92 days which corresponds to a SUM06 value "of 19.09*84/92 = 17.43 ppm-h over 84 days, at which biomass loss is 10%." The
calculation assumes that exposures can be scaled up or down in uniform fashion.
Based on GIS, TREGRO and ZELIG models projections. No data given in paper.
Source: Hogsett et al. (1993).
1 exposure of greater than 52.1 ppm-h. Comparison with Table 5-17 shows a SUM06 for
2 3 mo of 29.5 ppm-h at ambient concentrations, a value near that (33.3 ppm-h) expected to
3 prevent a 10% yield reduction in 50% of the cases Table 5-24). An individual year, such as
4 1988, might be significantly above the no-injury exposure value (Table 5-17). By further
5 grouping the seedlings by rate of growth (fast or slow), they were able to refine estimates of
6 the SUM06 exposure that would protect seedlings based on growth strategy. Deciduous
7 seedlings and/or fast growing species are more sensitive than evergreen and/or slow growing
8 seedlings (Table 5-24).
9 The response of a number of fruit and nut trees to 03 has been reported (McCool and
10 Musselman, 1990; Retzlaff et al., 1991, 1992a,b). Almond has been identified as the most
11 sensitive, but peach, apricot, pear, and plum have also been affected. Net growth of
12 almond, as well as stem diameter of peach and the stem diameter and number of shoots
13 produced on apricot were reduced by four months (the exposure duration specified by the
14 authors) of once-weekly exposure to 0.25 ppm for 4 h, a relatively small exposure
15 cumulatively (16 ppm h as a SUMOO or as a SUM 06) (McCool and Musselman, 1990) but
December 1993 5_197 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-26. SUM06 LEVELS ASSOCIATED WITH 10 AND 20% TOTAL
BIOMASS LOSS FOR 50 AND 75% OF THE SEEDLING STUDIES
(The SUM06 value is adjusted to an exposure length of 92 days per year)
Weibull Equations (all 51 seedling studies):
50th Percentile PRYL = 1 - exp(-[SUM06/176.342]**l.34962)
75th Percentile PRYL = 1 - exp(-[SUM06/104.281]**1.46719)
Weibull equations (27 fast-growing seedling studies):
50th Percentile PRYL = 1 - exp(-[SUM06/150.636]**l.43220)
75th Percentile PRYL = 1 - exp(-[SUM06/89.983]**l.49261)
Weibull equations (24 slow to moderate growing seedling studies):
501 h Percentile PRYL = I - exp(-[SUM06/190.900]**l.49986)
75th Percentile PRYL = 1 - exp(-[SUM06/172.443]**l. 14634)
Weibull equations (28 deciduous seedling studies):
50th Percentile PRYL = 1 - exp(-[SUM06/142.709]**l.48845)
75th Percentile PRYL = 1 - exp(-[SUM06/87.724]**1.53324)
Weibull equations (23 evergreen seedling studies):
50th Percentile PRYL = 1 - exp(-[SUM06/262.91 !]**!.23673)
75th Percentile PRYL = 1 - exp(-[SUM06/201.372]**1.01470)
Levels associated with prevention of a 10% and 20% total biomass loss for 50% and 75% of the seedlings.
All 51 seedling cases
Percent of seedlings
50% 75%
Relative 10% 33.3 11.4
Biomass Loss 20% 58.0 23.8
27 fast-growing seedling cases
Percent of seedlings
50% 75%
Relative 10% 31.3 19.4
Biomass Loss 20% 52.9 32.4
24 slow to moderate growth seedling cases
Percent of seedlings
50% 75%
Relative 10% 42.6 24.2
Biomass Loss 20% 70.2 46.6
28 deciduous seedling cases
Percent of seedlings
50% 75%
Relative 10% 31.5 20.2
Biomass Loss 20% 52.1 33.0
December 1993 5-198 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-26 (cont'd). SUM06 LEVELS ASSOCIATED WITH 10 AND 20% TOTAL
BIOMASS LOSS FOR 50 AND 75% OF THE SEEDLING STUDIES
(The SUM06 value is adjusted to an exposure length of 92 days per year)
23 Evergreen Seedling Cases
Percent of seedlings
50% 75%
Relative 10% 42.6 21.9
Biomass Loss 20% 78.2 45.9
aBased on CIS, TREGRO and ZELJG model projections. No data given in paper.
Hogsett et al. (1993).
1 one with a high peak value. Cross-sectional area of almond, plum, apricot, and pear sterns
2 decreased linearly with increasing O3, with a significant reduction at a 12 h seasonal mean of
3 0.051; dry weight of roots, trunk, and foliage was also reduced in one variety (Retzlaff
4 etal., 1991, 1992a,b).
5 Finally, two studies report the response of citrus and avocado to 03 (Eissenstat et al.,
6 1991: Olszyk et al., 1990). These species retain their leaves for more than one year, but fit
7 best in the deciduous category because, while evergreen, leaves are replaced more frequently
8 than in most evergreen species. Valencia orange trees (during a production year) exposed to
9 a seasonal 12 h mean of 0.04 or 0.075 ppm had 11 and 31% lower yields than trees grown
10 in filtered air at 0.012 ppm and atypical concentration. During an off-production year, yield
11 was not affected. Growth of Ruby Red grapefruit was not affected by concentrations of
12 3 times ambient (Eissenstat et al., 1991). Avocado growth was reduced by 20 or 61 % by
13 exposure during two growing seasons at 12 h seasonal mean concentrations of 0.068 and
14 0.096.
15 In summary, deciduous trees appear to be less sensitive to O3 than most crop plants,
16 but there are species that are as, or more sensitive due to their genetic composition, than
17 crops (e.g., Populus species, and perhaps black cherry—see discussion in Section 5.4.2).
18 Analysis of the crop data presented in Table 5-24 and discussed above suggests that a
19 7 h seasonal mean exposure of approximately 0.055 ppm over a three month period would
20 prevent injury to tree seedlings. However, the absence of multiple-year studies, or studies
December 1993 5499 DRAFT-DO NOT QUOTE OR CITE
-------
1 using older, more mature trees leaves unanswered the question of long-term and cumulative
2 effects.
3
4 5.6.4.3 Effects of Ozone on Evergreen Trees
5 As with hardwoods, little long-term data from controlled studies was available at the
6 time the literature was reviewed for U.S. Environmental Protection Agency (1986). The
7 1986 document did point out, however, that studies conducted on eastern white pine on the
8 Cumberland Plateau in Tennessee indicated that ambient O3 may have reduced the radial
9 growth of sensitive individuals as much as 30 to 50% annually over a period of 15 to
10 20 years (Mann et al., 1980). Also, field studies in the San Bernardino National Forest
11 indicated that over a period of 30 years O3 may have reduced the growth in height of
12 ponderosa pine by as much as 25%, radial growth by 37%, and total volume of wood
13 produced by 84% (Miller et al., 1982). Calculations of biomass in these studies were based
14 on apparent reductions in radial growth without standardization of the radial growth data with
15 respect to tree age. Since 1986, studies on the effects of O3 on evergreen trees have focused
16 primarily on three species or groups: red spruce in the eastern United States, southern pines
17 (loblolly and slash), and western conifers (primarily ponderosa pine). For the most part, the
18 research has been conducted with tree seedlings or saplings, and has involved exposures
19 lasting one to four growing seasons. In many cases, the research has concentrated on
20 defining the mode of action of O3 in conifers, and is discussed elsewhere in this document
21 (Section 5.3). Results of studies with evergreen trees are summarized in Table 5-27.
22 Hogsett et al. (1993) summarized the results from uncited data and produced composite
23 exposure-response functions for douglas-fir, ponderosa pine, eastern white pine, Virginia
24 pine, and loblolly pine (Table 5-25). They predict that a 10% loss in 50% of the
25 experimental cases could have been prevented by a SUM06 exposure of 42.6 ppm-h
26 (O3 concentrations and exposure times were not given).
27 Studies of the response of red spruce to O3 exposures, regardless of whether they have
28 been conducted in growth chambers (Lee et al., 1990a,b; Patton et al., 1991; Taylor et al.,
29 1986), or in the field (Kohut et al., 1990; Laurence et al., 1993b; Thornton et al., 1992)
30 have failed to detect effects on growth of seedlings or saplings, even after exposure to
31 12 h seasonal means of up to approximately 0.09 ppm each year for up to four years;
December 1993 5-200 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-27. A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
8
i-i
o
O
6
0
1
0
OZONE ON THE GROWTH OR PRODUCTIVITY OF EVERGREEN TREES
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
E/D/
Species PC/C Concentration
Avocado E/PC 10 to 108 ppb 12 h mean
Orange E/PC 10 to 108 ppb 12 h mean
Orange E/PC 12 to 75 ppb 12 h mean
Ponderosa Pine E ambient
Ponderosa Pine E 13 to 95 ppb 12 h mean 47
to 350 pprn-h over 3 years
Ponderosa Pine E 1 1 to 87 ppb 12 h mean
Ponderosa Pine E 5, 122 or 169 ppm-h
Ponderosa Pine E 67 to 71 ppb 7 h mean
Duration
4 and 8 mo in
2 growing
seasons
4 and 8 mo in
2 growing
seasons
7 mo/season
for 5 years
3 growing
seasons
3 growing
seasons
1 12 days
134 days
Facility0
OTCin
pots
OTCin
pots
OTC
F
OTC
OTC
OTCin
pots
OTCin
pots
Variable
leaf mass
leaf mass
fruit weight
radial
growth rate
growth
leaf weight
root growth
leaf, stem
and root
dry weight
Effect6
20 and 61 % reduction in leaf
mass at 86 and 108 ppb.
No effect.
"On" production year -
1 1 and 3 1 % reduction at
40 and 75 ppb "off year -
no effect.
No change in growth rate on
symptomatic trees.
19.5% reduction at 95 ppb.
70 and 48% loss of 2 and
1 year old needles at 87 ppb.
43 % reduction in coarse and
fine non-growing roots; 50,
65 and 62% reduction in
coarse, fine and new growing
roots.
20 to 33% reduction from
filtered air at 67 ppb.
Reference
Eissenstat et al.
(1991)
Eissenstat et al.
(1991)
Olszyk et al.
(1990)
Peterson and
Arbaugh (1988)
Beyers et al.
(1992)
Temple et al.
(1993)
Andersen et al.
(1991)
Hogsett et al.
(1989)
-------
TABLE 5-27 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
OZONE ON THE GROWTH OR PRODUCTIVITY OF EVERGREEN TREES
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
o
»— t
w>
to
8
o
E
i
2
o
H
^^
3
H
tn
O
*3
n
!-1
Species
Lodgepole Pine
Jeffrey Pine
Jeffrey Pine
Western Hemlock
Western Red Cedar
Douglas Fir
Giant Sequoia
Red Spruce
Red Spruce
E/D/
PC/Ca
E
E
E
E
E
E
E
E
E
Concentration
67 to 71 ppb 7 h mean
0-200 ppb 4 h/day
3 days/week
ambient
67 to 71 ppb 7 h mean
67 to 71 ppb 7 h mean
67 to 71 ppb 7 h mean
0-200 ppb 4 h/day
3 days/week
8 to 166 ppb 8 h mean
8 to 156 ppm-h
23 to 87 ppb 12 h mean
Duration
134 days
44 and
58 days in
2 growing
seasons
—
134 days
134 days
134 days
44 and
58 days in
2 growing
seasons
135 days
2 growing
seasons
Facility
OTCin
pots
GC
F
OTCin
pots
OTCin
pots
OTCin
pots
GC
OTC
OTCin
pots
Variable
leaf, stem
and rot
dry weight
root, stem
and
needles
dry weight
radial
growth
leaf, stem
and root
dry weight
leaf, stem
and root
dry weight
leaf, stem
and root
dry weight
root, stem
and
needles
dry weight
scion
growth
dry weight
Effect6
No effect.
Reduced 10-20% ppb in one
year.
1 1 % reduction in
symptomatic trees.
1 1 to 305 reduction at
71 ppb.
No effect.
No effect.
No effect.
No effect on juvenile or
mature scion growth.
No effect.
Reference
Hogsett et al.
(1989)
Temple (1988)
Peterson et al.
(1987)
Hogsett et al.
(1989)
Hogsett et al.
(1989)
Hogsett et al.
(1988)
Temple (1988)
Rebbeck et al.
(1992)
Kohut et al.
(1990)
-------
TABLE 5-27 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
CD
!
H- '
%
OJ
Ul
1
to
o
OJ
o
C
3
6
o
1
3
0
g
1
w
s
r ^
O
OZONE ON THE GROWTH OR PRODUCTIVITY OF EVERGREEN TREES
PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
Species
Red Spruce
Red Spruce
Red Spruce
Red Spruce
Norway Spruce
Norway Spruce
Norway Spruce
Sitka Spruce
Silver Fir
Fraser Fir
White Pine
Loblolly Pine
E/D/
PC/Ca
E
E
E
E
E
E
E
E
E
E
E
E
Concentration
120 ppb 4 h/day
twice/week
0, 150 ppb 6 h/day or 150
ppb 6 h plus 70 ppb 18
h/day
25 or 100 ppb 4 h/day
3 day /week
27 to 54 ppb 12 h mean
80 to 100 ppb 7 to 8 h/day
14 to 70 ppb 8 h mean
10 to 90 ppb weekly mean
5 to 170 ppb 7 h/day
5 days/week
10 to 90 ppb weekly mean
20 to 100 ppb 4 h/day
3 times/week
20 to 140 ppb 7 h/day
3 day/week
21 to 86 ppb 7 h mean
Duration
4 mo
195 days
10 weeks
3 growing
seasons
100 days
5 to 6 mo in 2
growing
seasons
5 years
65 days
5 years
10 weeks
3.5 mo
96 days
Facility0
GC
GC
GC
OTCin
pots
GC
OTCin
pots
OTC
GH
OTC
GC
GC
OTCin
pots
Variable*1
growth
dry weight
growth
dry weight,
diameter,
height
dry weight
growth
growth
growth and
winter
hardiness
growth
biomass
dry weight
dry weight
Effect6
No effect.
No effect.
No effect.
No effect.
0-14% reduction vs. filtered
air - 5 provenances.
No effect.
Reduced lateral shoot growth
in last year.
No effect on growth, reduced
winter hardiness.
Increased dry matter
production.
No effect.
No effect.
18% reduction at 86 ppb
20% reduction in foliage at
40 or 86 ppb.
Reference
Taylor et al. (1986)
Patton et al. (1991)
Lee et al. (1990b)
Thornton et al.
(1992)
Mortensen (1990a)
Nast et al. (1993)
Billen et al. (1990)
Lucas et al. (1988)
Billen et al. (1990)
Tseng et al. (1988)
Reich et al. (1987)
Adams et al.
(1988)
-------
TABLE 5-27 (cont'd). A SUMMARY OF STUDEES REPORTING THE EFFECTS OF
OZONE ON THE GROWTH OR PRODUCTIVITY OR EVERGREEN TREES
cr
Co
w
1
to
g
^
>
H
6
O
<-*
•^
H
O
C^
O
PUBLISHED SINCE U.S.
Species
Loblolly Pine
Loblolly Pine
Loblolly Pine
Loblolly Pine
Loblolly Pine
Loblolly Pine
Loblolly Pine
Loblolly Pine
Loblolly Pine
E/D/
PC/Ca
E
E
E
E
E
E
E
E
E
Concentration
21 to 117 ppb 7 h mean
22 to 94 ppb 7 h mean
32 to 108 ppb 7 h mean
23 to 90 ppb 12 h mean 46
to 209 max 12 h
22 to 92 ppb 12 h mean 37
to 143 ppb 1 h maximum
7 to 166 ppb 12 h mean 12
h maximum 248 ppb
7 to 132 ppb 12 h mean
17 to 382 ppm-h
21 to 137 ppb 12 h mean
60 to 397 ppm-h
20 to 137 ppb 12 h mean
50 to 286 ppb maximum
12 h mean
ENVIRONMENTAL PROTECTION AGENCY (1986)
Duration
3 growing
seasons
3 growing
seasons
18 weeks
150 days
3 growing
seasons
245 days
3 growing
seasons
241 days
2 growing
seasons
Facility
OTCin
pots
OTCin
pots
OTCin
pots
OTCin
pots
OTCin
pots
OTCin
pots
OTC
OTC
OTC
Variable11
growth
dry weight
dry weight
growth
dry weight
foliar
weight
foliage
abscission
shoot
growth
needle
retention,
fascicle
length
Effect6
No effect on 5 families.
4% reduction at 30 to
38 ppb, 8% reduction at
51 to 65 ppb.
20% reduction in needles at
108 ppb.
10% reduction at 46 ppb.
0 to 13% reduction after
3 years at about 45 to 50 ppb
12 h seasonal mean,
depending on family.
35% reduction at 166 ppb.
Initiated above 130 to
220 ppm-h in trees exposed
to ambient or above.
Shoot length reduced 30% at
137 ppb.
Needle retention decreased in
elevated ozone — fascicle
length reduced by ozone in
early flushes, increased in
later flushes.
Reference
Adams et al.
(1990)
Edwards et al.
(1992)
Tjoelker and
Luxmoore (1991)
Shafer et al.
(1987)
Shafer et al.
(1989)
Qiu et al. (1992)
Stow et al. (1992)
Mudano et al.
(1992)
Kress et al. (1992)
-------
TABLE 5-27 (cont'd). A SUMMARY OF STUDIES REPORTING THE EFFECTS OF
8 OZONE ON THE GROWTH OR PRODUCTIVITY OF EVERGREEN TREES
| PUBLISHED SINCE U.S. ENVIRONMENTAL PROTECTION AGENCY (1986)
H E/D/
jg Species PC/Ca Concentration15 Duration Facility0
w Loblolly Pine E 0 to 150 ppb 5 h/day 6-12 weeks GC
5 days/week
Loblolly Pine E 0 to 320 ppb 6 h/day 8 weeks GC
4 days/week
Loblolly Pine E 0 to 120 ppb 7 h/day 12 weeks GC
5 days/week
Loblolly Pine E 0 to 320 ppb 8 h/day 9 weeks GC
4 days/week
^ Loblolly Pine E 20 to 100 ppb 4 h/day 10 weeks GC
O 3 days/week
Slash Pine E 76 to 104 ppb 7 h mean 112 days GC
126 ppb 1 h maximum
122 and 155 ppm-h
§ Slash Pine E 200 to 1 ,000 ppm-h 28 mo OTC
§
H Slash Pine E 179 to 443 ppm-h 24 h 28 months OTC
6 SUMOO multiples of
O ambient
0 .
H E = evergreen, D = deciduous, C = crop, PC = perennial crop.
O Means are seasonal means unless specified. Maximums are 1 h seasonal maxima unless
5 unless otherwise specified, accumulation based on 24 h/day unless otherwise noted.
Variable Effect6 Reference
dry weight 8% reduction at 150 ppb. Meier et al. (1990)
height and 20 % reduction in height Horton et al. (1990)
diameter growth 36% reduction in
growth diameter growth in three
open-pollinated families.
dry weight Top dry weight increased up Spence et al. (1990)
to 60 % root dry weight
reduced 6%.
relative 36% reduction in height RGR Wiselogel et al.
growth rate 10% reduction in diameter (1991)
(RGR) RGR.
dry weight No effect. Lee et al. (1990a)
top and root 18% reduction in top dry Hogsett et al.
dry weight weight and 39% reduction in (1985)
root dry weight at 122 ppm-h.
litterfall Twice as much litterfall at Byres et al. (1992)
ozone above 220 ppnrh.
leaf area Reduced up to 33 % by Dean and Johnson
443 ppm-h. (1992)
otherwise specified. Cumulative exposures are SUMOO
>-j COTC = open-top chamber with plants in ground unless specified in pots; CC = closed chamber, outside; GC = controlled environment growth
W chamber or CSTR; GH = greenhouse; F = field; OF = open-field fumigation.
O The effect reported in the study that is a measure of growth, yield, or productivity.
_ "^Effect measured at specified ozone concentration, over the range specified under concentration, or predicted (if specified) to occur based on
3 relationships developed in the experiment.
3
-------
1 concentrations that are considerably greater than those expected in ambient air. There was
2 an indication that total non-structural carbohydrate content was reduced by O3, which might
3 be an indicator of cumulative stress (Woodbury et al., 1991). However, results of these
4 studies indicate red spruce is tolerant of O3, at least for exposures of a few years.
5 Growth of seedlings of loblolly pine (a much faster growing species than red spruce)
6 has been reduced by O3 under some conditions. In growth chamber experiments, height
7 growth was reduced after exposure to 0.10 ppm for 4 h per day, three days per week for
8 10 weeks, but only in combination with a "control" rain treatment. The effect was not
9 observed in trees that received significant inputs of potential nutrients in simulated rain.
10 Conversely, Tjoelker and Luxmoore (1991) reported a significant reduction hi the weight of
11 current year needles following an open-top chamber exposure to O3 at a 7 h seasonal mean
12 of 0.056 or 0.108 ppm only in a high nitrogen treatment.
13 Multiple-year open-top chamber exposures of loblolly pine have resulted in decreased
14 foliar weight, partly through accelerated abscission, and decreased root surface area in the
15 first year following exposure to a 2.5 tunes ambient O3 treatment (0.10 ppm 12 h seasonal
16 average, 318 ppnvh) (Qiu et al., 1992). In a two year study, Kress et al. (1992) found that
17 fascicle length and number of early season needle flushes decreased linearly with increasing
18 O3, but the reverse was true in flushes produced later in the season. This may only occur in
19 seedlings that produce more than two leaf flushes per year. Foliage retention decreased with
20 O3, and fewer fascicles were retained on trees exposed to ambient concentrations of
21 O3 (12 h seasonal mean of 0.045 averaged over two years). Shafer and Heagle (1989)
22 exposed seedlings of four families of loblolly pine to O3 over three growing seasons and,
23 based on their data, predicted growth suppressions of above ground plant parts at a 12 h
24 seasonal mean of 0.05 ppm of 0 to 19% (depending on the sensitivity of the family) after two
25 years, and of 13% in the most sensitive family after three years. Cumulative effects of
26 multiple-year exposures were not apparent from the above study, but no measures of root
27 growth, which has been reported to be affected in other species (Andersen et al., 1991;
28 Edwards et al., 1992; Temple et al., 1993) were reported. Edwards et al. (1992) also
29 conducted a three year exposure and found a 4% reduction in whole plant biomass after
30 exposure to a 7 h seasonal concentration of about 0.050 ppm. An 8% reduction was
December 1993 5-206 DRAFT-DO NOT QUOTE OR CITE
-------
1 associated with a 7 h concentration of about 0.10 ppm. Growth reductions occurred in both
2 above and below ground plant parts.
3 Many studies with loblolly pine have used multiple families with a range of reported
4 tolerance to O3 (Adams et al., 1988, 1990; Kress et al., 1992;, Qiu et al., 1992; Shafer and
5 Heagle, 1989; Wiselogel et al., 1991). These studies have demonstrated the range of
6 response, from tolerant to sensitive, in the species. Adams et al. (1990) suggest that
7 resistance to natural stresses, such as drought, may be linked to tolerance to O3, thereby
8 affecting the response of the species to multiple stresses.
9 The response of slash pine to O3 has also been characterized. Dean and Johnson (1992)
10 found leaf area to be reduced by O3 in all three growing seasons studied, with an
11 intensification of the effect each year at an O3 exposure of about 0.03 to 0.04 ppm
12 (12 h seasonal means) or 77 to 216 ppm-h. Leaf litterfall was also increased by O3 (Byres
13 et al., 1992). Volume increment of the trees was affected, with an increased sensitivity to
14 simulated acid rain in trees exposed to twice ambient. Hogsett et al. (1985) found reduced
15 height (22%), diameter (25%), top (18%), and root growth (39%) in slash pine exposed to a
16 7 h seasonal mean of 0.076 ppm, with a maximum concentration of 0.094 ppm. From these
17 studies, it is clear that slash pine is relatively sensitive to O3 on an annual basis.
18 Hogsett et al. (1989) report the results of exposing 5 western conifers to Oj at a
19 seasonal 7 h mean concentration of 0.067 or 0.071 ppm (SUM07 for 134 day was 49.5 and
20 63 ppm-h; SUMOO was 140 and 153 ppnvh). Ponderosa pine and western hemlock had
21 reduced needle, stem, and root dry weight after 134 days of exposure. Douglas-fir and
22 western redcedar were not different from the charcoal-filtered air control, but douglas-fir
23 showed consistent decreases in weight of plant components. Lodgepole pine was not affected
24 by either O3 treatment. Carry-over effects were observed in bud elongation the following
25 spring in lodgepole pine, ponderosa pine and hemlock. Andersen et al. (1991) also observed
26 reduced root dry weight in ponderosa pine after exposure to SUMOO of 122 or 169 ppnvh
27 during a 120-day growing season. In addition, they observed a reduction in the weight of
28 newly formed roots the following spring, possibly due to reduced levels of root starch.
29 In a three year field study, Temple et al. (1993) and Beyers et al. (1992) found that
30 ponderosa pine trees exposed to a 24 h seasonal mean of 0.087 ppm had a 48 and 70% loss
31 of 2 and 3 year old needles, respectively. Radial stem growth and coarse root growth were
December 1993 5-207 DRAFT-DO NOT QUOTE OR CITE
-------
1 also reduced, but not as severely as needle weight (due to abscission). After three seasons of
2 exposure, current year needles in elevated O3 treatments had a higher photosynthetic capacity
3 than those in filtered air. The compensation was apparently due to higher foliar nitrogen in
4 O3-exposed needles, a product of redistribution of nitrogen before abscission of needles.
5 Cumulative responses would suggest that eventually, reductions in growth of the trees would
6 occur at lower concentrations of O3.
7 A number of field studies have been conducted in North America in which an attempt
8 was made to relate air quality to growth or injury of forest trees. Two field studies have
9 correlated radial growth with visible injury in ponderosa and Jeffrey pine in California
10 (Peterson and Arbaugh, 1988; Peterson et al., 1987). An 11% reduction in radial growth
11 was measured in symptomatic Jeffrey pine compared to trees that did not show symptoms of
12 O3 injury, but no reduction could be demonstrated in ponderosa pine, however, the authors
13 point out that the trees they measured were not under competitive stress, which might alter
14 their response.
15 The response of evergreen trees varies widely, depending on species, and genotype
16 within species. It is clear, however, that major forest species, such as ponderosa, loblolly,
17 and slash pine are sensitive to O3 at, or slightly above the concentrations of O3 (0.04 to
18 0.05 ppm) that occur over wide areas of the United States. Furthermore, because of the long
19 life span of these trees, including those that have not been reported sensitive to O3, there is
20 ample opportunity for a long-term, cumulative effect on growth of the trees. Most of the
21 experiments are conducted over only 2% or less of the life expectancy of the tree; an
22 equivalent exposure in field crop plants would be two to three days. Consideration must also
23 be given to the fact that most of these trees grow as part of mixed forests, in competition
24 with many other species. Small changes in growth might be translated into large changes in
25 stand dynamics, with concomitant effects on the structure and function of the ecosystem.
26
27 5.6.5 Assessments Using Ethylene Diurea (EDU) as a Protectant
28 Vegetational response to air pollutants can be modified by agricultural chemicals
29 commonly used by growers to control diseases, insects and other pests on crops (see
30 Section 5.4.7; U.S. Environmental Protection Agency, 1986). A chemical protectant,
31 ethylene diurea (EDU; N-[2-(2-oxo-l-imidagolidinyl)ethyl]-N-phenylurea), has been used to
December 1993 5-208 DRAFT-DO NOT QUOTE OR CITE
-------
1 study the response of plants to O3 without attempting to control the concentration of the
2 pollutant during the exposure (U.S. Environmental Protection Agency, 1986) Table 5-28.
3 Disadvantages of the use of open-top chambers for assessing the effects of Oj on the
4 growth of plants include relatively high cost, the need for electrical power, and potential
5 effects of the chambers themselves on the growth of the plants. In many cases, no chamber
6 effects can be detected, and since most studies compare against a control, chamber effects
7 would have a minimal effect on interpretation of results. While the number of experiments
8 conducted with open-top chambers has led to a firm understanding of plant response to a
9 chamber environment, the possibility of interactions with treatment cannot be ruled out. The
10 use of EDU is attractive due to low cost and ease of application, however, it is essential to
11 establish the correct dosage for protection from O3 without direct effects of EDU on the
12 plant, and an estimate of the level of protection from O3 achieved (Kostka-Rick and
13 Manning, 1992a,b, 1993). EDU is known to be phytotoxic, so studies under controlled
14 O3 conditions to establish an effective level of protection without phytotoxicity are essential
15 before it can be used as an assessment tool.
16 Previous studies with EDU led to the conclusion, as did experiments with open-top
17 chambers, that ambient concentrations of O3 were sufficient to reduce crop yields (U.S.
18 Environmental Protection Agency, 1986). If hourly O3 concentrations exceeded 0.08 ppm
19 for 5 to 18 days during the growing season, yields of crops might be reduced 18 to 41 %
20 (U.S. Environmental Protection Agency, 1986).
21 Inspection of Table 5-28 shows that in many cases there were clear-cut reductions in
22 O3-induced injury, and increases in yield resulting from the application of EDU. However,
23 the conflicting results for field-grown soybean indicated that, at the rate of EDU application
24 used, no beneficial effects could be demonstrated. Similarly, experiments with corn and
25 cotton suggest that any possible effects of O3 may have been confounded by direct effects on
26 growth of EDU itself.
27 A few studies using EDU have been conducted since 1986. Kostka-Rick and Manning
28 (1992a,b, 1993) conducted studies to determine the direct effects of EDU on growth and to
29 develop an understanding of dose-response to EDU itself. Their studies using EDU and
30 radish in the presence or absence of a controlled O3 fumigation in a greenhouse and found
31 that the chemical did suppress O3-induced reductions in below-ground plant organs. It also
December 1993 5-209 DRAFT-DO NOT QUOTE OR CITE
-------
u
e>o
a
&
I
1"
1
I
O
2
X
8
i
December 1993
5-210 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-28 (cont'd). EFFECTS OF ETHYLENE DIUREA (EDU) ON OZONE RESPONSES
u
I
Crop/Species
EDU Application
O3 Exposure
Effects of EDU
Reference
Ni
Radish (cont'd)
Soybean
Tobacco
Beech
H- Black cherry
Other woody species:
O Red maple
Soil drench, 150 mg/1,
60 ml/pot
Soil drench, 500 ppm,
0.51/pot
Soil drench, 500 ppm,
4l/6m row
Spray to runoff, 1 kg/ha
7 applications
Stem injection 1 g/1;
0.25ml
Spray to runoff, 1,000 ppm
7 applications/year
Spray to runoff, 500 ppm or
soil drench, 500 or
2,000 ppm, 250 ml/pot
Greenhouse, 0.07 ppm/7 h,
5 day/week, with two
weekly peaks to 0.12 ppm
Greenhouse; 0.2 ppm,
6 h/day, 2 days
Field; 78 h >0.12 ppm
(0.2 ppm max.)
Field; >0.08 ppm on
2 days
OTC; ambient and ambient
+0.08 ppm, 8 h/day
Field; 75 h >0.08 ppm
(over 4 years)
Up to 0.95 ppm, 3 h
Kostka-Rick & Manning
(1992b)
Reduced O3 injury,
90-100%; less reduction hi
hypocotyl weight
Reduced O3 injury, 80-90% Brennan et al. (1987)
No effect on loss of
chlorophyll; no effect on
seed weight
Increased growth, 22%
No consistent effect
2-fold increase in growth
Reduced O3 injury
Smith et al. (1987) and
Brennan et al. (1990)
Bisessar and Palmer (1984)
Ainsworth and Ashmore
(1992)
Long and Davis (1991)
Cathey and Heggestad
(1982)
6
o
g
H
O
0
$
0
o
Paper birch
White ash
Honey locust
Golden-rain
London plane
Lilac
Basswood
Reduced O3 injury
Reduced O3 injury
Reduced O3 injury
Reduced O3 injury
Reduced O3 injury
Reduced O3 injury
Reduced O3 injury
-------
1 protected the plants from foliar injury. The EDU itself did not cause effects on growth at a
2 concentration of 150 mg 1" applied as a 60 ml drench to each plant, a dosage much lower
3 than has often been used (e.g., Long and Davis, 1991; Smith, et al., 1987 discussed below).
4 Kosta-Rick and Manning emphasize that it is essential to establish the appropriate dose for
5 the species under consideration. Armed with this background, they used EDU in a field
6 study and found an O3-induced decrease in the relative growth rates of sink organs of field-
7 grown radish plants above a threshold level of about 0.052 to 0.058 ppm (7 h daily mean),
8 an exposure that is near ambient O3 concentrations.
9 EDU has also been used to estimate the effect of O3 on field-grown soybean in New
10 Jersey (Smith, et al., 1987; Brennan, et al., 1990). In this case, the researchers did not
11 establish the appropriate dose level for O3 protection as was done by Kostka-Rick and
12 Manning. No differences in yield were found and the authors conclude that O3 does not
13 impact soybean yield of the tested cultivars in New Jersey. However, they did not
14 demonstrate that EDU was an effective protectant at the concentrations used and on the
15 cultivars grown.
16 In a similar study, potato yields were measured and related to foliar injury in EDU
17 treated and non-treated plots over a four year period (Clarke, et al., 1990). The cumulative
18 O3 dose ranged from 45 to 110 ppm-h, depending on the year, producing a range of foliar
19 injury from 1 to 75 %. The authors found that significant differences in yield between EDU-
20 treated and control plants occurred only when foliar injury on untreated plants was 75 % of
21 leaf area. No level of protection, other than from foliar injury, could be assessed.
22 In a three-year study of potted green ash, no significant effects on growth were
23 measured using EDU (two years) or by comparison of filtered and non-filtered air in open-
24 top chambers (one year) (Elliot, et al., 1987). Foliar injury was observed only late in the
25 season of the first year in the non-filtered chambers.
26 An effort by Ensing, et al. (1986) to assess the impact of Oj on yield of peanut in
27 Ontario found that year-to-year variation was greater than that they could account for either
28 by correlation of O3 concentration with yield of test plots or by EDU treatment. They
29 conclude that a correlative approach to assessing losses due to O3 will not work.
30 Finally, a 4 year study of black cherry using EDU as a protectant was conducted by
31 Long and Davis (1991). They found significant effects with a 47% reduction in above
December 1993 5-212 DRAFT-DO NOT QUOTE OR CITE
-------
1 ground biomass compared to EDU-treated trees. The authors do not believe the difference
2 was due to a stimulation in growth due to nitrogen in the EDU, but did not conduct studies
3 as recommended by Kostka-Rick and Manning to characterize the EDU system for black
4 cherry.
5 The consistency of the results on field-grown white bean in southern Ontario led to the
6 development of an economic loss model based on the improved yields obtained as a result of
7 applications of EDU (Adomait et al., 1987).
8 In summary, the EDU method for assessing the impact of 03 is promising, particularly
9 for remote areas, or as a validation tool for existing crop-loss models. The system must be
10 carefully characterized, however, as pointed out by many of its users.
11 It should be noted that, in spite of the promise shown by EDU as a field protectant over
12 many years, it has not been developed commercially, and until recently was unavailable for
13 further experimentation.
14
15 5.6.6 Summary
16 Several conclusions were drawn from the various approaches used to estimate cropyield
17 loss. In 1986, U.S. Environmental Protection Agency (1986) established that 7 h/day
18 growing season mean exposures to O3 concentrations above 0.05 were likely to cause
19 measurable yield loss in agricultural crops. At that tune, few conclusions could be drawn
20 about the response of deciduous or evergreen trees or shrubs because of the lack of
21 information about the response of such plants to season-long exposures to O3 concentrations
22 of 0.04 to 0.06 ppm and above. However, the 1978 and 1986 criteria documents (U.S.
23 Environmental Protection Agency, 1986) indicate that the limiting values for foliar injury to
24 trees and shrubs was 0.06 to 0.10 ppm for 4 h. Since 1986, considerable research has been
25 conducted and the sensitivity of many tree species has been established.
26 Based on research published since U.S. Environmental Protection Agency (1986), the
27 following conclusions can be drawn:
28 1. An analysis of 10 years of monitoring data from more than 80 to almost
29 200 nonurban sites in the U.S. established ambient 7 h growing season average
30 concentrations of O3 for 3 or 5 mo of 0.051 to 0.060 and 0.047 to 0.054 ppm,
31 respectively. SUM06 exposures ranges from 25.8 to 45.2 ppm-h for 3 mo, and
32 32.7 to 44.4 ppm-h for 5 mo (Tingey et al., 1991).
33
December 1993 5-213 DRAFT-DO NOT QUOTE OR CITE
-------
1 2. The results of open-top chamber studies that compare yields at ambient
2 O3 exposures with those in filtered-air, and retrospective analyses of crop data
3 summarized in this section, establish that the current ambient concentrations of O^ at
4 some sites are sufficient to reduce the yield of major crops in the United States.
5 The results of research since 1978 do not invalidate the conclusions of the U.S.
6 Environmental Protection Agency (1976, 1986) that visible injury due to
7 O3 exposures reduces the market value of certain crops and ornamentals (spinach,
8 petunia, geranium and poinsettia, for instance), and that such injury occurs at
9 O3 concentrations (0.04 to 0.10) that presently occur in the United States.
10
11 3. A growing season SUM06 exposure of 26.4 ppm-h, corresponding to a 7 h growing
12 season mean of 0.049 ppm and a 2HDM of 0.094 ppm will prevent a 10% loss in
13 50% of the 54 experimental cases analyzed by Tingey, et al., (1991) and Lee, et al.,
14 (1993a,b). A 12 h growing season mean of 0.045 should restrict yield losses to
15 10% in major crop species (Lesser, et al., 1990).
16
17 4. Concentrations of O3 and SUM06 exposures that occur at present in the U.S. are
18 sufficient to affect the growth of a number of trees species. Given the fact that
19 multiple year exposures may cause a cumulative effect of the growth of some trees
20 (Simini, et al., 1992; Temple, et al., 1992), it is likely that a number of species are
21 currently being impacted, even if threshold exposures are not being reached.
22
23 5. A study of 51 cases of seedling response to O3 (Hogsett, et al., 1993), including
24 11 species representing deciduous and evergreen growth habits, concluded that a
25 SUM06 exposure for 5 mo of 31.5 ppm-h would protect hardwoods from a 10%
26 yield (growth) loss in 50% of the cases studied. A SUM06 exposure of 42.6 ppm-h
27 would provide the same level of protection for evergreen seedlings. Research by
28 others support these conclusions. It should be noted that these conclusions do not
29 take into the account the possibility of effects on growth in subsequent years, an
30 important consideration in the case of long-lived species.
31
32 6. Studies of the response of trees to O3 have established that, in some cases (Populous
33 and black cherry, for instance) trees are as sensitive to O3 as annual plants, in spite
34 of the fact that they are longer lived and have lower rates of gas exchange, and
35 therefore a lower uptake of O3.
36
37 7. The use of the chemical protectant, EDU is of value to establish O3 related losses in
38 crop yield and tree growth, providing care is exercised in establishing the
39 appropriate dosage of the compound to protect the plants without affecting growth.
40 EDU cannot be used to predict the response of plants at concentrations greater than
41 those that exist in ambient air.
42
43
44
December 1993 5-214 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.7 EFFECTS OF OZONE ON NATURAL ECOSYSTEMS
2 5.7.1 Introduction
3 The previous section presents the response of different species of crops and cultivated
4 plants, as well as data from deciduous shrubs and trees (Table 5-25) and evergreen tree
5 species (Table 5-26) to a variety of O3 exposure times and concentrations. The exposure-
6 response studies cited in the tables in the previous section, whether conducted in chambers,
7 greenhouses, or open-tops suggest that all sensitive plants will respond to O3 concentrations
8 above 0.06 ppm within hours. In general, depending upon the length of exposure, the
9 number and height of peaks, and the sensitivity of the vegetation, data from the field
10 supports this contention.
11 The environment in both natural and agricultural ecosystems is seldom optimal for plant
12 growth. In fact, most natural environments are suboptimal with respect to one or more
13 environmental parameters. Plant attempting to maintain a balance of resources by integrate
14 their responses (Chapin, 1991). Most plants undergo some form of stress during the various
15 stages of their life cycle, however, the multiple stresses they encounter during their lifetimes
16 do not usually all occur at once. In addition, the sensitivity to stress varies with the age of
17 the plant (Osmond et al., 1987). Cultivated plants are fertilized and frequently watered so
18 that they will have a balance of resources and produce better crops. Plants growing in their
19 natural habitats must compensate not only for the lack of resources but also for the multiple
20 stresses they usually encounter. How plants respond to O3 exposures and may compensate
21 for stresses is pointed out in the section on Mode of Action (5.3). The importance of genetic
22 variability in plant response, plant competition, and also the multiple biological and physical
23 factors that may modify plant response and, in some cases cause stress, are discussed in
24 Factors That Modify Plant Response (5.4). The discussion regarding modifying factors is of
25 particular importance in understanding ecosystem response to stresses because plants growing
26 in their natural habitats are much more likely to encounter them.
27 In order to place the known plant responses to O3 exposure in the ecosystem context, a
28 brief presentation of ecosystems characteristics and their importance to human existence is
29 given in the next section. The responses of forest ecosystems to pollutant exposure have
30 received more study than unmanaged ecosystems of other biomes (grasslands, shrublands, or
31 deserts). The following discussion relies mainly on forest ecosystems for its examples.
December 1993 5-215 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.7.2 Ecosystem Characteristics
2 5.7.2.1 Expected Sequence of Events
3 Ecosystems are composed of populations of "self-supporting" and "self-maintaining"
4 living plants, animals and microorganisms (producers, consumers, and decomposers)
5 interacting with one another and with the nonliving chemical and physical environment within
6 which they exist (Odum, 1989; U.S. Environmental Protection Agency, 1993). Mature
7 ecosystems are seldom stable. They must continually respond and adapt to changing
8 environments (Koslowski, 1985). Structurally complex communities, they are held in an
9 oscillating steady state by the operation of a particular combination of biotic and abiotic
10 factors. Ecosystems usually have definable limits, the boundaries of which, and the
11 organisms that can live there, are determined by the environmental conditions of that
12 particular area or region. They may be large or small (e.g., fallen logs, forests, grasslands,
13 cultivated or uncultivated fields, ponds, lakes, rivers, estuaries, oceans, the earth) (Odum,
14 1971). Together, the environment, the organisms, and the physiological processes resulting
15 from their interactions form the life-support systems that are essential for the existence of
16 any species on earth, including man (Odum, 1993).
17 The importance of ecosystems to human existence is pointed out in the nitrogen oxides
18 criteria document (U.S. Environmental Protection Agency, 1993).
19 Human welfare is dependent on ecological systems and processes. Natural ecosystems
20 are traditionally spoken of in terms of their structure and functions. Ecosystem structure
21 includes the species (richness and abundance) and their mass and arrangement in an
22 ecosystem. This is termed an ecosystem's standing stock—nature's free "goods" (Westman,
23 1977; U.S. Environmental Protection Agency). Society reaps two kinds of benefits from the
24 structural aspects of an ecosystem: (1) products with market value such as fish, minerals,
25 forest products and Pharmaceuticals, and genetic resources of valuable species (e.g., plants
26 for crops and timber and animals for domestication); and (2) the use and appreciation of
27 ecosystems for recreation, aesthetic enjoyment, and study (Westman, 1977; U.S.
28 Environmental Protection Agency).
29 More difficult to comprehend, but no less vital, are the functional aspects of an
30 ecosystem. They are the dynamics of ecosystems and impart to society a variety of benefits,
31 nature's free "services". Ecosystem functions encompass the interactions of the ecosystem
December 1993 5-216 DRAFT-DO NOT QUOTE OR CITE
-------
1 components and their environment and maintain clean air, pure water, a green earth, and a
2 balance of creatures; the functions that enable humans to obtain the food, fiber, energy, and
3 other material needs for survival (Westman, 1977; U.S. Environmental Protection Agency).
4 Ecosystems have both structure and function. The several most visible levels of
5 biological organization are (1) the individual and its environment; (2) the population and its
6 environment; (3) the community and its environment; and (4) the ecosystem (Billings, 1978).
7 These populations of plants, animals, and microorganisms (producers, consumers and
8 decomposers) in an ecosystem, live together and interact as communities. Individual
9 organisms within a population vary in their ability to withstand the stress of environmental
10 changes. The range of variability within which they can exist and function, determines the
11 ability of the population to survive. Communities, due to the interaction of their populations,
12 respond to pollutant stresses differently from individuals (U.S. Environmental Protection
13 Agency, 1993).
14 Intense competition among plants for light, water, nutrients, and space, along with
15 recurrent natural climatic (temperature) and biological (herbivory, disease, pathogens)
16 stresses, can alter the species composition of communities by eliminating those individuals
17 sensitive to specific stresses, a common response in communities under stress (Woodell,
18 1970; Guderian, 1985). Those organisms able to cope with stresses survive and reproduce.
19 Competition among different species results in succession (community change over time) and
20 ultimately produces ecosystems composed of populations of plant species that have a capacity
21 to tolerate the stresses (Kozlowski, 1980). Pollutant stresses, such as those caused by
22 exposure to O3, are superimposed upon the naturally occurring competitional stresses
23 mentioned above (See also Section 5.4). Air pollutants are known to alter the diversity and
24 structure of plant communities (Guderian et al., 1985). The extent of change that may occur
25 in a community depends on the condition and type of community as well as the pollutant
26 exposure.
27 In unpolluted atmospheres, the number of species in a community usually increases
28 during succession. Productivity, biomass, community height, and structural complexity
29 increase. Severe stresses, on the other hand, divert energy from growth and reproduction to
30 maintenance and alter succession (Waring and Schlesinger, 1995). In addition, biomass
31 accumulation and production decrease and structural complexity, biodiversity, environmental
December 1993 5-217 DRAFT-DO NOT QUOTE OR CITE
-------
1 modification, and nutrient control are reduced (Borman, 1985). When catastrophic
2 disturbances or injury, whether from natural (e.g., fire, flood, or windstorm) or
3 anthropogenic stresses (e.g., O3), alter the species composition (biodiversity) of a forest
4 enough to disrupt food chains and modify rates of energy flow and nutrient cycling,
5 succession is returned to and earlier less complex stage. The effects of stresses upon
6 ecosystems, unless they are catastrophic disturbances are frequently difficult to determine
7 (Koslowski, 1985; Garner et al., 1989). In a mature forest, a mild disturbance becomes part
8 of the oscillating steady state of the forest community or ecosystem. Responses to
9 catastrophic disturbances, however, as a rule are readily observable and measurable (Garner,
10 1993).
11 The structural changes within ecosystems are the result of functional changes that
12 occurred within the individuals, populations and community (e.g., altering of nutrient cycling
13 Milleretal., 1982). Ecosystem functions include the processes and interactions among the
14 various components and their environments that involve the movement of nutrients and
15 energy through a community as organic matter. Vegetation, through the process of
16 photosynthesis, plays and integral role in energy and nutrient transfer through ecosystems.
17 During photosynthesis, plants utilize energy from sunlight to convert carbon dioxide (CO^)
18 from the atmosphere and water from the soil into carbohydrates (See Section 5.3). The
19 energy accumulated and stored by vegetation as carbohydrates is also available to other
20 organisms such as herbivores, carnivores, and decomposers. As energy and nutrients move
21 from organism to organism in food chains or webs, they become more complex as ecosystem
22 diversity increases (Odum, 1993). Energy flow through the food chains is unidirectional.
23 At each step some is dissipated until ultimately, the amount left is dissipates into the
24 atmosphere as heat and must be replaced. Nutrients and water can be recycled, fed back into
25 the system, and used over and over again (Odum, 1993). The plant processes of
26 photosynthesis, nutrient uptake, respiration, translocation, carbon allocation, and biosynthesis
27 are directly related to the ecosystem function of energy flow and nutrient cycling. Reduction
28 in structure and diversity in ecosystems shortens food chains, reduces the total nutrient
29 inventory, and returns the ecosystems to a simpler successional stage (Woodwell, 1970).
30 How changes in these processes influence an ecosystem is discussed in the following text.
December 1993 5-218 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.7.3 Ecosystem Response to Stress
2 5.7.3.1 Forest Ecosystems
3 Ecosystem response to stress begins with individuals (Figure 5-23). Individuals, not
4 ecosystems respond to O3 exposure. Ecosystem responses are hierarchical ranging from
5 those that are characteristic of individuals to those characteristic of the entire ecosystem (see
6 Section 5.7.2). Table 5-29 lists the components that might be adversely affected by air
7 pollutants in hierarchical level, the possible functional responses that might occur at each
8 level of organization and the associated structural changes that might be expected (Sigal and
9 Suter, 1987). Changes in ecosystems begin with the death of individual plants. Even though
10 ecosystems integrate individual responses and propagate them through trophic and
11 competitive relationships, there can be no ecosystem response without an individual having
12 first responded. How the structural and functional components interact beginning with the
13 death of sensitive individuals as observed in the San Bernardino Forest ecosystem are
14 described below. Individuals sensitive to O3 have been described from the forests of both
15 coasts of the United States (U.S. Environmental Protection Agency, 1986). The impacts on
16 the west coast, however, have been much more severe, possibly because the exposures have
17 been chronic, continuing day and night at higher levels and the overall O3 concentrations
18 have been higher. In the Appalachian Mountains and the east in general O3 exposures have
19 been episodic varying in length from days to weeks (see U.S. Environmental Protection
20 Agency, 1986). The varying responses of the two ecosystems to O3 stress will be discussed
21 in the text that follows.
22 The extent of injury an ecosystem will experience from O3 exposure is determined by
23 the responses of its individuals. Leaf injury, as has been stated previously, is usually the
24 first visible indication of O3 exposure. Structural effects develop when functional responses
25 are severe enough (see Table 5-29). Stresses, whose primary effects occur at the molecular
26 or cellular physiology level in the individual, must be propagated progressively up through
27 more integrative levels of organ physiology (e.g., leaf, branch, root) to whole plant
28 physiology, stand dynamics, and then to the landscape level to produce ecosystem effects
29 (Figure 5-24; Table 5-29). Particularly, this is true if the stress is of low-level because only
30 a small fraction of stresses at the molecular and cellular level become disturbances at the
31 tree, stand, or landscape level. Insect defoliation, for example, may severely reduce the
December 1993 5_219 DRAFT-DO NOT QUOTE OR CITE
-------
Ozone Effects
Atmospheric Processes
Other
Stresses
Canopy Processes
Carbohydrate
Allocation
Leaf Processes/
Ozone Uptake
Reproduction
Leaf Processes/
Mode of Action
Mineral Uptake
Water Uptake
Ecosystem Response
Nitrogen
Phosphoi
Sulfur
J [j
Mycorrhizae Formation]
Figure 5-23. Effects of ozone on plant function and growth. Reduced carbohydrate
production decreases allocation to for resources need for plant growth
processes. Solid black arrows indicate affects of ozone absorption; gray
stripped arrows indicate affects on plant growth.
1 growth of one or several branches while growth of the tree appears not to be affected
2 (Hinckley et al., 1992).
3 Two properties that are important in determining the effect a stress at one hierarchial
4 level of organization will have on a higher level are variability and compensation.
5 Variability in response to stress may mean that, because of genetic variation, not all trees are
6 equally susceptible. At the stand level, the slower growth of some trees may be compensated
7 for by the relatively faster growth of others that are experiencing reduced competition so that
8 the overall growth of the stand is not affected (Hinkley et al., 1992). These properties when
9 taken together will determine the extent and rate at which stress at one hierarchical level will
December 1993
5-220
DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-29. PROPERTIES OF ECOLOGICAL SYSTEMS SUSCEPTIBLE TO
OZONE AT FOUR LEVELS OF BIOLOGICAL ORGANIZATION
Level of Organization
Properties
Structural
Functional Properties
Organism
Population
Community
Ecosystem
Leaf area and distribution
Biomass and allometry
Age and size structure
Population density
Genetic composition
Spatial distribution
Dispersion (spatial pattern)
Species composition
(diversity)
Trophic levels and food
webs
Physical structure (leaf area
index)
Biomass
Element pools
Soil properties
Photosynthesis, respiration
Nutrient uptake and release
Carbon allocation
Natality (reproduction,
mortality)
Competition
Productivity
Redundancy and resilience
Succession (the integration
of all species processes such
as competition and
predation)
Ecosystem productivity
Nutrient cycling
Hydrologic cycling
Energy flow
Source: Adapted from Sigal and Suter (1987).
1 impact the next highest level. To understand the effects of a stress, one must utilize the
2 framework of hierarchical scales (Figure 5-24) developed by Hinckley et al. (1992) to
3 provide a means by which the effects on forest trees of the eruption of St. Helens could be
4 followed and understood. This figure is also applicable for use when considering O3 effects
5 and can be used to explain the difference between the response of the San Bernardino Forest
6 ecosystem and the forests in the eastern United States. As pointed out above, variability and
7 compensation determine the severity of the response of the individual.
8 Variability and compensation can also occur at the population level, all populations do
9 not respond equally (Taylor and Pitelka, 1992). Plant populations can respond in four
10 different ways: (1) no response, the individuals are resistant to the stress; (2) mortality of all
11 individuals and local extinction of the extremely sensitive population—the most severe
December 1993
5-221
DRAFT-DO NOT QUOTE OR CITE
-------
Injury Symptom
Key Changes in Processes
Needle necrosis
& abscission
Branch length,
bifurcation ratio &
ring width growth
altered
Reduction In
diameter & death
of tree
Decreases In
stand productivity.
Increases In mortality
and alterations In
regeneration patterns
Reduced carbon assimilation
because of reduced radiation
Reduced carbon available for foliage
replacement & branch growth/export
Synerglstic Interaction between
mistletoe and tephra deposition
Reduced carbon available for
height, crown and stem growth
Influence of crown class on Initial
Impact & subsequent recovery
Interaction between stand
composition and recovery
Evaluating Impacts within a Level of Organization
Leaf Level Carbon exchange 1
Carbon pools 2
Needle #/slze 3
Needle retention/abscission 4
Branch Level Carbon allocation 5
Branch growth 6
Branch morphology 7
Branch vigor 8
Branch retention 9
Tree Level Hgt & diameter growth 1 0
Crown shape & size 1 1
Tree vigor 12
Mortality 13
Stand Level Productivity 1 4
Mortality 15
Species composition 16
Evaluating Interactions between Different Levels of Organization
The diagonal arrow Indicates the Interaction between any two levels of organization.
The types of Interaction are due to the properties of variability, and compensation.
A - refers to the Interaction between the leaf and branch levels where, for example,
variability at the branch level determines leaf quantity and compensation at the leaf
level In photosynthesis may compensate for the reduction In foliage amount.
B- refers to the Interaction between the branch and the tree where variability In branches
determines Initial Interception, branch vigor and branch location In the crown, and
compensation may be related to Increased radiation reaching lower branches.
C- refers to the Interaction between the tree and the stand. Both genetic and
environmental variability, Inter- and Intraspecfflc compensations and tree historical
and competitive synerglsms Involved.
Figure 5-24. Effects of environmental stress on forest trees are presented on a
hierarchial scale for the leaf, branch, tree, and strand levels of
organization. The evaluation of impacts within a level of organization are
indicated by horizontal arrows. The evaluation of interactions between
different levels of organization are indicated by diagonal arrows.
Source: Hinckley et al. (1992).
December 1993
5-222 DRAFT-DO NOT QUOTE OR CITE
-------
1 response; (3) physiological accommodation, growth, and reproductive success of individuals
2 are unaffected because the stress is accommodated physiologically; and (4) differential
3 response, members of the population respond differentially with some individuals exhibiting
4 better growth and reproductive success due to genetically determined traits (Taylor and
5 Pitelka, 1992). Differential response results in the progressive elimination over several
6 generations of the sensitive individuals and a shift in the genetic structure of the population
7 toward greater resistance (microevolution). Physiological accommodation or microevolution,
8 with only the latter affecting biodiversity, are the most likely responses for exposure to
9 chronic stress (i.e., stresses that are of intermediate-to-low intensity and of prolonged
10 duration). The primary effect of O3 on the more susceptible members of the plant
11 community is that the plants can no longer compete effectively for essential nutrients, water,
12 light, and space and are eliminated. The extent of change that can occur in a community
13 depends on the condition and type of community, as well as the exposure (Garner, 1993).
14 Forest stands differ greatly in age, species composition, stability, and capacity to recover
15 from disturbance. For this reason, data dealing with the responses of one forest type may
16 not be applicable to another forest type (Kozlowski, 1980).
17
18 5.7.3.2 The San Bernardino Forest Ecosystem—Before 1986
19 The mixed conifer forest ecosystem in the San Bernardino Mountains of southern
20 California is one of the most thoroughly studied ecosystems in the United States. The
21 changes observed in the mixed conifer forest ecosystem exemplify the information presented
22 in the foregoing discussion. Chronic O3 exposures over a period of 50 or more years caused
23 major changes in the San Bernardino National Forest ecosystem. The primary effect was on
24 the more susceptible members of the forest community, individuals of ponderosa and Jeffrey
25 pine, in that they were no longer able to compete effectively for essential nutrients, water,
26 light and space. As a consequence of altered competitive conditions in the community, there
27 was a decline in the sensitive species, permitting the enhanced growth of more tolerant
28 species (Miller et al., 1982; U.S. Environmental Protection Agency, 1978, 1986). The
29 results of the studies of the San Bernardino Forest ecosystem were reported in both the 1978
30 and 1986 criteria documents (U.S. Environmental Protection Agency, 1978, 1986). The
31 information summarized below is from these two documents.
December 1993 5-223 DRAFT-DO NOT QUOTE OR CITE
-------
1 An inventory of the forest was begun in 1968 and conducted through 1972 to determine
2 the results of more than 30 years of exposure to O3. Based on that inventory and
3 accompanying studies, the conclusions reached at that time are presented in Table 5-30.
4 Data from the inventory indicated that for 5 mo per year from 1968 through 1972 trees were
5 exposed to O3 concentrations greater than 0.08 ppm for more than 1300 h with
6 concentrations rarely decreasing below 0.05 ppm at night near the crest of the mountain
7 slope, elevation approximately 5,500 ft (Miller, 1973). The importance of altitude in plant
8 response was discussed in the 1986 criteria document (U.S. Environmental Protection
9 Agency, 1986).
10
TABLE 5-30. SAN BERNARDINO FOREST-STATUS 1972
1. Ponderosa and Jeffrey pine were suffering the most injury. Mortality of one population of ponderosa
pine (n = 160) was 8% Between 1969 and 1971 (p = 0.01); in a second population (n = 40),
mortality was 10% between 1968 and 1972. White fir populations had suffered slight damage, with
scattered individual trees showing severe symptoms. Sugar pine, incense cedar, and black oak
exhibited only slight foliar injury from oxidant exposure.
2. A substantial shift occurred in ponderosa pines from the "slight injury" category in 1969 to the
"moderate injury" category in 1971, indicating that there was continuing oxidant stress and that the
selective death of ponderosa pines was occurring.
3. Suppression of photosynthesis in seedlings was observed (Miller et al., 1969). In ponderosa pine
saplings, needles shortened by exposure to oxidants returned to normal length when the seedlings were
moved to ozone-free air during 1968 to 1973 (Miller and Elderman, 1977).
4. Bark beetles were judged to be responsible for the death of weakened trees in the majority of cases.
Elimination of ponderosa pine from the mixed-conifer forest was postulated to occur in the future if the
rate of bark beetle attack were to continue unabated (Cobb and Stark, 1970).
5. Aerial portions of ozone-injured pine trees showed a decrease hi vigor that was associated with
deterioration of the feeder root system (Parameter et al., 1969).
6. Seed production was decreased in injured pines. Ordinarily, trees 25 to 50 inches dbh produce the
most cones, but they were also the most sensitive to oxidants (Luck, 1980).
7. Understory plant species sensitive to oxidant pollution may already have been removed by air pollution
stress at the time of these early studies (Miller and Elderman, 1977).
Source: U.S. Environmental Protection Agency (1986).
December 1993 5-224 DRAFT-DO NOT QUOTE OR CITE
-------
1 The survey cited above indicated the need for further information and resulted in the
2 development of an interdisciplinary research team to more accurately determine the effects of
3 the 30 years of exposure to O3 of the San Bernardino Forest ecosystem. The study was
4 designed to answer two questions: How do organisms and biological processes of the conifer
5 forest response to different levels of chronic oxidant exposure? and (2) How can these
6 responses be interpreted within and ecosystem context?
7 The research plan included study of the limited aspects of the following ecosystem
8 processes: (1) carbon (energy) flow (the movement of carbon dioxide into the plants, its
9 incorporation into carbohydrates; and then its partitioning among, consumers, decomposers,
10 litter, and the soil; and finally its return to the atmosphere); (2) the movement of water in the
11 soil-plant-atmosphere compartments; and (4) the shift in diversity patterns in time and space
12 as represented by changes in structure, and density in the composition of tree species in
13 communities.
14 The major abiotic components studied were water (precipitation), temperature, light,
15 mineral nutrients (soil substrate, and oxidant pollution. The biotic components studied
16 included producers (an assortment of tree species and lichens), consumers (wildlife, insects,
17 disease organisms), and decomposers, (populations of saprophytic fungi responsible for the
18 decay of leaf and woody litter (U.S. Environmental Protection Agency, 1978, 1986).
19 During the period of the study, 1973 to 1978, average 24 h Qj concentrations ranged
20 from a background of 0.03 to 0.04 ppm in the eastern part of the San Bernardino Mountains
21 to a maximum of 0.10 to 0.12 ppm in the western part during May through September.
22 In addition to total oxidant, PAN and NO2 concentrations were measured. Peroxyacetyl
23 nitrate injury symptoms could not distinguished from O3 symptoms on herb-layer plant
24 species, while NO2 remained at non-toxic concentrations (Miller et al., 1982; U.S.
25 Environmental Protection Agency, 1978, 1986).
26 The study indicated that the major changes in the ecosystem began with injury to
27 ponderosa and Jeffrey pine.
28 Ponderosa pine was the most sensitive of the trees to O3 with Jeffrey pine, white fir,
29 black oak, incense cedar, and sugar pine following in decreasing order of sensitivity. Foliar
30 injury on sensitive ponderosa and Jeffrey pine was observed when the 24-h-average
31 O3 concentrations were 0.05 to 0.06 ppm (Miller et al., 1982). Foliar injury, premature
December 1993 5-225 DRAFT-DO NOT QUOTE OR CITE
-------
1 senescence, and needle fall decreased the photosynthetic capacity of stressed pines and
2 reduced the production of carbohydrates needed for use in growth and reproduction by the
3 trees. Nutrient availability to the trees was also reduced by their retention of smaller
4 amounts of green foliage (Miller et al., 1982). Decreased carbohydrate resulted in a
5 decrease in radial growth and in height of stressed trees (McBride et al., 1975; Miller and
6 Elderman, 1977).
7 Tree reproduction also was influenced by a reduction in carbohydrate. Injured
8 ponderosa and Jeffrey pines older than 130 years produced significantly fewer cones per tree
9 than uninjured trees of the same age (Luck, 1980). Tree ring analysis indicated declines in
10 ring width indices for many trees. Stand thinning however reversed the trend (Miller et al.,
11 1982).
12 Summarized, the responses of individual conifers sensitive to O3 include (1) visible
13 foliar injury; (2) premature needle senescence; (3) reduced photosynthesis; (4) reduced
14 carbohydrate production and allocation; (5) reduced plant vigor; and (6) reduced growth or
15 reproduction or both (Miller et al., 1982). The majority of these same responses were
16 observed on the eastern white pine growing in the Appalachian Mountains of the eastern
17 United States (McLaughlin et al., 1982; Skelly et al., 1984; U.S. Environmental Protection
18 Agency, 1986).
19 Injury to or changes in the functioning of other living ecosystem components affected,
20 either directly or indirectly, the processes of carbon (energy) flow and mineral nutrient
21 cycling, water movement and changed vegetational community patterns (Miller et al., 1982).
22 Change in decomposition patterns, altered nutrient cycling. Early senescence and abscission
23 caused an accumulation of pine needles into a thick layer under the stands of 03 injured trees
24 and changed the successional patterns of the fungal microflora as well. Altering the
25 taxonomic diversity and population density of the microflora that normally develop on
26 needles while they are on the tree, influenced its relationship with the decomposer
27 community. Change in the type of fungi on needles weakened the decomposer community
28 and the rate of decomposition (Bruhn, 1980). Carbon and mineral nutrients accumulated in
29 the heavy litter and the thick needle layer under stands with the most severe needle injury
30 and defoliation and influenced nutrient availability.
December 1993 5-226 DRAFT-DO NOT QUOTE OR CITE
-------
1 A comparison of lichen species found on conifers during the years 1976 to 1979 with
2 collections from the early 1900's indicated 50% reduction in species in the more recent
3 period. Marked morphological deterioration of the common species Hypogymnia
4 enteromorpha was documented in areas of high oxidant concentrations (Sigal and Nash,
5 1983).
6 Biotic interactions associated with predators, pathogens, and symbionts were influenced
7 by changes in the energy available to the trees. The decrease in vigor and lack of ability to
8 recover from ozone injury associated with reduced carbohydrates made the ponderosa pines
9 more susceptible to attack by predators and pathogens (Stark and Cobb, 1969). Dahlsten and
10 Rowney (1980) have pointed out that oxidant-weakened pines can be killed by fewer western
11 pine beetles than are required to kill healthier trees. In stands with a high proportion of
12 O3-injured trees, a given population of western pine beetles could therefore kill more trees.
13 James et al. (1980a,b) observed that the root rot fungus, Heterobasidium annosum, increased
14 more rapidly because freshly cut stumps and roots of weakened trees were more vulnerable
15 to attack (U.S. Environmental Protection Agency, 1986).
16 From the presentation above, it can be seen that temporal dynamics at the level of the
17 individual organisms can upset the equilibrium and be disorganizing in ecosystems that are
18 dominated by nonmoble organisms (Shugart, 1987). In forests with large canopy trees, a
19 canopy tree dominates the space where it is growing, reduces the amount of light reaching
20 the forest floor and alters the microclimate, conditions that help to determine the plant
21 species that can survive beneath the canopy. Death of a canopy tree increases the resources
22 of light, nutrients, water, and energy available to other organisms. This change initiates a
23 struggle for dominance among understory trees and seedlings. In time a new canopy
24 becomes established (Shugart, 1987; Garner et al., 1989).
25 The mode of death of a tree is ecologically important as it determines the regeneration
26 success of trees that form the next forest generation. Trees may dies catastrophically as
27 when high winds or ice storms break off the crown or branches or when they are blown
28 over, exposing roots; or they may dies slowly and tend to waste away as in the case of those
29 injured by pathogens or insects (Shugart, 1987). Tree death, as influenced by O3, is usually
30 gradual rather than catastrophic unless the tree is extremely sensitive or the pollutant
31 concentration is extremely high. Growth responses require time. Therefore, because growth
December 1993 5_227 DRAFT-DO NOT QUOTE OR CITE
-------
1
2
3
4
5
6
responses to cumulative stresses require time and trees are continually being subjected to
many other stresses, the cause of death is frequently difficult to determine (Garner et al.,
1989). The continuum of vegetation changes associated with increasing pollutant stress as
seen in the San Bernardino mixed forest ecosystem is presented in Table 5-31 (Garner et al.,
1989; U.S. Environmental Protection Agency, 1986).
TABLE 5-31. ECOSYSTEM RESPONSE TO POLLUTANT STRESS
Stage of Response of Vegetation
Response of Ecosystem
Continuum
0
I
n
ra
Anthropogenic pollutants insignificant.
Pollutant concentrations low; no measurable
physiological response.
Pollutant concentrations injurious to sensitive
species;
(1) Reduced photosynthesis, altered carbon
allocation, reduced growth and vigor;
(2) Reduced reproduction;
(3) Predisposition to entomological or
microbiological stress.
Severe pollution stress. Large plants of
sensitive species die. Forest layers are peeled
off; first trees, tall shrubs and, under most
severe conditions, short shrubs and herbs.
Unaffected; systems Pristine.
Ecosystem functions unaffected; pollutants
transferred from atmosphere to organic or
available nutrient compartments.
Altered species composition; populations of
sensitive species decline; some individuals are
lost. Their effectiveness as functional
ecosystem members diminishes; they could be
lost from the system. Ecosystem reverts to
an earlier stage.
(1) Simplification, basic ecosystem structure
changes, becomes dominated by weedy
species not previously present.
(2) Reduced stability and productivity; loss
of capability for repairing itself. Runoff
increases, nutrient loss, and erosion
accelerates; a barren zone results.
Ecosystem collapses.
Source: Garner et al. (1989); adapted from Bormann (1985); Kozlowski (1985); Smith (1974).
December 1993
5-228
DRAFT-DO NOT QUOTE OR CITE
-------
1 5.7.3.3 The San Bernardino Forest Ecosystem—Since 1986
2 The source for pollutants transported to the mixed conifer forests of the San Bernardino
3 Mountains is the South Coast Air Basin of southern California. Between 1976 and 1991 the
4 weather adjusted O3 data for the May through October "smog season" indicates that the
5 number of Basin days exceeding 0.12 ppm, 1 h average have declined at an average annual
6 rate of 2.27 days per year (Davidson, 1993). The number of days with episodes greater than
7 0.2 ppm, 1 h average have declined at an average annual rate of 4.70 days per year over the
8 same time period. The total days per year with concentrations greater than 0.12 ppm was as
9 high as 159 in 1978 with the lowest number being 105 days in 1990. The 1974 to 1988
10 trends of the May through October hourly average and the average of monthly maximum
11 O3 concentrations for Lake Gregory, a forested area in the western section of the San
12 Bernardino Mountains, have also shown a decline (Miller et al., 1989). Similarly there has
13 been an improvement shown in the injury index used to describe chronic injury to crowns of
14 ponderosa and Jeffrey pines between 1974 and 1988 in 13 of 15 plots located on the gradient
15 of decreasing O3 exposure in the San Bernardino Mountains (Miller et al., 1989). The two
16 exceptions were plots located at the highest exposure end of the gradient.
17 Miller et al. (1991) reported that for the 1974 to 1988 period the basal area increase of
18 ponderosa pines was generally less than competing species at 12 of the 13 plots evaluated.
19 The total basal area for each species as percent of the total basal area for all species
20 (Figure 5-25) shows that ponderosa and Jeffrey pines lost basal area in relation to competing
21 species that are more tolerant to O3, namely, white fir, incense cedar, sugar pine and black
22 oak at plots with slight to severe crown injury to ponderosa or Jeffrey pine. In effect stand
23 development has been forced in reverse, that is, the development of the normal fire climax
24 mixture dominated by fire tolerant ponderosa and Jeffrey pines is altered. The accumulation
25 of more stems of O3 tolerant species in the understory presents a fuel ladder situation that
26 jeopardizes the remaining overstory trees in the event of a catastrophic fire. The 03 tolerant
27 species are inherently more susceptible to fire damage because of thinner bark and branches
28 close to the ground. The important question for the future is whether the declining
29 O3 exposure will eventually allow ponderosa and Jeffrey pine to resume dominance in basal
30 area.
December 1993 5-229 DRAFT-DO NOT QUOTE OR CITE
-------
100
80-
80
20
gSofTotoltt
Borrow 74 (a)
. Camp PaMka
i
I
Breezy Point
1
Sky Forest
i I, i
Schi
\
wider Creek
• II.
Dogwood
ft i Eh il
PPICSPBO PPWFICSPBOLO PPWFICSPBOOW PP JPWFSPBOQCCP PPWFICSPBO
100
80-
60
40-
20-
u.c.
Center
SrfTMIUBA
So(Totil74BA
(b)
CampAngalus
Green Valley Creek
Barton Flats
T 1 1 1 T n 1 1 "I "1—-"I "I 1 r
PP BO PP WF SP BO PP JP WF JC SP BO PP JP SP BO QC
100-
80
r.
20
S of Total 74
CampOsosola
Holcomb Valtey
Heart Bar
(c)
Bluff Lake
PP JP WF BO
JP WF BO
JP WF
JP WF
Figure 5-25. Total basal area for each species as percent of the total basal area for all
species in 1974 and 1988 on (a) plots with severe to moderate damage,
(b) plots with slight damage, and (c) plots with very slight damage or no
visible symptoms.
Source: Miller, McBride, and Schilling (1991).
December 1993
5-230 DRAFT-DO NOT QUOTE OR CITE
-------
1 5.7.3.4 The Sierra Nevada Mountains
2 Since 1991 there has been an annual survey of the amount of crown injury by 03 to the
3 same trees in approximately 33 sample plots located in the Sierra Nevada including Tahoe
4 National Forest, Eldorado National Forest, Stanislaus National Forest, Yosemite National
5 Park, Sierra National Forest, Sequoi-Kings Canyon National Park, and Sequoia National
6 Forest plus the San Bernardino National Forest. In general, the results of this study
7 document the regional nature of the O3 pollution problem originating primarily from the San
8 Jaoquin Valley Air Basin (as well as the San Francisco Bay Air Basin further to the west).
9 Oxidant air pollution is transported southward in the San Jaoquin Valley Air Basin until it
10 reaches the southern boundary of the air basin—the Tehachipi Mountains. Because of this
11 barrier polluted air masses circulate back northward. This circulation cell causes higher
12 O3 levels to be adverted to the southernmost administrative units, namely the Sequoia
13 National Forest and the Sequoia-Kings Canyon National Park. The early tree injury results
14 (unpublished) corroborate the north to south increase of chronic O3 symptoms described by
15 Peterson et al. (1991). The most important long-range object of the study is to provide a
16 data base that will allow hypothesis testing regarding exposure response of ponderosa and
17 Jeffrey pines as a function of O3 concentration statistics, summer meteorology, and stomatal
18 conductance measurements.
19 The region-wide survey (Peterson et al., 1991) of ponderosa pine provides a useful
20 backdrop for reporting a number of other studies or surveys in the Sierra Nevada that were
21 more narrowly focused. Another tree ring analysis and crown injury study concentrated on
22 Jeffrey pines in Sequoia-Kings Canyon National Park (Peterson et al., 1989). This study
23 suggested that decreases of radial growth of large, dominant Jeffrey pines growing on thin
24 soils with low moisture holding capacity and direct exposure to upslope transport of
25 O3 amounted to as much as 11 % less in recent years when compared with similar trees
26 without symptoms.
27 Both permanent plots and cruise surveys have been employed in Sequoia, Kings Canyon
28 and Yosemite National Parks to determine the spatial distribution and temporal changes of
29 injury to ponderosa and Jeffrey pine within the Parks (Duriscoe and Stolte, 1989).
30 In Sequoia—Kings Canyon, O3 injury to individual trees and the number of trees injured in
31 each plot increased between 1980 to 1982 and 1984 to 1985 evaluations was the most
December 1993 5_23i DRAFT-DO NOT QUOTE OR CITE
-------
1 common response among the 28 plots studied. Ozone injury tends to decrease with
2 increasing elevation of plots. The O3 concentrations associated with the highest levels of tree
3 injury in the Marble Fork drainage of the Kaweah River at approximately 1,800 m elevation
4 are hourly averages frequently peaking at 80 to 100 ppb but seldom exceeding 120 ppb.
5 During a cruise survey in 1986 (Duriscoe and Stolte, 1989) to identifiy the partial
6 distribution of injury, there were 3,120 ponderosa or Jeffrey pines evaluated for O3 injury in
7 Sequoia, Kings Canyon and Yosemite National Parks. Approximately one third of this
8 number were found to have some level of chlorotic mottle. At Sequoia, Kings Canyon
9 symptomatic trees comprised 39% of the sample (574 out of 1,470) and at Yosemite they
10 comprised 29% (479 out of 1,650). Ponderosa pines were generally more severely injured
11 than Jeffrey pines.
12 In Sequoia, Kings Canyon observations at field plots showed that giant sequoia
13 seedlings developed O3 injury symptoms at both ambient O3 concentrations and 1.5 X
14 ambient O3 (0.08 to 0.1 ppm hourly peaks) (in open top chambers) during the 8 to 10 weeks
15 following germination (Miller and Grulke, submitted). Field plot observations of seedling
16 health and mortality in natural giant sequoia groves over a 4 year period showed that
17 seedling numbers were reduced drastically from drought and other abiotic factors. Any
18 variable such as O3 that could stress seedlings sufficiently to reduce root growth immediately
19 after germination could increase vulnerability to late summer drought. Significant differences
20 in light compensation point, net assimilation at light saturation, and dark respiration were
21 found between seedlings in charcoal filtered air treatments and 1.5 x ambient O3 treatments
22 (0.08 to 0.1 ppm hourly peaks) (Grulke et al., 1989). One interpretation of these results is
23 that O3 could be a new selection pressure during the regeneration phase of giant sequoia,
24 possibly reducing genetic diversity.
25 The Lake Tahoe Basin is located at the northern end of the Sierra Nevada sampling
26 transect (near the Eldorado National Forest) (Peterson et al., 1991). Because it is an air
27 basin unto itself the air quality situation is distinct from other Sierra Nevada sites. In 1987,
28 a survey of 24 randomly selected plots in the basin included a total of 360 trees of which
29 105 (29.2%) had some level of foliar injury (Pedersen, 1989). Seventeen of these plots had
30 FPM injury scores (Pronos et al., 1978) that fell in the slight injury category. Of 190 trees
December 1993 5-232 DRAFT-DO NOT QUOTE OR CITE
-------
1 in 16 cruise plots that extended observations to the east outside the basin 21.6% had injury—
2 less than in the basin.
3
4 Site Variables Affecting Ozone Response in the California Ecosystems
5 Structural changes in forest stands are highly related to their position on the landscape,
6 sometimes referred to as site. Site variables can be defined at regional and local levels. For
7 example, the regional level is defined in California by the location of forested mountain
8 slopes and summits in relation to polluted urban air basins. In both the Sierra Nevada
9 mountains and the San Bernardino mountains in California the worst tree injury is found on
10 ridges that overlook the polluted air basins. The polluted air masses are transported up-slope
11 or up-canyon in terrain that is usually sunlit in the afternoon and early evening, thus the
12 thermal convection on warm slopes is a major means by which O3 and associated pollutants
13 are delivered to the first forested ridges. Both vertical mixing and horizontal diffusion into
14 cleaner air results in a distinct gradient of decreasing O3 concentration in more distant forest
15 stands. Two such gradients have been described in the San Bernardino mountains (Miller
16 et al., 1986). Across the longer axis of the west to east orientation of the mountain range
17 O3 concentrations range from the highest summer months 24 h averages of 90 to 140 ppb
18 nearest the polluted South Coast Air Basin to 40 to 50 ppb at a downwind distance of 35 to
19 40 km. In the more narrow south to north direction the same concentration gradient is seen
20 over a much shorter distance of 5 to 8 km because of a more rapid transition to the warm
21 desert influence which causes mixing and dilution (Miller et al., 1972). Accordingly,
22 O3 injury to sensitive vegetation ranges from severe to none over these distances.
23 In the Sierra Nevada mountains a gradient of decreasing injury is observed from west to
24 east and south to north (Peterson and Arbaugh, 1992). But the worst level of chronic injury
25 is generally much less than observed in the San Bernardino mountains.
26 With respect to localized site variables there is evidence from repeated surveys in
27 Sequoia and Kings Canyon National Parks that percent of trees injured and the severity of
28 foliar injury both increased with decreasing elevation in the 1,500 to 2,500 m zone on
29 generally west facing slopes adjacent to the polluted San Joaquin Valley Air Basin (Stolte
30 et al., 1992). In Sequoia-Kings Canyon National Parks radial growth reductions in Jeffrey
31 pine with foliar injury by O3 was documented only for large, dominant trees growing on
December 1993 5-233 DRAFT-DO NOT QUOTE OR CITE
-------
1 shallow soils (Peterson et al., 1987). Soil moisture availability is generally lower on such
2 sites. One hypothesis for explaining radial growth decline on these sites and not more
3 favorable sites with greater moisture holding capacity is that O3 defoliation in favorable
4 moisture years and water stress in dry years integrate sequentially to suppress growth.
5 In the San Bernardino mountains, radial growth of ponderosa and Jeffrey pines in plots
6 along the decreasing O3 gradient was not well correlated with level of chronic injury but was
7 better correlated with soil moisture holding capacity. Within a single plot with relatively
8 uniform moisture availability there was a good correlation between increased radial growth
9 and a decreasing level of chronic O3 visible injury to crowns. Chappelka et al. (1992) have
10 suggested that some of the variability in foliar injury response of hardwood species to Oj in
11 Shenandoah National Park and the Great Smoky Mountains National Park is due to elevation
12 and microsite conditions, including the proximity to streams.
13
14 5.7.3.5 The Appalachian Mountains—Before 1986
15 Oxidant induced injury on vegetation in the Applachian Mountains has been observed
16 for many years but has not produced the same ecosystem responses. Results of studies in the
17 eastern United States was reported in the 1986 criteria document and is summarized in the
18 following passages (U.S. Environmental Protection Agency, 1986). Needle blight of eastern
19 white pine was first reported in the early 1900s but it was not known until 1963 that it was
20 the result of acute and chronic O3 exposure (Berry and Ripperton, 1963). The U.S. Forest
21 Service in the 1950s studied the decline of eastern white pine in an area covering several
22 hundred square miles on the Cumberland Plateau in Tennessee and concluded that
23 atmospheric constituents were the cause (Berry and Hepting, 1964). Despite this and other
24 early reports of field observations by Berry (1962, 1964), no concerted effort was made to
25 determine the effects of O3 on the vegetation of the Appalachian Mountains until the 1970s
26 when Skelly and his coworkers began monitoring total oxidants and recorded associated
27 injury to eastern white pine in three rural Virginia sites between April 1975 and March 1976
28 (Hayes and Skelly, 1977). Reductions in overall growth of eastern white pine was reported
29 by Benoit (1982). The mean ages of trees in the study plots classified as O3-tolerant,
30 intermediate and O3-sensitive were 53, 52 and 56 years, respectively. A comparison of
31 growth from 1974 to 1978 with that from 1955 to 1959 based on tree rings showed decreases
December 1993 5-234 DRAFT-DO NOT QUOTE OR CITE
-------
1 of 26, 37, and 51% for tolerant, intermediate and sensitive trees, respectively (Benoit, 1982).
2 The authors assumed the reduction in radial growth had occurred because there had been no
3 significant changes in seasonal precipitation between the 1955 to 1959 and the 1963 to 1978
4 periods. Ozone concentrations for 1 h at the study cites for 1979 to 1982 ranged from a low
5 of 0.063 ppm to a high of 0.129 ppm (U.S Environmental Protection Agency, 1986).
6 Growth reductions in trees growing on the were studied by Mann et al. (1980) and
7 McLaughlin et al. (1982). A steady growth decline in annual ring increment was observed
8 during the years 1962 through 1979. Reductions in average annual growth and 90% of bole
9 growth of 70% in sensitive trees when compared to the growth of tolerant and in
10 intermediate trees were observed. Tolerant trees consistently showed a higher growth rate of
11 from 5 to 15% than intermediate trees for the 1960 to 1968 interval, similar growth from
12 1969 through 1975, and a reduction in growth of 5 to 15% for the period 1976 through 1979
13 when compared to trees intermediate in sensitivity. The decline was attributed to chronic
14 O3 which frequently exceeded 1-h average concentrations greater than 0.08 ppm. Maximum
15 1-h concentrations ranged from 0.12 to 0.20 ppm (U.S Environmental Protection Agency,
16 1986).
17 McLaughlin et al. (1982) observed that the decline in vigor and reduction in growth in
18 trees and the production of carbohydrates (carbon flow) were associated with the following
19 sequence of events and conditions: (1) premature senescence of mature needles at the end of
20 the growing season; (2) reduced carbohydrate storage capacity in the fall and reduced
21 resupply capacity in the spring to support new needle growth; (3) increased reliance of new
22 needles of self-support during growth; (4) shorter new needles, resulting in lower gross
23 photosynthetic productivity; and (5) higher retention of current photosynthate (carbohydrate)
24 by foliage, resulting in reduced availability for transport to for external use including repair
25 of chronically stressed tissues of older needles (U.S. Environmental Protection Agency,
26 1986).
27
28 5.7.3.6 The Appalachian Mountains—Since 1986
29 In a survey of eastern white pine stands in the southern Appalachians fifty white pines
30 were examined for foliar symptoms (chlorotic mottle) believed to be O3-caused at 201 sites
31 distributed on a 24 x 24 km grid across the natural range of the species in South Carolina,
December 1993 5.235 DRAFT-DO NOT QUOTE OR CITE
-------
1 Tennessee, Virginia, North Carolina, Kentucky and Georgia (Anderson et al., 1988), The
2 survey was conducted from September through November 1985. The percentage of stands
3 with at least one symptomatic tree was highest in Kentucky (77%), followed by Tennessee
4 (31 %), and the lowest in Georgia (10%). The mean percentage of symptomatic trees per
5 plot for all six states was 27 %. The mean volume difference of 48 pairs of symptomatic and
6 non-symptomatic trees was 49% less for symptomatic trees. Elevation and percent slope
7 were not correlated with occurrence of symptomatic trees but most symptomatic trees were
8 found on southwest-facing slopes. Plantations had a higher percentage of symptomatic trees
9 than did natural stands. Ozone exposure concentrations were not reported but it may be
10 possible to make estimates of exposure at the nearest O3 monitoring sites.
11 Shenandoah National Park and Great Smoky Mountains National Park are contained
12 within the survey area investigated by Anderson et al. (1988). These parks have been the
13 subject of several surveys for O3 injury to native vegetation. For example, Winner et al.
14 (1989) surveyed 7 to 10 individuals of 5 native species at 24 sites in Shenandoah National
15 Park. These species included tulip poplar (Liriodendron tulipifera, wild grape (Vitis sp.),
16 black locust (Robinia pseudoacacia, virgin's bower (Clematis virginiana), and milkweed
17 (Asclepias sp.). Visible foliar injury due to O3 was most prevalent on milkweed species (up
18 to 70%) while the remaining species had injury approaching 20%. In each case the level of
19 foliar injury increased with elevation of the sites. The summer monthly 24-h mean
20 O3 concentrations at Blacksburg, Rocky Knob, Salt Pond and Big Meadows did not exceed
21 0.06 ppm and still foliar injury was observed.
22 Another recent survey in the Shenandoah National Park included black cherry, yellow
23 poplar, and white ash; and, in the Great Smoky Mountains National Park black cherry,
24 sassafras and yellow poplar (Chappelka et al., 1992). Black cherry exhibited symptoms in
25 both parks. In former the percentage of leaves injured ranged from 18 to 40 while in the
26 latter the range was 8 to 29% in 1991. Black cherry at Cove Mountain in the latter exhibited
27 the highest percentage of symptomatic trees (97%). This site also had the highest number of
28 hours exceeding 0.08 ppm. The majority of occurrences of concentrations exceeding
29 0.08 ppm occurred during evening hours.
30 In the previous O3 document (U.S. Environmental Protection Agency, 1986f), Duchelle
31 et al. (1982, 1983) reported that exposing native tree seedlings and herbaceous vegetation in
December 1993 5-236 DRAFT-DO NOT QUOTE OR CITE
-------
1 the Big Meadows area of the Shenandoah National Park in the Blue Ridge Mountains of
2 Virginia to ambient O3 reduced both the growth of the native trees other than eastern white
3 pine and the productivity of the native herbaceous vegetation found growing in forested
4 areas. Comparison of growth of seedlings in open plots or open-top chambers with charcoal-
5 filtered air suppressed the growth of wild-type seedlings of tulip poplar, green ash, sweet
6 gum, black locust (Robinia pseudoacacia L.), Eastern hemlock [(tsugs canadensis (L.)
7 Carr.), Table Mountain (Pinus pungens Lamb.), Virginia (P. virginiana Mill,) and pitch
8 (P. rigida Mill.) pines, usually without visible foliar injury symptoms. Open-top chambers
9 were operated continuously from May 9 until October 9 during the 1979 and from April 24
10 until September 15 in 1980 (U.S. Environmental Protection Agency, 1986). Common
11 milkweed (Asclepias syriaca L.) and common blackberry (Rubus allegheniensis Porter) were
12 two species of native vegetation to exhibit visible injury symptoms (Duchelle and Skelly,
13 1981).
14 It was mentioned previously that though there has been evidence of widespread injury
15 to native tree and other vegetation from exposure to O3 the amount of injury has not been
16 great enough for it to be transferred from the tree level to the stand level (Figure 5-23).
17 Undoubtably, there has been selection for and removal of the most sensitive tree species of
18 eastern white pine, for example. However, the numbers of sensitive individuals in a stand
19 have not been great enough to make a visible impact on the forest. Simulations suggest that
20 in forests with mixed species of uneven-aged stands suggest that long-term responses are
21 likely to be shifts in species composition rather than widespread degradation (Taylor and
22 Norby, 1985; U.S. Environmental Protection Agency, 1986).
23
24 5.7.3.7 Foliage and Soil-Mediated Effects—Combined Stress
25 Previously, it has been mentioned that the environment is seldom optimal in either
26 natural or agricultural ecosystems. It is not unusual, therefore, for plants growing in natural
27 habitats to encounter multiple stresses. Plant response to multiple stresses depend on
28 resource, particularly, carbohydrate and nitrogen, interactions at levels ranging from the cell
29 to the ecosystem (Chapin et al., 1987). Plant responses are either foliage-mediated, soil-
30 mediated or both. The discussions in the previous section has focused on the foliaged-
31 mediated response of plant species in an ecosystem to O3. This section discusses the possible
December 1993 5.237 DRAFT-DO NOT QUOTE OR CITE
-------
1 interactive effects of soil-mediated responses. It has usually been thought that nitrogen input
2 to natural ecosystem alleviated deficiencies in the soil and increased plant growth (U.S.
3 Environmental Protection Agency, 1993). Increased nitrogen additions can lead to nutrient
4 deficiencies of other elements (e.g., calcium and magnesium). Aber et al., (1989) state that
5 when nitrogen becomes readily available, other resources (e.g., phosphorus for plants and
6 carbon for microorganisms) become limiting. Changes in nitrogen supply can have an
7 impact on an ecosystem's nutrient balance and alter many plant and soil processes (U.S.
8 Environmental Protection-Agency, 1993).
9 The possible interactive effects of nitrogen on the forests of the San Bernardino
10 Mountains has come under consideration more recently. For some time there has been a
11 concern that O3 is not the only pollutant in the photochemical mixture that may be causing
12 lasting changes in the mixed conifer forest ecosystem. A multidisciplinary study to
13 investigate the possibility of the combined impacts on ecosystem processes from chronic
14 O3 injury and both wet and dry deposition of acidic nitrogen compounds has been under way
15 since 1991 at Barton Flats in the San Bernardino Mountains. The data base includes frequent
16 measurements of stomatal conductance in relation to weather and O3 exposure.
17 The nitrogen oxides criteria document (U.S. Environmental Protection Agency, 1993)
18 explored the possible effects of increased nitrogen on litter content and decomposition. That
19 discussion is presented here.
20 An increase in the nitrogen litter content and in litter decomposition rates and an
21 alteration in nitrogen cycling have been observed in the more highly polluted areas when
22 compared with moderate- and low-polluted areas of the San Bernardino Mountains of
23 Southern California (Fenn and Dunn, 1989). A pollutant concentration gradient exists with
24 24-h O3 concentrations at the high sites in the west averaging 0.1 ppm or higher, moderate
25 sites ranging from 0.06 to 0.08 ppm, and low sites in the east averaging 0.05 ppm or less
26 (Fenn, 1991). Nitrogen and sulfur compounds also occur in the pollutant mixture to which
27 the mountains downwind of the Los Angeles Basin are exposed (Bytnerowicz et al., 1987a,b;
28 Solomon et al., 1992). A nitrogen deposition gradient from west to east parallels the
29 decreasing O3 gradient. Deposition of nitrogen exceeds that of sulfur (Fenn and
30 Bytnerowicz, 1992). Annual average HNO3 concentrations in 1986 ranged from 1.2 ppb
December 1993 5-238 DRAFT-DO NOT QUOTE OR CITE
-------
1 near the Southern California coast to 2.7 ppb in the San Gabriel Mountains (Solomon et al.,
2 1992).
3 The effects of O3 exposure and injury to ponderosa (Pinus ponderosa Laws.) and
4 Jeffrey pine (P. Jeffreyi Grev. & Balf.) on a mixed conifer forest in the San Bernardino
5 Mountains, east of Los Angeles, have been studied for many years (Miller, 1973; Miller,
6 1984; U.S. Environmental Protection Agency, 1986). The litter layers under trees severely
7 injured by O3 is deeper than that under trees less severely injured (Fenn and Dunn, 1989).
8 A comparison study of litter decomposition rates of L-layer litter indicates that litter from the
9 more polluted areas in the west decomposed at a significantly (p = 0.01) faster rate than
10 litter from moderate to low pollution levels (Fenn and Dunn, 1989; Fenn, 1991). Nitrogen
11 content of litter was greatest at the high pollution sites and was positively correlated with the
12 litter decomposition rate. The higher nitrogen and lower Ca content of the litter suggests
13 that litter hi the western plots originated from younger needles than at the less polluted sites,
14 possibly due to O3-induced needle abscission. Fungal diversity was also greater in the litter
15 from the western San Bernardino Mountains (Fenn and Dunn, 1989).
16 When the factors associated with enhanced litter decomposition were investigated, it
17 was found that nitrogen concentrations of soil, foliage, and litter of ponderosa and Jeffrey
18 pine were greater in the plots where pollution concentrations were high than in moderate- or
19 low-pollution sites. This was also true for sugar pine (Pinus lambertiana Dougl.) and for
20 incense cedar (Calocedrus decurrens [Torr.] Florin.), two O3-tolerant species. The rate of
21 litter decomposition for all three pine species was greater at the high-pollution sites.
22 Therefore, the increased rate of litter decomposition in the high-pollution plots does not
23 appear to be related to O3 sensitivity or premature needle abscission, but rather due to higher
24 levels of nitrogen in the soils (Fenn, 1991).
25 Nitrogen is the mineral nutrient that most frequently limits growth in both agricultural
26 and natural systems (Chapin et al., 1987). The uptake of nitrogen and its allocation is of
27 overriding importance in plant metabolism and governs, to a large extent, the utilization of
28 phosphorus, potassium, and other nutrients, and plant growth. Plants usually obtain nitrogen
29 by absorbing ammonium (or ammonia) or nitrate (or nitrite) through their roots or through
30 fixation by symbiotic organisms. Nitrogen availability via the nitrogen cycle typically
31 controls net primary productivity. Normally, the acquisition of nitrogen is a major carbon
December 1993 5-239 DRAFT-DO NOT QUOTE OR CITE
-------
1 expense for plants. Plants expend a predominant fraction of the total energy available to
2 them in the form of carbohydrates in the acquisition of nitrogen through the processes of
3 (1) absorption, bringing nitrogen into the plant from the environment; (2) translocation,
4 moving inorganic nitrogen within the plant; and (3) assimilation, conversion of inorganic to
5 organic nitrogen (Chapin et al., 1987). Absorption of nitrogen from the soil requires
6 constant and extensive root growth to meet the needs of a rapidly growing plant because the
7 soil pools of mineral nitrogen, ammonium, or nitrate in the immediate vicinity of the roots
8 are usually so small that they are quickly depleted. A crude estimate suggests that the
9 fraction of carbon budget spent on absorption, translocation, and assimilation ranges from
10 25 to 45% for ammonium, 20 to 50% for nitrate, 40 to 45% nitrogen fixation, and 25 to
11 50% for formation of mycorrhizae (Chapin et al., 1987).
12 Nitrogen is the mineral that most frequently limits growth in both natural and
13 agricultural ecosystems (Chapin et al., 1987). The uptake of nitrogen and its allocation is of
14 overriding importance in plant metabolism and governs, to a large extent, the utilization of
15 phosphorus, potassium, and other nutrients, and plant growth. Nitrogen availability via the
16 nitrogen cycle typically controls net primary productivity. Normally, the aquisition of
17 nitrogen is a major carbohydrate expense for plants.
18 Nitrogen uptake influences photosynthesis because in the leaves of plants with
19 C3 photosynthesis (the pathway used by most of the world's plants), approximately 75% of
20 the total nitrogen is contained in the choloroplasts and is used during photosynthesis. The
21 nitrogen-photosynthesis relationship is, therefore, critical to the growth of trees (Chapin
22 et al., 1987). As a rule, plants allocate resources most efficiently when growth is equally
23 limited by all resources. When a specific resource such as nitrogen limits growth, plants
24 adjust by allocating carbohydrates to the organs that acquire the most strongly limiting
25 resources (Figure 5-26).
26 Among boreal and subalpine conifers, nitrogen added to the soil may not increase
27 growth. Depending on the plant species, nitrogen use efficiency above a critical level
28 decreases. All plants do not necessarily benefit from the added nitrogen in the leaves. More
29 nitrogen in the soil is not mirrored directly by increased nitrogen uptake except at low levels.
30 This is particularly true of conifers that are adapted to low-resource environments and tend to
31 have low potential growth rates. The photosynthetic capacity of conifer foliage is low and
December 1993 5-240 DRAFT-DO NOT QUOTE OR CITE
-------
Leaf
Biomass
Photosynthetic
Rate
c /
" K \
Root
Biomass
Nutrient \
Uptake )
Initial
Allocation
State
Environmental
Stress
Carbon
Reduce Carbon Supply
Low Light
SOj/Og
Leaf
Biomass
Reduce Nitrogen/Water Supply
Drought
Low Soil Fertility
Root
Biomass
New
Allocation
State -*^_«
"Carbon- "^ ^" —Carbon'
Figure 5-26. Impact of a reduced supply of carbon to the shoot, or water and nitrogen
t(> the roots, on subsequent allocation of carbon.
Source: Winner and Atkinson (1986).
1 not greatly enhanced by increased nitrogen content (Waring, 1985; Chapin, 1991). High leaf
2 nitrogen content is not always an advantage when other resources, among which are light and
3 water, are limited. At the present time, data dealing with the response of trees or other
4 vegetation to the combined stresses of O3 exposure above ground and nitrate deposition
5 through the soil are sparse. Tjoelker and Luxmoore (1991), however, have assessed the
6 effects of soil nitrogen availability and chronic O3 stress on carbon and nutrient economy in
7 1-year-old seedlings of loblolly pine (Pinus taeda L.) and yellow poplar (Liriodendron
8 tulipifera L). Elevated O3 concentrations altered biomass partitioning to needles of the
9 current year. Ozone concentrations of 0.108 ppm reduced the biomass of current-year
10 needles in loblolly pine seedlings grown at the highest (172 ^tg/g) nitrogen supply by 20%,
11 but not those grown with a low (59 jug/g) supply of nitrogen. The interaction between
12 O3 and nitrogen suggests that plants grown with a high nitrogen supply are more sensitive to
13 chronic O3 stress in terms of biomass reduction (Tjoelker and Luxmoore, 1991). Similar
14 results in the growth of domestic radish (Raphanus saliva L., cv. Cherry Bell) were obtained
15 by Pell et a). (1990). Brewer et al. (1961) and Harkov and Brennan (1980) observed
December 1993
5-241
DRAFT-DO NOT QUOTE OR CITE
-------
1 increased foliar injury when plants were grown with an adequate nitrogen supply (U.S.
2 Environmental Protection Agency).
3
4 5.7.3.8 Mycorrhizae-Plant Interactions
5 Mycorrhizae are quite literally, "fungus-roots", an association between plant roots and
6 fungi that is beneficial to both. The majority of plants develop this relationship and could
7 not reach optimum growth without it. Mycorrhizal fungi increase the uptake of mineral
8 nutrients for the host plants, protect roots against pathogens, produce plant growth hormones,
9 and move carbohydrates from one plant to another (Hacskaylo, 1972). The fungi in return
10 receive food in the form of simple sugars from the plant roots. Ozone has the capability of
11 disrupting the association between the mycorrhizal fungi and host plants by inhibiting
12 photosynthesis and reducing the amount of sugars available for transfer from the shoots to
13 the roots (see Figure 5-25). The benefits of this relationship was discussed in the previous
14 criteria document (U.S. Environmental Protection Agency, 1986).
15 Mycorrhizal fungi are an integral part of the below-ground ecosystem of terrestrial
16 communities. Some plants appear to be obligately mycorrhizal, while others are facultative
17 or non-mycorrhizal; however, mycorrhizae form on the vast majority of terrestrial plants and
18 contribute substantially to ecosystem function (Allen, 1991; Harley and Smith, 1983). The
19 symbiotic association is mutually beneficial and is characterized by a flow of nutrients from
20 fungus to host and flow of carbohydrates from host to fungus. Since oxidants have been
21 shown to alter photosynthesis and carbohydrate allocation within the plant, often decreasing
22 allocation to roots, they can disrupt the association by reducing carbohydrate availability to
23 the fungus. Any change in the vigor of the association will influence the nutrient harvesting
24 capability of the mycorrhizal hyphae, the physiology of the host plant, and subsequently
25 ecosystem function.
26 Several studies of the effects of 03 on tree species have included investigations of the
27 effects on ectomycorrhizal associations. The topic has been discussed in a series of articles
28 in a special issue of the journal Environmental Pollution (Vol. 73, No. 2), and selected
29 studies are summarized in Table 5-32. The understanding of oxidant effects on root
30 symbioses has not changed substantially since 1986 (U.S. Environmental Protection Agency,
31 1986), however, understanding of the importance of symbiotic organisms in ecosystem
December 1993 5-242 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-32. INTERACTIONS OF OZONE AND FOREST TREE ECTOMYCORRHIZAE INTERACTIONS
3
8*
u>
Host Plant
Mycorrhizae Exposure Condtions Effect of O3 on Mycorrhiza
Reference
Ul
to
Loblolly pine
Scots pine
White pine
Norway Spruce
Paper Birch
Red Oak
Pisolithus tinctorius
not stated
not stated
not stated
10 spp.
Pisolithus tinctorius
6 spp.
4 spp.
Pisolithus tinctorius
not stated
OTC, field reduced root infection
OTC, field, 3 years reduced root infection
CEC no effect
CSTR reduced root infection
open air, field, 3 years no significant effects
CEC reduced root infection
open air, field, 3 years no significant effects
CEC no consistent effects
CEC no effects
CEC and OTC significant increase
Adams and O'Neill (1991)
Edwards and Kelly (1992)
Mahoney et al. (1985)
Meier et al. (1990)
Shaw et al. (1992)
Stroo et al. (1992)
Shaw et al. (1992)
Blaschke and Weiss (1990)
Keane and Manning (1988)
Reich et al. (1985)
CEC: controlled environment chamber; CSTR: continuous stirred tank reactor; OTC: open-top filed chamber.
I
§
-------
1 function has improved. The basic hypothesis on mechanisms remains the same (i.e., effects
2 are mediated through host carbohydrate metabolism) since oxidants do not penetrate the soil
3 more than a few centimeters. Most of the research has been conducted on individual plant
4 species in controlled environments, and while the role of mycorrhizae in community stmcture
5 has been recognized, it has not specifically been addressed experimentally. Below is a
6 review of recent studies addressing oxidants effects on mycorrhizae.
7 Several studies have refined our understanding of oxidant stress effects on roots.
8 In Pseudotsuga menziesii [Mirb.] Franco root/soil respiration was significantly reduced
9 during the first 1 to 2 weeks after exposure to O3 or SO2, followed by a recovery period
10 which resulted in similar total respiratory release between treatments and controls (Gorissen
11 and Van Veen, 1988; Gorissen et al., 1991). Total allocation to roots did not appear to be
14
12 reduced, but 03 apparently reduced translocation to roots in that respiration of C was
13 suppressed. Edwards (1991) found that root and soil respiration were reduced in Pinus
14 taeda L. seedlings exposed to O3 levels ranging from 0.07 to 0.11 ppm (7 h mean) compared
15 to seedlings exposed to levels below ambient (0.02 to 0.04 ppm). Nouchi et al. (1991) found
16 that 03 at 0.1 ppm reduced root respiration by 16% in Oryza. sativa L. after one week of
17 exposure. However, exposure to 3 to 7 weeks of 0.1 ppm O3 resulted in elevated levels of
18 root respiration.
19 Several studies have examined the effects of O3 on carbohydrate allocation to roots and
20 subsequent shifts in biomass allocation (Cooley and Manning, 1987; Kostka-Rick and WJ.
21 Manning, 1992; Karnosky et al., 1992; Temmerman, Vandermeiren and Guns, 1992; Qui
22 et al., 1992; Sharpe et al., 1989; Gorissen and Van Veen, 1988; Gorissen et al., 1991;
23 Spence et al., 1990). Gorissen et al. (1991) examined the combined effects of 03 and
24 ammonium sulfate application to Psuedotsuga menziesii [Mirb.] Franco in association with
25 Khizopogon vinicolor A.H. Smith and Lactarius rufus (Scop. :Fr.)Fr. They found greater
26 needle retention of 14C labelled compounds in the new needles of O3 treated plants, and a
27 trend towards less 14C labelled substrates recovered in roots and root/soil fractions.
28 Short-term transport of UC labelled substrates were followed throughout P. taeda (Spence
29 et al., 1990). A 45% reduction in transport of photosynthates to roots occurred in O3 treated
30 plants compared to controls. Collectively, the studies have shown a general trend of
31 diversion of carbohydrate resources from roots and retention in the photosynthetically active
December 1993 5-244 DRAFT-DO NOT QUOTE OR CITE
-------
1 portions of plants. A reduction in allocation to roots may be associated with a change in the
2 availability of carbohydrate for maintenance of root symbioses.
3 The effects of O3 on mycorrhizal colonization have varied depending on the
4 experimental conditions and the species used. Stroo et al. (1988) studied the effects of 03 on
5 mycorrhizal infection in Pinus strobus seedlings grown for 4 mo in several soils. Results
6 varied by soil type and nitrogen availability; however, in several soils the number of
7 mycorrhizal short roots increased slightly at low O3 levels and decreased significantly at
8 higher O3 concentrations. Reich et al. (1986) found similar results in P. strobus L. and
9 Quercus rubra L., and concluded that O3 may stimulate mycorrhizal infection at low
10 O3 concentrations. Simmons and Kelly (1989) observed a trend of greater mycorrhizal short
11 roots in Pinus taeda seedlings exposed to subambient O3 treatment than those exposed to
12 ambient or twice ambient O3 levels; results were not statistically significant. In another
13 study with two families of P. taeda, Adams and O'Neill (1991) found that mycorrhizal
14 colonization tended to increase with O3 during the first 6 weeks of exposure, and decrease
15 with O3 after 12 weeks of exposure. Meier et al. (1990) found a decrease in ectomycorrhizal
16 root tips and percentage of feeder roots in P. taeda L. seedlings. Keane and Manning (1988)
17 found significant interactions among O3, soil type and pH, however the direct effects of
18 O3 were difficult to elucidate. Collectively, these results suggest that O3 does impact
19 colonization of roots by mycorrhizal fungi; however the results illustrate the variability in
20 response due to such factors as soil condition, duration of experiment, and timing of
21 measurements.
22 Altered root carbohydrate allocation resulting from O3 exposure may affect host-fungus
23 compatibility (Edwards and Kelly, 1992; Simmons and Kelly, 1989). Combined effects of
24 O3, rainfall acidity and soil Mg status on growth and ectomycorrhizal colonization of Pinus
25 taeda L has been studied (Simmons and Kelly, 1989). Although variation was high, there
26 was a trend towards altered species composition and reduced mycorrhizal infection in
27 O3 treated seedlings. Edwards and Kelly (1992) found high variability in morphotype
28 (morphologically different) frequency in response to O3 treatments in P. taeda L., and noted
29 changes in morphotype frequency over the 3 year study that suggested fungal succession had
30 occurred. Fungal succession and the effects of oxidant stress on normal successional patterns
December 1993 5.245 DRAFT-DO NOT QUOTE OR CITE
-------
1 are poorly understood. Shaw et al. (1992), using an open field exposure system, found no
2 differences in morphotype frequency or fruit-body succession in response to O3 treatments.
3 In summary, there is evidence that O3 affects mycorrhizal symbioses in a number of
4 plant species. The mechanism appears to be through altered carbohydrate allocation and
5 availability in the root. Efforts are currently underway to address successional patterns of
6 mycorrrhizal fungi under oxidant stress and should provide a more integrative understanding
7 of the effects at the ecosystem level.
8
9 5.7.3.9 Rhizosphere and Soil Processes
10 The importance of the below-ground ecosystem has largely been overlooked when
11 evaluating ecological responses to oxidant exposure. Although the soil system is part of the
12 larger terrestrial ecosystem, it is a system that operates independently and therefore is itself
13 an ecosystem (Richards, 1987). While above-ground components of the terrestrial ecosystem
14 are dominated by producers, the below-ground system is composed primarily of consumers.
15 Thus, the below-ground system is dependant on the above-ground system for inputs of
16 energy-containing substrates. Bacteria, fungi, protozoa, nematodes, microarthropods,
17 earthworms and enchytraeids all serve various functions in maintaining biological, physical
18 and chemical characteristics of soil, and all are dependent on plant residues for their
19 maintenance. While the uniqueness of the below-ground ecosystem needs to be recognized,
20 the interdependence between the above- and below-ground systems cannot be over
21 emphasized.
22 Even though the effects of O3 on rhizosphere and soil processes have not been studied,
23 potential impacts can be hypothesized based on known plant responses to O3. As noted
24 above, O3 stress reduces photosynthesis and growth, and roots often are more affected than
25 shoots (Winner et al., 1991; McCool and Menge, 1984; Blum and Tingey, 1977; Manning
26 et al., 1971; Tingey and Blum, 1973; Hogsett, 1985; Tingey et al., 1976; Spence et al.,
27 1990; McLaughlin et al., 1982). Oxidant stress has been shown to affect both leaf
28 senescence and root production in plants, thereby disnipting carbon availability for
29 maintenance of the below-ground system (Gorissen et al., 1991; Andersen and Rygiewicz,
30 1991).
December 1993 5-246 DRAFT-DO NOT QUOTE OR CITE
-------
1 The availability of current photosynthate for root growth appears to be reduced under
2 O3 stress, and maintenance of below-ground processes dependent on roots for their carbon
3 substrates may be affected. As noted in the previous section, a 45% reduction in transport of
4 photosynthates to roots occurred in O3 treated P. taeda (Spence et al., 1990). Ozone reduces
5 concentrations of root carbohydrates (Jensen, 1981; Tingey et al., 1976; Meier et al., 1990,
6 Andersen et al., 1991). Starch in roots was significantly reduced in P. ponderosa by the end
7 of one growing season of O3 exposure (Tingey et al., 1976). Reductions in coarse and fine
8 root starch concentrations persisted over the winter in O3-treated P. ponderosa and were
9 lower during the subsequent years shoot flush (Andersen et al., 1991). In this study, lower
10 starch concentrations in O3-treated seedlings were associated with suppressed growth of new
11 roots. The consequences of a reduction in carbon allocation below-ground includes reduced
12 substrate availability for soil flora and fauna, altered soil physical characteristics such as total
13 organic matter and aggregation, and altered soil chemical characteristics including cation
14 exchange capacity.
15 Premature leaf senescence has been observed in plants exposed to O3 stress (U.S.
16 Environmental Protection Agency, 1986). Premature senescence affects the below-ground
17 ecosystem by reducing canopy photosynthesis and carbon availability for transport to the
18 below-ground system, and by increasing leaf litter inputs to the forest floor (Miller, 1982;
19 Fenn and Dunn, 1989). The result is increased flux of nutrients, especially N, below-ground
20 due to oxidant exposure.
21 The increased flux of nitrogen due to premature needle senescence in oxidant exposed
22 plants may act to disrupt nutrient flow of the ecosystem. Allocation of carbon resources
23 throughout a plant is based on a priority scheme that is driven by carbon and nutrient
24 availability (Waring and Schlesinger, 1985). When soil nutrient levels are high, allocation to
25 the shoot is favored over the roots. By shifting carbon allocation to organs in this fashion,
26 plants can adjust to shifts in resource availability in their environment. Oxidant stress alters
27 typical allocation schemes and in the process may impair the plant's ability to cope with
28 drought or other stresses. In addition, reductions in allocation to roots can alter root system
29 size, architecture and spatial arrangement, which in turn can influence populations of soil
30 organisms.
December 1993 5-247 DRAFT-DO NOT QUOTE OR CITE
-------
1 Bacteria and fungi are particularly important in nutrient cycles and act to immobilize N,
2 C, P, and other nutrients in their biomass. The turnover of these nutrient pools are relatively
3 short as bacterial and fungal preditors act to release these nutrients. The majority of plant
4 available N during the growing season comes from these predatory interactions in the soil
5 (Kuikman et al., 1991; Ingham et al., 1985), emphasizing their importance in the
6 maintenance of terrestrial ecosystems. Currently there are no data available on the effect of
7 O3 on soil fauna.
8
9 5.7.4 Summary
10 Ecosystems are composed of populations of "self-supporting" and "self-maintaining"
11 living plants, animals and microorganisms (producers, consumers, and decomposers)
12 interacting with one another and with the nonliving chemical and physical environment within
13 which they exist (Odum, 1989; U.S. Environmental Protection Agency, 1993). Mature
14 ecosystems are seldom stable. They must continually respond and adapt to changing
15 environments (Koslowski, 1985). Structurally complex communities, they are held in an
16 oscillating steady state by the operation of a particular combination of biotic and abiotic
17 factors.
18 Ecosystem response to stress begins with individuals. Intense competition among plants
19 for light, water, nutrients, and space, along with recurrent natural climatic (temperature) and
20 biological (herbivory, disease, pathogens) stresses, can alter the species composition of
21 communities by eliminating those individuals sensitive to specific stresses, a common
22 response in communities under stress (Woodell, 1970; Guderian, 1985). Those organisms
23 able to cope with stresses survive and reproduce. The effects of stresses upon ecosystems,
24 unless they are catastrophic disturbances are frequently difficult to determine (Koslowski,
25 1985; Garner et al., 1989) In a mature forest, a mild disturbance becomes part of the
26 oscillating steady state of the forest community or ecosystem. Responses to catastrophic
27 disturbances, however, as a rule are readily observable and measurable (Garner, 1993).
28 Ecosystem responses are hierarchical ranging from those that are characteristic of
29 individuals to those characteristic of the entire ecosystem. Ecosystems integrate individual
30 responses and propagate them through trophic and competitive relationships. Two properties
31 that are important in determining the effect a stress at one hierarchial level of organization
December 1993 5-248 DRAFT-DO NOT QUOTE OR CITE
-------
1 will have on a higher level are variability and compensation. Variability in response to stress
2 may mean that, because of genetic variation, not all trees are equally susceptible. At the
3 stand level, the slower growth of some trees may be compensated for by the relatively faster
4 growth of others that are experiencing reduced competition so that the overall growth of the
5 stand is not affected (Hinkley et al., 1992). These properties when taken together will
6 determine the extent and rate at which stress at one hierarchical level will impact the next
7 highest level.
8 The mixed conifer forest ecosystem in the San Bernardino Mountains of southern
9 California is one of the most thoroughly studied ecosystems in the United States. The
10 changes observed in the mixed conifer forest ecosystem exemplify those expected in a
11 severly disturbed ecosystem. Chronic O3 exposures over a period of 50 or more years
12 caused major changes in the San Bernardino National Forest ecosystem. The primary effect
13 was on the more susceptible members of the forest community, individuals of ponderosa and
14 Jeffrey pine, in that they were no longer able to compete effectively for essential nutrients,
15 water, light and space. As a consequence of altered competitive conditions in the
16 community, there was a decline in the sensitive species, permitting the enhanced growth of
17 more tolerant species (Miller et al., 1982; U.S. Environmental Protection Agency, 1978,
18 1986). The results of the studies of the San Bernardino Forest ecosystem were reported in
19 both the 1978 and 1986 criteria documents (U.S Environmental Protection Agency, 1978,
20 1986). The more recent data from the San Bernardino Forest and from other ecosystems in
21 California indicate that there continue to be 03 concentrations injurious to forest vegetation.
22 The concentrations and durations, however, are not as high nor for as long periods as in
23 former times. For this reason, the vegetational injury has not been as great.
24 There is some indication from new data that O3 may not have been the only stress
25 encountered by the San Bernardino Forest ecosystem. Nitrate deposition gradients similar to
26 those measured for O3 suggests the possible soil-mediated exposures to nitrate could have
27 been anc continue to be combined with the foliage-mediated O3 exposures. Research in this
28 area is continuing.
29 Studies of O3 induced vegetational injury in the Appalachian Mountains and the
30 southeastern United States is ongoing. Preliminary results indicate that injury to sensitive
31 vegetation continues to occur.
December 1993 5-249 DRAFT-DO NOT QUOTE OR CITE
-------
1 Inhibition of photosynthesis by O3 exposure alters carbohydrate allocation from the
2 shoot to the roots. Reduced carbohydrate allocation to the roots has been shown to affect
3 mycorrhizae formation. Mycorrhizae are extremely important for mineral nutrient uptake by
4 trees and other vegetation reduction in their formation has been shown to have an detrimental
5 impact on plant growth.
6
7
8 5.8 EFFECTS OF OZONE ON AGRICULTURE, FORESTRY, AND
9 ECOSYSTEMS: ECONOMICS
10 5.8.1 Introduction
11 Evidence from the plant science literature cited hi the 1986 O3 Criteria Document (U.S.
12 Environmental Protection Agency, 1986) and in the present document is unambiguous with
13 respect to the adverse effects of tropospheric O3 on some types of vegetation. For example,
14 findings from U.S. EPA's multiyear NCLAN program provides rigorous corroboration of at
15 least a decade of previous research which showed that O3 at ambient levels caused physical
16 damage to important species, Specifically, NCLAN established that ambient O3 levels
17 resulted in statistically significant reductions in yields for these crops. Literature reviewed in
18 Section 5.6 of this document assesses the state of natural science findings regarding
19 O3 effects on crops, forests and other types of vegetation in more detail.
20 Information on the benefits and costs of alternative policy options or states of the world
21 (such as changes in air pollution) is of use to decision makers in a variety of settings. For
22 example, economic information provides one means by which to choose from alternative
23 policies or public investments. The role of cost-benefit analysis in federal rule making or
24 standard setting was enhanced in 1981 by President Reagan's Executive Order 12291
25 (February 19, 1981) which required that such calculations be performed on any rule or
26 regulation promulgated by the federal government. That executive order provided the
27 stimulus for a large increase in the use of economic analysis in evaluating federal actions,
28 including environmental policies. While the Clean Air Act and its amendments do not allow
29 the use of cost-benefit analysis in the standard setting process for primary (human health)
30 effects, economic information has been introduced into the discussion of secondary or
December 1993 5-250 DRAFT-DO NOT QUOTE OR CITE
-------
1 welfare effects. A number of economic studies addressing vegetation and other welfare
2 effects have been performed in the last decade.
3 Assessments of the economic consequences of O3 on vegetation reflect the state of
4 natural science information on each vegetation category. The natural science evidence
5 concerning effects of O3 on individual tree species or plant communities is less secure than
6 for agricultural crops (see Section 5.6). As a result, most economic assessments focus on
7 agricultural crops. The economics literature on effects of O3 and other air pollutants on
8 forest productivity is very sparse; the few assessments are confined to evaluations of assumed
9 or hypothetical changes in output (e.g., board feet of lumber). The economic effects of
10 O3 on plant communities or ecosystems have not been measured in any systematic fashion.
11 This section reviews economic assessments across these vegetation categories. The
12 discussion of economic valuation of ecosystem effects is limited to conceptual and
13 methodological issues in performing such assessments, given the absence of empirical
14 analyses in this category.
15
16 5.8.2 Agriculture
17 In view of the importance of U.S. agriculture to both domestic and world consumption
18 of food and fiber, reductions in crop yields could adversely affect human welfare. The
19 plausibility of this premise resulted in numerous attempts to assess, in monetary terms, the
20 losses from ambient O3 or the benefits of O3 control, to agriculture. Fourteen assessments
21 of the economic effects of O3 on agriculture were reviewed in the 1986 document (U.S.
22 Environmental Protection Agency, 1986). Since the preparation of the 1986 document, there
23 have been at least nine other studies published in the peer review literature which provide
24 estimates of the economic consequences of O3 on agriculture.
25 The 1986 document highlighted key issues in judging the validity of economic
26 assessments which are applicable to post-1986 studies (i.e., how well the biological,
27 aerometric, and economic inputs used in the assessment conform to specific criteria). First,
28 the evidence on crop response to O3 should reflect how crop yields will respond under actual
29 field conditions. Second, the air quality data used to frame current or hypothetical effects of
30 O3 on crops should represent actual exposures sustained by crops at individual sites or
31 production areas. Finally, the assessment methodology into which such data are entered
December 1993 5_25i DRAFT-DO NOT QUOTE OR CITE
-------
1 should (1) capture the economic behavior of producers and consumers as they adjust to
2 changes in crop yields and prices that may accompany changes in O3 air quality; (2) should
3 accurately reflect institutional considerations, such as regulatory programs and income
4 support policies (e.g., provisions of federal "Farm Bill" legislation), that may result in
5 market distortions; and (3) use measures of well-being that are consistent with principles of
6 welfare economics.
7
8 5.8.2.1 Review of Key Studies from the 1986 Document
9 Assessments of O3 damages to agricultural crops reported in the 1986 document
10 displayed a range of procedures for calculating economic losses, from simple monetary
11 calculation procedures to more complex assessment methodologies which conform to some or
12 all of the economic criteria above. As noted in the 1986 document, the simple procedures
13 calculate monetary effects by multiplying predicted changes in yield or production resulting
14 from exposure to O3 by an assumed constant crop price. By failing to recognize possible
15 crop price changes arising from yield changes and not accounting for potential producer
16 responses, such assessments are flawed, except for highly restricted situations such as
17 localized pollutant events. Conversely, some assessments provide estimates of the economic
18 consequences of O3 and other air pollutants that reflect producer-consumer decision-making
19 processes, associated market adjustments, and some measure of distributional consequences
20 between affected parties. The distinctions between studies based on naive or simple models
21 and those based on correct procedures is important at the regional or national level, since the
22 simple procedures may be biased, leading to potentially incorrect policy decisions.
23 Most of the economic assessments reviewed in the 1986 document (nine of the fourteen)
24 focused on O3 effects in specific regions, primarily California and the Corn Belt (Illinois,
25 Indiana, Iowa, Ohio, and Missouri). There have been a number of additional regional
26 assessments since the 1986 document; most are non-peer reviewed reports arising from
27 consulting or contract research. This regional emphasis in the earlier literature may be
28 attributed to the relative abundance of data on crop response and air quality for selected
29 regions, as well as (he importance of sonic agricultural regions, such as California, in the
30 national agricultural economy. Most of the recent state or regional assessments are
31 commissioned by state public utility commissioners or similar regulatory agencies and use
December 1993 5-252 DRAFT-DO NOT QUOTE OR CITE
-------
1 variants of the simple "price times yield" approach, where yields are calculated from
2 response functions arising from the NCLAN data. While perhaps of use to Public Utility
3 Commissioners concerned with effects from single power plants or other localized sources,
4 these studies generally contribute little to the assessment of pollution effects at the national
5 level. (Most local or regional studies abstract from physical and economic interdependencies
6 between regions, which limits their utility in evaluating secondary national ambient air
7 quality standards or NAAQS.)
8 National-level studies which account for economic linkages between groups and regions
9 can overcome some limitations of regional analyses. A proper accounting of these linkages,
10 however, requires additional data and more complex models, and frequently poses more
11 difficult analytical problems. Thus, detailed national assessments tend to be more costly to
12 perform. As a result, there are fewer assessments of pollution effects at the national level
13 than at the regional level.
14 Two national studies reported in the 1986 document were judged to be "adequate" in
15 terms of the three critical areas of data inputs. Together, they provided a reasonably
16 comprehensive estimate of the economic consequences of changes in ambient air Qj levels on
17 agriculture. Because of their central role in the 1986 document, these two studies are
18 reported in Table 5-33 and are reviewed briefly below.
19 In the first of these studies, Kopp et al. (cited as 1984 in the earlier document but
20 subsequently published as a journal article in 1985) measured the national economic effects
21 of changes in ambient O3 levels on the production of corn, soybeans, cotton, wheat, and
22 peanuts. In addition to accounting for price effects on producers and consumers, the
23 assessment methodology used is notable in that it placed emphasis on developing
24 producer-level responses to O3-induced yield changes (from NCLAN data available at the
25 tune) in 200 production regions. The results of the Kopp et al. study indicated that a
26 reduction in O3 from 1978 regional ambient levels to a seasonal 7-h average of
27 approximately 0.04 ppm would result in a $1.2 billion net benefit in 1978 dollars.
28 Conversely, an increase in O3 to an assumed ambient concentration of 0.08 ppm (seasonal
29 7-h average) across all regions produced a net loss of approximately $3.0 billion.
30 The second study, by Adams et al. (originally cited as 1984b, but subsequently
31 published as a journal article in 1986a), was a component of the NCLAN program. The
December 1993 5.253 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-33. RECENT STUDIES OF THE ECONOMIC EFFECTS OF OZONE AND
OTHER POLLUTANTS ON AGRICULTURE3
a
sr
i
i—*
\o
VO
u>
LA
hJ
Ul
•^
U
C
2
i
O
O
2!
/-)
e
^-<
S
i
Model Features
Pollutant and Price Output Input Quality
Region Concentration Changes Substitutions Substitutions Changes
Illinois Ozone. 10% increase No Yes Yes No
from 46.5 ppb
U.S. Ozone. 25% reduction Yes Yes Yes No
from 1980 level for each
slate
U.S. Ozone, universal Yes Yes Yes No
reduction from 53 ppb to
40ppbb
U.S. Ozone, universal Yes No No Yes
reduction from 53 ppb to
49 ppba
U S Acid deposition, 50% Yeh Yes Yes No
reduction in wet acidic
deposition
U.S Ozone. 10% reduction Yes Yes Yes No
from annual levels
(1986- 1990) for rural
areas. Includes
adjustments for 1985
Farm Bill.
U.S. Ozone, seasonal standard Yes Yes Yes No
of 50 ppb with 95%
compliance ; includes
adjustments for 1985
Farm Bill.
U.S. Increased UV-B radiation Yes Yes Yes No
and associated increase
of tropospheric pj (of
16%)
"All studies except Garcia et al. use NCLAN data to generate yield changes due to ozone.
Seven-hour growing season geometric mean. Given a log-normal distribution of air pollution events,
not to be exceeded more than once a year Heck et al. (1982).
Crops
Corn, soybeans
Corn, soybeans,
cotton, wheat,
sorghum, barley
Corn, soybean,
wheat, cotton,
peanuts
Soybeans
Soybeans
Corn, cotton,
soybeans, wheat
Corn, soybeans,
cotton, wheat,
sorghum, rice, hay,
barley
Soybeans (for
UV-B) and all
crops in Adams
et al. (1989) for
tropospheric 03
Results (Annual 1980 U.S. Dollars)
Consumer Producer Total Benefits
Benefits Benefits (Costs)
None $226 x 10° $226 x 106
$1.160 x 106 $550 x 106 $1,700 x 106
Not reported Not reported $1,300 x 106
$880 x 106 $-90 x 106 $790 x 106
$172 x 106 $-30 x 106 $142 x 106
NA NA $2,500 x 106
(sum of
discounted
values at 5 % ,
1986-1990)
$905 x 106 $769 x 106 $1,674 x 106
NA NA -830 x 106 (for
the increase in
tropospheric Qj
only)
Study
Garcia et al.
(1986)
Adams et al.
(1986)d
Kopp et al.
(1985)d
Shortle et al.
(1986)
Adams et al.
(1986)
Kopp and
Krupnick (1987)
Adams et al.
(1989)
Adams and Rowe
(1990)
a 7-h seasonal ozone level of 40 ppb is approximately equal to an hourly standard of 80 ppb,
cSevenand 12-h growing season geometric mean. Analysis includes both fixed roll-backs (e.g., 25%) and seasonal standards (with variable compliance rates).
Tteported in 1986 Criteria Document.
-------
1 results were derived from an economic model of the U.S. agricultural sector that includes
2 individual farm models for 63 production regions integrated with national supply and demand
3 relationships for a range of crop and livestock activities. Using NCLAN data, the analysis
4 examined yield changes for six major crops (corn, soybeans, wheat, cotton, grain, sorghum,
5 and barley) that together account for over 75 % of U.S. crop acreage. The estimated annual
6 benefits (in 1980 dollars) from O3 adjustments are substantial, but make up a relatively small
7 percentage of total agricultural output (about 4%). Specifically, in this analysis, a 25%
8 reduction in O3 from 1980 ambient levels resulted in benefits of $1.7 billion. A 25%
9 increase in O3 resulted in an annual loss (negative benefit) of $2.4 billion. When adjusted
10 for differences in years and crop coverages, these estimates are close to the Kopp et al.
11 (1986a) benefit estimates.
12 The Kopp et al. (1985) and Adams et al. (1986a) studies indicated that ambient levels
13 of O3 were imposing substantial economic costs on agriculture. However, both Kopp et al.
14 (1985) and Adams et al. (1986a) were judged to suffer from several sources of uncertainty.
15 These include the issue of exposure dynamics (7-h per day exposures from the NCLAN
16 experiments versus longer exposure periods, such as 12-h exposures), and the lack of
17 environmental interactions, particularly O3-moisture stress interactions, in many of the
18 response experiments. Also, the O3 data in both studies are based on a limited set of the
19 monitoring sites in the AIRS system, mainly sites in urban and suburban areas. While the
20 spatial interpolation process used for obtaining O3 concentration data (Kriging) resulted in a
21 fairly close correspondence between predicted and actual O3 levels at selected validation
22 points, validation for rural sites was limited (Lefohn et al., 1987). The economic models,
23 with their large number of variables, and parameters, and the underlying data used to derive
24 these values, were also noted as potential sources of uncertainty, including the effects on
25 economic estimates of market-distorting factors such as the federal farm programs. Concern
26 over farm programs stems from the evidence that reductions in O3 will increase yields and
27 hence total production of some crops. If the crop is covered (eligible for deficiency
28 payments) under the provisions of the farm program, then the total costs to the government
29 (of the farm program) may increase as a result of reduced O3 (McGartland, 1987). Thus,
30 the benefits of the O3 reduction may not be as great as estimated.
December 1993 5.255 DRAFT-DO NOT QUOTE OR CITE
-------
1 The 1986 criteria document concluded that these possible improvements in future
2 assessments were not likely to alter greatly the range of agricultural benefit estimates for
3 several reasons. First, the studies covered about 75 to 80% of U.S. agricultural crops (by
4 value). For inclusion of the other 20% to change the estimates significantly would require
5 that their sensitivities to O3 be much greater than for the crops included to date. Second,
6 model sensitivity analyses reported in past studies indicate that changes in plant exposure
7 response relationships must be substantial to translate into major changes in economic
8 estimates. For example, it was believed unlikely that use of different exposure measures or
9 inclusion of interaction effects would greatly alter the magnitude of the economic estimates.
10 Third, it was believed that there were likely to be countervailing effects that would mitigate
11 against large swings in the estimates (e.g., longer exposure periods may predict greater yield
12 losses), but O3-water stress tends to dampen or reduce the yield estimates. Finally, the
13 document noted that potential improvements in economic estimates are policy-relevant only to
14 the extent that they alter the relationship between total benefits and total costs of that policy.
15 The possible exception to this generally optimistic assessment of the robustness of the
16 estimates was inclusion of market-distorting factors (i.e., farm programs), an issue which is
17 addressed in some of the post-1986 assessments reviewed below.
18
19 5.8.2.2 A Review of Post-1986 Assessments
20 The 1986 document concluded that the O3 assessments by Kopp et al. (1985) and
21 Adams et al. (1986a) provided the most defensible evidence in the literature at that time of
22 the general magnitude of such effects. These two studies, in combination with the
23 underlying NCLAN data on yield effects, were judged to be the most comprehensive
24 information available on which to evaluate the economic impact of ambient air quality (O3)
25 on crops.
26 Seven national-level assessments performed since the last criteria document are reported
27 in Table 5-33. Of these, all use defensible economic approaches to quantify dollar effects,
28 where "defensible" is measured in terms of conforming to the criteria cited earlier.
29 An evaluation of these studies in terms of the adequacy of critical plant science, aerometric,
30 and economic data is presented in the table, along with estimates of benefits or damages
31 associated with changes in O3.
December 1993 5-256 DRAFT-DO NOT QUOTE OR CITE
-------
1 The concluding statements in the 1986 document are a benchwork against which to
2 judge these seven national level studies published since the last document. Most of the
3 contemporary studies build on either Kopp et al. (1985) or Adams et al. (1986a); indeed, the
4 motivation of some of the more recent studies is to test whether the problems noted above
5 (such as exclusion of farm programs) are sufficient to alter the original estimates in a
6 meaningful manner. A relevant question is whether these new studies provided any
7 "surprises" in terms of magnitude of economic effects. These studies are summarized in
8 Table 5-33.
9 In discussing these latest evaluations, there are several points which relate to their
10 comparability with Kopp et al. (1985) and Adams et al. (1986a). First, all studies use
11 NCLAN response data to generate yield effects (for inclusion in the respective economic
12 models). In most cases, data used in the post-1986 assessments reflect improvements of
13 earlier NCLAN data. Second, these studies may be characterized as second generation
14 assessments. They build on the first generation of studies reported in the 1986 document by
15 refining selected aspects of those earlier studies, including (1) interactions with other
16 stresses; (2) use of aerometric data and assumptions which, in some cases, more closely
17 follow the seasonal and regional characteristics of O3 exposure (Adams et al., 1989); and
18 (3) effects of O3 on quality of commodities (Shortle et al., 1986). Several of the studies use
19 updated versions of the economic models in Adams et al. (1986a) and Kopp et al. (1985).
20 In addition, some of the studies model the effects of government programs to judge the
21 potential consequences of such distortions on economic estimates (Kopp and Krupnik, 1987;
22 Adams et al., 1989). Third, there are differences in underlying aerometric assumptions;
23 some studies include both O3 and other environmental stresses (e.g., acid deposition, UV-B
24 radiation); others reflect O3 data for more recent time periods. Since ambient O3 levels vary
25 across years, the choice of year will influence the yield estimates and ultimately the
26 economic estimates.
27 Common themes or findings from these (and earlier) O3 and other air pollution studies
28 have been summarized in two recent synthesis papers (Adams and Crocker, 1989; Segerson,
29 1991). The results of the post-1986 assessments in Table 5-33 and the recent synthesis
30 papers corroborate the general findings of the 1986 document. Specifically, the agricultural
31 effects of tropospheric O3 at ambient levels impose economic costs to society (or conversely,
December 1993 5.257 DRAFT-DO NOT QUOTE OR CITE
-------
1 that reductions in ambient O3 result in societal benefits). The magnitude of the economic
2 costs reported in the more recent studies is similar to the estimates in Kopp et al. (1985) and
3 Adams et al. (1986a). Such a similarity is not surprising, given the points noted above
4 concerning use of similar data and economic models.
5 One important recent finding pertains to farm program. In each case, the inclusion of
6 farm programs in the economic models resulted in modest changes (reductions) in the
7 economic benefits of O3 control (due to increased farm program costs). As Segerson notes,
8 however, it is not clear that these increased costs should be charged against the potential
9 benefits of an O3 regulatory standard but rather as an additional cost associated with the
10 inefficiencies of the federal farm program. Even with the inclusion of farm programs and
11 other elements the general magnitude of further effects reported in the 1986 document are
12 only reduced by approximately 20%.
13 In addition to including farm programs, there are a couple of other notable additions to
14 the assessment literature. One study (Adams et al., 1989) attempts to analyze economic
15 benefits under a regulatory alternative involving a seasonal (crop growing season)
16 O3 exposure index measured as a 12 h mean, instead of hourly levels or percent changes
17 from ambient reported in earlier studies. Specifically, a seasonal average of 50 ppb
18 O3 (measured as a 12-h seasonal average) with a 95% compliance level, is reported in
19 Adams et al. (1989). The result (of a $1.7 billion benefit) is similar to the assumed 25%
20 reduction across all regions reported by Adams et al. in 1986a. At least one study has also
21 combined environmental stresses (e.g., O3, UV-B, radiation) in preforming economic
22 assessments. Adams and Rowe (1990), using the same model as Adams et al. (1986a,
23 1989), report that a 15% depletion of stratospheric O3 (which results in a 13% increase in
24 tropospheric O3) caused an economic loss of approximately $0.8 billion attributed to the
25 tropospheric O3 increase.
26
27 5.8.2.3 Limitations and Future Research Issues
28 The recent literature (post-1986) on economic effects of O3 on agriculture supports the
29 general conclusions drawn in the 1986 document. That is, ambient levels of O3 are imposing
30 economic costs on producers and consumers. However, there are at least three issues which
31 are not addressed in the extant literature on the topic. First, the existing assessments do not
December 1993 5-258 DRAFT-DO NOT QUOTE OR CITE
-------
1 consider the external costs of changes in agricultural production arising from changes in
2 O3 exposures (Sergerson, 1990). These costs are important if changes in O3 result in
3 changes in crop mixes or production practices, which in turn result in changes in soil
4 erosion, fertilizer and pesticide runoff or other agricultural externalities. For example, if
5 reductions in O3 increase the relative profitability of a crop which uses higher levels of
6 chemical inputs, then some increase in chemical effluent may result. Given that some
7 assessments suggest that such changes in crop mixes and production practices are likely to
8 accompany O3 changes, these costs/benefits need to be addressed.
9 A second issue not directly assessed in the current literature is the relationship between
10 climate change and tropospheric O3 effects. This relationship is important if global warming
11 is expected to increase tropospheric O3 levels. In addition, research indicates that global
12 climate change will lead to a relocation of crops (Adams et al., 1990). This relocation may
13 change the vulnerability of crop species to O3, given the spatial distribution of O3 across the
14 U.S. (i.e., increased crop production in areas of relatively low ambient O3, such as the
15 Pacific Northwest, implies lower O3 damage).
16 A third issue involves the institutional setting in which agricultural production occurs.
17 Several recent studies have assessed O3 effects in the presence of federal farm programs.
18 However, the U.S. and most industrialized economies are moving away from price supports,
19 production quotas and import restrictions, the traditional form of government intervention in
20 agriculture. At the same time that these market distortions are being removed, there is
21 increasing government regulation of agricultural production practices to reduce agricultural
22 externalities. Future assessments of O3 effects may need to pay less attention to farm
23 program effects and instead include other institutional features of U.S. Agriculture.
24
25 5.8.3 Forests (Tree Species)
26 The plant science literature on O3 and other air pollutant effects on tree species is
27 evolving rapidly as a result of recent research initiatives by EPA and other agencies. The
28 long-term nature of air pollution effects of perennial species creates challenges to plant
29 scientists in sorting out the specific effects of individual stresses from among the many
30 potential explanatory factors, such as O3 (Skelly, 1989). It also creates problems in terms of
December 1993 5-259 DRAFT-DO NOT QUOTE OR CITE
-------
1 measuring impacts of direct economic value, such as reductions in board-feet of lumber
2 produced per unit of time.
3 To date, most natural science literature on forest species reports O3 effects in terms of
4 foliar injury or similar measures (Taylor and Hanson, 1992; Davis and Skelly, 1992; Simini
5 et al., 1992; Freer-Smith and Taylor, 1992). This emphasis on foliar effects (rather than
6 marketable yield) is similar to the state of science for agricultural crops prior to 1975. Such
7 visible foliar effects information is of limited use in economic assessments . The exception
8 is in measuring the economic value of aesthetic changes in a forest stock (see Crocker,
9 1985a).
10 The lack of usable data concerning changes in marketed output, such as board-feet of
11 lumber (or even changes in growth rates), has limited the number of economic assessment of
12 O3 effects on forests. The few studies which attempt to measure economic losses arising
13 from O3 or other pollutants circumvent the lack of plant science data by assuming (arbitrary)
14 reductions in forest species growth or harvest rates (Callaway et al., 1985; Haynes and
15 Adams, 1992 ; Adams, 1986; Crocker and Forster, 1985).
16 These studies are summarized in Table 5-34. While the economic estimates reported in
17 Table 5-34 are comparable to these reported for agricultural crops (e.g., $1.5 billion for
18 eastern Canada, $1.7 billion for eastern U.S. forests), the lack of defensible natural science
19 data makes these studies suggestive at best, of possible economic consequences of forest (tree
20 species effects) of O3 or other environmental stresses. In addition, the economic
21 methodology used in the assessments varies, from simple price times quantity calculations
22 (e.g., Crocker, 1985b) to the use of large, econometric-based representations of the U.S.
23 timber market (Haynes and Adams, 1992). With appropriate data, the TAMM methodology
24 laid out by Haynes and Adams holds promise for assessing the economic consequences of
25 O3 when requisite natural science data become available.
26 In summary, the plant science literature shows that O3 adversely influences the
27 physiological performance of tree species; the limited economic literature also demonstrates
28 that changes in growth have economic consequences. However, the natural science and
29 economic literature on the topic is not yet mature enough to conclude unambiguously that
30 ambient O3 is imposing economic costs. The output from on-going natural science research
December 1993 5-260 DRAFT-DO NOT QUOTE OR CITE
-------
TABLE 5-34. STUDIES OF THE ECONOMIC EFFECTS OF OZONE AND
OTHER POLLUTANTS ON FORESTS
Pollutant/Coverage
All pollutants. Forest
products (hardwood
and softwood) in the
eastern U.S.
Acid deposition.
Forest products and
forest ecosystem
service flows for
eastern U.S.
Response and Air
Quality Data
Assumes three
arbitrary growth
reductions (10%,
15% and 20%) for
hardwood and
softwood tree
species.
Assumes a 5%
reduction in products
due to acid
deposition: assumes
a pristine background
pH of approximately
5.2
Economic Model
Spatial equilibrium
models of
softwood and
hardwood
stumpage and
forest products
industries in the
U.S.
Naive; assumed
changes in output
multiplied by
average value of
those goods or
services.
Annual
Damages or
Benefits of
Control
($ billion)
-270 X 106 to
563 X 106
damage in 1984
dollars for
assumed
reductions in
growth levels
-1,750 X 106
damage in 1978
dollars from
current levels of
acid deposition
Study
Callaway et al.
(1985)
Crocker (1985b)
Acid deposition.
Forest products and
forest ecosystem
services for eastern
Canada.
Air pollutants.
including acid
precipitation. Losses
estimated for eastern
U.S softwoods.
Assumes 5%
reduction in forest
productivity for all
eastern Canadian
forests receiving
> 10 kg/ha/year
sulphate deposition.
None; paper
demonstrates a
methodology for
assessing economic
effects of yield
(growth and
inventory) reductions
due to any course.
Assumes losses from
6 to 21 % for
softwoods.
Naive; assumed
changes in output
multiplied by
average value of
goods or services.
Econometric model
of U.S. timber
sector (TAMM).
-1,500 X 10° Crocker and
damage in 1981 Forster (1985)
Canadian dollars
from current
levels of acid
deposition
-1,500 x
-7,200 X
1986 dollars
106to
106in
Haynes and
Adams (1992)
December 1993
5-261
DRAFT-DO NOT QUOTE OR CITE
-------
1 on this topic will be important to our understanding of this potentially important class of
2 effects.
3
4 5.8.4 Valuing Ecosystem Service Flows
5 5.8.4.1 Background
6 Over the last thirty years economists have developed a variety of techniques for
7 assessing the value of nonmarket goods and services [recently surveyed by Braden and
8 Kolstad (1991) and Smith (1993)]. "Nonmarket" refers to those goods and services not
9 priced and traded in markets. While most applications are to natural resources and
10 environmental assets, the concepts extend to a range of goods not usually traded in markets.
11 Early applications focused primarily on commodities used directly by the consumer, such as
12 outdoor recreation. Within the last decade, attention has shifted to estimating nonuse
13 (or passive) values, such as what individuals are willing to pay to insure the existence of
14 species or unique natural settings. The values elicited with these techniques are being used
15 in an increasing array of settings; however, their use is not without controversy.
16 Valuing complex ecological functions and the associated range of ecosystem service
17 flows is relatively uncharted territory and raises a number of conceptual and practical issues.
18 Some difficulties in valuing ecosystem services lie in the inability of ecologists to
19 unambiguously define and measure ecosystem performance and endpoints (see Section 5.7 of
20 this chapter). Other problems arise from the inability of economic science to measure
21 adequately the consequences of long-term and complex phenomenon. A related problem is
22 the difference in disciplinary perspectives between ecologists and economists. As a result,
23 the current state-of-the-art for valuing ecosystem service flows is inadequate for benefit-cost
24 assessments used in environmental regulatory processes. As Costanza et al. (1991) state:
25 Because of inherent difficulties and uncertainties in determining values,
26 ecological economics acknowledges several different independent approaches.
27 There is no consensus on which approach is right or wrong — they all tell us
28 something — but there is agreement that better valuation of ecosystem services
29 is an important goal for ecological economics. [Bold emphasis added]
30
31 Improvement in valuation of ecosystem service flows will require increased interdisciplinary
32 cooperation and research between ecologists and economists, including the development of a
33 shared language. The objectives of this section are to provide a brief background on
December 1993 5-262 DRAFT-DO NOT QUOTE OR CITE
-------
1 nonmarket valuation techniques, outline some of the conceptual issues which require
2 resolution, and to offer an agenda for future research.
3
4 5.8.4.2 The Economic Perspective
5 Economic analysis is one input to the full social accounting of environmental planning
6 and management. For example, Tingey et al. (1990) discuss a variety of viewpoints for
7 defining "adverse effects" from ambient air quality, with economics being just one
8 perspective. More specifically, economic values are only one type of assigned values
9 (Brown, 1984). They reflect human preferences for a good or service, and are not inherent
10 in the good or service itself. Further, economic values are exchange values; they reflect the
11 terms of trade—dollars for services. Decision criteria which are based on economic values,
12 such as efficiency and benefit-cost analysis (BCA), reflect a particular utilitarian
13 philosophical perspective. Such a perspective is typically based on the notions of
14 "welfarism" (only individual assessments of value count), and "consequentialism"
15 (assessment of value is based on identified outcomes or consequences) (Sen, 1987). While
16 economists do not deny the existence or validity of alternative perspectives of value, they do
17 assert the importance of economic values in making some types of private and social choices.
18 For example, terms such as "ecosystem management" are likely to reflect a broad
19 spectrum of choices for policy makers. Each choice may imply an alternative path for
20 attaining a given "state" of biological diversity or ecological health. Each state will, in turn,
21 imply a different mix of endpoints (or outputs). Ecosystem management will alter the mix of
22 goods and services provided by a natural system, requiring tradeoffs between these
23 endpoints. Accurate assessment of these tradeoffs in selecting the optimal path and
24 subsequent endpoints poses numerous challenges. Economists believe that economic values
25 can aid in making such difficult social choices.
26 A critical first step in applying economy reasoning is the distinction between different
27 economic value components. The most familiar component involves market values. Market
28 values (prices) convey incentive information that guides the numerous independent decisions
29 that directly affect ecosystems (Perrings et al., 1992). However, market values typically will
30 not reflect the full social welfare change that results from a change in ecosystem service
December 1993 5-263 DRAFT-DO NOT QUOTE OR CITE
-------
1 flows. The distinction between market values and total social valuation requires the
2 introduction of nonmarket values and the notion of total economic value.
3 One taxonomy of economic values can be given as follows:
4
5 (1) Direct use values (DUV): determined by the contribution of an
6 environmental asset to current production or consumption (e.g., timber
7 harvest, recreation).
8
9 (2) Indirect use values (IUV): determined by the functional service flows
10 from the environment which support current and future production and
11 consumption (e.g., watershed protection, natural filtration of polluted
12 water, ecological functions).
13
14 (3) Option value (OV): refers to the value individuals place on the potential
15 future use of a resource (willingness to pay today for option to exercise
16 future use of an environmental asset).
17
18 (4) Bequest value (BV): refers to the present generation's preference for
19 bequest to future generations.
20
21 (5) Existence value (EV): contemplative values for the existence of a
22 resource arising independent of any current or future in situ use of the
23 resource).
24
25
26 A number of similar value taxonomies exist (e.g., Munasinghe, 1992; Pearce, 1993).
27 The critical distinction for decision-making is between goods and services whose economic
28 values are or are not fully captured in market prices. For example, timber products may
29 have a direct use value which is accurately reflected by market prices. Recreation may also
30 have a direct use value, but minimal or nonexistent fees do not accurately reflect this value.
31 Nonuse values, by definition, have no discernible link to market behavior. Specialized
32 techniques must be used to assess these values in a manner commensurate with more
33 conventional commodities (e.g., timber production sold in a market). Of the class of
34 nonmarket goods and services, the critical distinction is between use values and nonuse
35 values.
36 The measurement of total economic value (TEV) refers to systematic attempts to assess
37 the combined economic values of an environmental asset or resource system (Pearce, 1993;
38 Peterson and Sorg, 1987; Randall, 1991b). In the same way that physical resource functions
39 are interconnected, economic values for the various goods and services produced are
December 1993 5-264 DRAFT-DO NOT QUOTE OR CITE
-------
1 interconnected. A valid TEV measurement must account for this interconnectedness (e.g.,
2 substitution and complementarity issues, and non-additivity of component parts). While the
3 concept of TEV is generally accepted by economists, systematic attempts to measure TEV in
4 a regional policy or planning context are rare.
5
6 5.8.4.3 Nonmarket Valuation: Implications for Ecosystem Service Flows
7 Nonmarket valuation techniques consist of two basic types. Indirect approaches rely on
8 observed behavior to infer values. Direct approaches use a variety of survey-based
9 techniques to directly elicit preferences for nonmarket goods and services. Both sets of
10 techniques share a common foundation in welfare economics, where measures of willingness-
11 to-pay (WTP) and willingness-to-accept (WTA) compensation are taken as the basic data for
12 individual benefits and costs.
13
14 Indirect Approaches
15 Indirect approaches, sometimes referred to as revealed preference approaches, rely on
16 observed behavior to infer values. Examples include the travel cost method (TCM) where
17 the relationship between visits and travel expenditures is used to infer the value of a
18 recreational site, and hedonic pricing method (HPM) which attempt to decompose the value
19 of market goods, say recreational real estate adjacent to a national forest, to extract the
20 embedded values for environmental assets. Travel cost methods encompasses a variety of
21 models ranging from the simple single site travel cost model, to regional and generalized
22 models which incorporate quality indices and account for substitution across sites. Hedonic
23 pricing method encompasses both land price and wage models which account for variations
24 due to environmental attributes (e.g., air and water quality, noise, aesthetics, environmental
25 hazards). The indirect approaches can only measure use values. This limitation is brought
26 out in one fairly strong assumption (weak complimentarity) which requires that associated
27 with consumption of an environmental good or service is the purchase of some market good,
28 and when consumption of the market good is zero then demand for the environmental good
29 or service is almost zero (Adamowicz, 1991). Recent summaries of the indirect approaches
30 can be found in Braden and Kolstad (1991), Mendelsohn and Markstrom (1988), Peterson
31 et al. (1992) and Smith (1989, 1993).
December 1993 5_265 DRAFT-DO NOT QUOTE OR CITE
-------
1 A brief hypothetical example illustrates the use of an indirect approach to measuring
2 nonmarket value. The HPM can be applied to housing prices to estimate the value of
3 environmental attributes, such as "clean air" or proximity to wetlands, which vary across a
4 region. It is assumed that variations in housing prices can be linked to real or perceived
5 variations in these environmental attributes, (controlling for a variety of other statistical
6 determinants). In practice, the approach involves collection of cross sectional data on house
7 sales (or possibly assessed values) and information on a menu of potential determinants of
8 value (lot size, number of bedrooms, etc.). These factors would include one or more indices
9 of environmental attributes or services. Through multivariate statistical techniques, the
10 marginal value of either positive or negative environmental externalities can be inferred. For
11 example, it might be found that the average homeowner in a particular county would pay
12 $X to be Y meters closer to an open-water wetland, and would require a reduction in price
13 of $Z to be Y meters closer to a smoke-producing factory.
14 Applications of HPM are limited to use values and work best where there is some
15 identifiable spatial distribution of value. Continued improvements in available natural
16 science information (e.g., geographic information systems—GIS) will improve the efficacy
17 and precision of future HPM applications to environmental services. While direct use of
18 GIS-based information is rare in HPM models (Doss and Taff, 1993), this is a likely area of
19 future research expansion.
20
21 Direct Approaches
22 Direct approaches to nonmarket valuation are survey-based techniques to directly elicit
23 preferences. The hypothetical nature of these experiments requires that markets (private
24 goods or political) be "constructed" to convey a set of changes to be valued. While there are
25 a number of variants on these constructed markets, the most common is the contingent
26 valuation method (CVM).
27 Contingent valuation method can be viewed as a highly structured conversation (Smith,
28 1993) which provides respondents with background information concerning the available
29 choices and specific increments or decrements in one or more environmental goods. Values
30 are elicited directly in the form of statements of maximum WTP or minimum WTA
31 compensation for the hypothetical changes in environmental goods. For example,
December 1993 5-266 DRAFT-DO NOT QUOTE OR CITE
-------
1 recreational anglers may be asked about their WTP for specific increments in run size for a
2 fishery (Johnson and Adams, 1989). Typically, multivariate statistical techniques are used to
3 model a WTP function. Such models allow the analyst to control for variation in the
4 personal characteristics of the respondents, check for consistency of results with economic
5 theory, and possibly estimate an entire WTP response surface across varying levels of
6 environmental goods.
7 The contingent valuation method can be applied to both use and nonuse values. The
8 flexibility of constructing hypothetical markets accounts for much of the popularity of the
9 technique. There are numerous methodological issues associated with application of CVM
10 including how the hypothetical environmental change is to be specified, the elicitation format
11 for asking valuation questions, the appropriate welfare measure to be elicited (i.e., WTP or
12 WTA), and various types of response biases. Randall (1991a) argues that because of the
13 importance of nonuse values, CVM is likely to be the primary tool for measuring the
14 environmental benefits of biodiversity. Recent summaries of CVM can be found in Mitchell
15 and Carson (1989) and Carson (1991).
16
17 Nonuse Values
18 From a measurement perspective, "passive" or nonuse values (i.e., option, existence
19 and bequest) are the most problematic component of TEV. The contingent valuation method
20 is the only technique available for assessing these values. The topic of existence values for
21 environmental assets is one of the most controversial in environmental economics (Bishop
22 and Welsh, 1992; Edwards, 1992; Kopp, 1992; Rosenthal and Nelson, 1992). Evidence
23 shows that individuals will contribute to environmental organizations, and express positive
24 WTP to preserve environmental assets on CVM surveys, with no expectation of current
25 period or future use of the resource. However, evidence that existence values "exist" is
26 something less than arguing that they can be measured on a sufficiently comprehensive and
27 reliable basis for use in formal decision rules (Castle and Berrens, 1993; Rosenthal and
28 Nelson, 1992).
29 An example of some of the ambiguities in existence value estimation can be seem in a
30 CVM study by Stevens et al. (1991a) on bald eagles, wild turkeys, and Atlantic Salmon in
31 New England. While Stevens et al. (1991a) found substantial economic benefit from
December 1993 5_267 DRAFT-DO NOT QUOTE OR CITE
-------
1 protection and restoration programs, their results also indicate that in a setting of potential
2 irreversibiiity, existence values were difficult to quantify and sensitive to how the species
3 were aggregated. Further, a majority of respondents viewed species protection as important,
4 but were unwilling to pay anything. Follow-up questions indicated that many respondents
5 were uncertain of their values or protested the WTP question for ethical reasons.
6 Much of the controversy over nonuse values has been stimulated by debates
7 surrounding natural resource damage assessment and liability cases (e.g., the Exxon Valdez
8 oil spill in Alaska). This controversy among economists is highlighted by the recent blue-
9 ribbon panel, containing several Nobel Laureate economists, convened by the U.S.
10 Department of Commerce's National Oceanic and Atmospheric Administration (NOAA) in
11 1992. The panel was convened to provide guidance concerning the potential use of CVM in
12 measuring lost passive or nonuse values in promulgating regulations, pursuant to the Oil
13 Pollution Control Act of 1990. The potential for assessing nonuse values through application
14 of CVM was essentially reaffirmed by the NOAA panel, provided rigorous guidelines are
15 followed (Arrow et al., 1993).
16 There are a variety of recent examples of the application of CVM to measure existence
17 values, including Columbia River salmon (Olsen et al., 1991), and forest protection (Hagen
18 et al., 1992; Loomis et al., 1993; Rubin et al., 1991). Some economists remain skeptical
19 that existence values can be reliably measured for endangered species and irreversibiiity
20 problems (Castle and Berrens, 1993; Stevens et al., 1991b), and turn from the traditional
21 benefit-cost framework towards more ecologically conservative decision criteria such as the
22 safe minimum standard (SMS) of conservation (Ciriacy-Wantrup, 1952).
23
24 Use Values
25 Empirical estimates of nonmarket use values are less controversial, and are important
26 inputs to some types of research planning processes. The extant literature contains hundreds
27 of site-specific studies valuing recreational services and environmental quality. Recent
28 examples involving natural resources include Donnelly et al. (1990), Duffield et al. (1992),
29 Johnson and Adams (1989), Morey et al. (1991). Viewed in the aggregate, the numerous
30 valuation studies document the considerable economic worth of nonmarket goods and
31 services.
December 1993 5-268 DRAFT-DO NOT QUOTE OR CITE
-------
1 An emerging issue in the valuation literature is the need to develop acceptable
2 procedures for transferring values (Brookshire and Neil, 1992; Walsh et al., 1992). Benefit
3 transfer refers to the transfer of some existing valuation estimate (or function) from the study
4 site to a policy site. For example, incorporation of "typical" nonmarket values for
5 recreation, fish and wildlife occurs in the U.S. Forest Service planning process under the
6 Resource Planning Act (RPA) of 1974, as amended (Duffield, 1989). It should be noted,
7 however, that these "RPA values" differ greatly from state-of-the-art primary data studies
8 (Olsen, 1989), perhaps because the RPA values do not incorporate nonuse values (Duffield,
9 1989).
10
11 5.8.4.4 Challenges in Linking Valuation Techniques to Ecosystem Service Flows
12 The need for and interest in values of nonmarket goods and services has arisen
13 independently of concerns regarding ecosystem management and sustainability.
14 As environmental planning and management change to accommodate new issues, the need for
15 de novo valuation studies may increase (e.g., standard RPA values may be poor indicators of
16 the economic benefits and costs produced by forest quality changes under, say, alternative air
17 pollution regimes). The process of developing a tractable framework for ecosystem
18 management may require that valuation studies also co-evolve to aid critical management
19 decisions. For example, explicitly linking valuation techniques to physical resource functions
20 through bioeconomic models, remains an important research area (Adams et al., 1990a).
21 Linking valuation measures, from both market and nonmarket studies, to indices of biological
22 diversity is a fundamental challenge (e.g., Niese and Strong, 1992).
23 Ecologists have a traditional skepticism of attempts to assign monetary values to
24 ecosystem functioning, due both to inherent limitations in BCA and the inadequacy of
25 quantitative information about ecological and social factors (Westman, 1977; Higgs, 1987).
26 Attempts to monetize environmental benefits are also seen as having an inherent "quantitative
27 bias"; poorly understood ecological functions are neglected, while traditional commodities
28 (outdoor recreation) are paid full attention (Foy, 1990).
29 A further question is whether TEV really captures total value. Economists make no
30 claim that all values are being considered, only total economic value. But do traditional
31 measurement approaches really capture all economic value? As Pearce (1993) states:
December 1993 5_269 DRAFT-DO NOT QUOTE OR CITE
-------
1 There is in some sense a "glue" that holds everything together and that glue
2 has economic value. If this is true, and it is difficult to pinpoint what is at
3 issue here, then there is a total value to an ecosystem or ecological process
4 which exceeds the sum of the values of the individual functions. Total
5 economic value may not, after all, be total.
6
7 Can complex ecological functions be accurately expressed in monetary terms? It is
8 tempting to think that because of its inherent flexibility CVM may be up to the task (Russell
9 1993). The contingent valuation method has been applied to an impressive array of
10 nonmarket goods. While the traditional focus has been on outdoor recreation, applications
11 include endangered species and unique natural landscapes. However, precise valuation of
12 ecosystem services with CVM will require a precisely defined commodity. As researchers
13 move from valuing single environmental endpoints or services to addressing more complex
14 "bundles" of endpoints and services, it will become more difficult to define the commodity in
15 a CVM survey. This may prevent unambiguous estimation of such values.
16
17 Indirect Use Values
18 The common use/nonuse dichotomy may miss the fundamental interconnectedness and
19 transparency of complex ecological functions. The full range of ecosystem services is
20 comprised of multiple endpoints. This vector of endpoints is supported by complex
21 interactions. Uniqueness of the system lies in its functional structure, rather than discrete
22 environmental commodities—often the visual or spectacular—which satisfy human
23 preferences. Within the TEV taxonomy, indirect use values comes the closest to capturing
24 the system characteristics of complex ecological functions (Pearce, 1993).
25
26 Potential Misuse/Abuse of Economic Values
27 If understanding of functional structure of ecosystems is inadequate, then can any
28 attempt to express the valuation of ecosystem service flows be satisfactory? While services
29 may indeed be important, grossly imperfect attempts at monetary valuation may do more
30 harm than good (Norton, 1987; Rolston, 1988; Sagoff, 1988).
31 The first possibility is that precise quantification may introduce a degree of rigidity or
32 inflexibility that reduces the efficacy of future planning and management. There may be an
33 inherent need to seize onto any monetary value, no matter how poorly measured. Once
December 1993 5-270 DRAFT-DO NOT QUOTE OR CITE
-------
1 seized, the value can then take on a life of its own. A second possibility is that focusing on
2 satisfaction of preferences of the current generation and distributional status quo ignores
3 other collective social priorities. Examples include intergenerational equity and the notion of
4 preserving critical natural capital (biogeochemical cycles, genetic information, etc.) embodied
5 in the vast literature on sustainable development (Pezzey, 1992; Pearce, 1993).
6 If attempts at assigning monetary values are found to be unsatisfactory for the full
7 complex of ecosystem service flows, then alternative decision rules must be given closer
8 scrutiny. One such decision rule is the often-discussed safe minimum standard.
9
10 Safe Minimum Standard
11 The safe minimum standard (SMS) of conservation for ecosystem or other resources
12 which give rise to a steady stream of benefits or servcies ("flow resources") was originally
13 introduced by Ciriacy-Wantrup (1952). The SMS is a decision rule to protect critical flow
14 resources unless the costs of doing so are intolerably large (Randall, 1989, 1991). The SMS
15 rule is designed to provide flexibility and protect future options in the presence of true
16 uncertainty (where probability assessments cannot be meaningfully made), and the possibility
17 of irreversible ecological changes exist (Westman, 1977). Determination of "intolerable" is a
18 social choice which can be informed by both economists (by identifying the costs associated
19 with alternative degrees of protection or preservation), and ecologists (by identifying
20 ecological indicators and threshold effects which can be tied to preservation costs). The SMS
21 concept has received increasing attention (e.g., Bishop and Woodward, 1993; Randall, 1991)
22 and is closely linked to the notion of "critical natural capital" that occurs in some
23 formulations of sustainable development in agriculture and forestry.
24 Recent discussions of the SMS to threatened and endangered species applications
25 include Hyde (1989), Castle and Berrens (1993), and Stevens et al. (1991).
26
27 5.8.4.5 The Research Agenda
28 The previous sections highlight some deficiencies in the natural science and economics
29 literature which prevent valuation of ecosystem services. This section offers a set of
30 opportunities for future research efforts to addresses these deficiencies. No priority is
31 assigned and some opportunities are obviously overlapping.
December 1993 5-271 DRAFT-DO NOT QUOTE OR CITE
-------
1 (1985) suggests that the value of additional natural science information (in terms of reducing
2 confidence materials around economic estimates) declines rapidly. Thus, not all sources of
3 uncertainty need be removed before economic analyses can proceed.
4
5 Explore Relationship of Ecosystem Functions and Service Flows to Economic Values
6 An important research question is whether measures based on subjective preferences
7 (e.g., stated WTP from a CVM study, or observed WTP from an HPM study) have any
8 relationship to indices of ecosystem functions and service flows (Costanza et al., 1991). For
9 example, Hanley and Ruffell (1993) find that "forest characteristics are in general poor
10 predictors" of recreational value in a CVM study. However, they also conclude that the
11 chief problem lies in quantifying changes in natural science information. Utilization of GIS-
12 based tools offer new opportunities for connecting natural science information with a variety
13 of economic models.
14 A critical building block in identifying such connections is the development of a shared
15 language. For example, the connections between the concept of "endpoints" for ecological
16 risk assessment and economic valuation are unclear. The EPA (1992) framework for risk
17 assessment defines an assessment endpoint as: "an explicit expression of the environmental
18 value that is to be protected." However, as Suter (1990) makes clear, assessment endpoint,
19 "must be valued by society, but they are not ultimate values." An important, but not the
20 sole, measure for expressing social value are economic values.
21 Future valuation studies may demonstrate that complex, and often transparent,
22 ecological functions do not map into individual preferences. This would appear to be useful
23 information worthy of documentation. It would also not be a reason to discard experimental
24 techniques such as CVM.
25
26 Identify Limits to Application of Valuation Techniques
21 The measurement of nonuse values using the contingent valuation method remains
28 controversial. Given ethical limitations to markets in various social spheres (Anderson,
29 1990), there is a limit to the things to which we can meaningfully assign values. As one
30 example, the assignment of monetary values as expressions of the worth of certain personal
31 or social relations may change the nature of the relationship. As a second example,
December 1993 5-272 DRAFT-DO NOT QUOTE OR CITE
-------
1 following Bishop and Woodward (1993), and Howarth and Norgaard (1992), we cannot
2 expect nonmarket values to resolve large-scale sustainability and intergenerational equity
3 issues. Economic efficiency criteria provide no guarantee of the protection of any particular
4 environmental assets (Foy, 1990) or ecological sustainability (Common and Perrings, 1992).
5
6 Identify the Degree of Substitutability Between Natural and Man-Made Capital
1 Neoclassical economics has no traditional notion of sustainability other than
8 intertemporal efficiency. A critical issue in the sustainability literature is the degree of
9 substitutability between man-made and natural capital. Substitutability can occur on several
10 levels. The first question is whether man-made capital is an adequate substitute for the full
11 range of complex ecological functions. The corollary question is the degree of
12 substitutability within the structure of human preferences. Valuation methods based on
13 willingness to pay and willingness to accept compensation assume an underlying
14 substitutability. Economic values are exchange values based on the notion of indifference—
15 they express the terms of trade (money for environmental services) and money is the good
16 which accesses other exchange commodities. In short, assigning an economic value to
17 critical natural capital assumes an inherent substitutability for that asset in individual
18 preference functions.
19
20 5.8.4.6 Valuing Ecosystem Service Flows: Summary
21 Economists have a variety of valuation techniques to help guide policy choices
22 concerning the effects of air pollution or other environmental change on environmental
23 assets. Applying these techniques to ecosystem management issues and valuing the full range
24 of ecosystem service flows is a new, and as yet, unresolved challenge. Current
25 state-of-the-art assessment methods are judged inadequate for environmental regulations.
26 Many scholars, in both ecology and economics, are inherently skeptical of any economic
27 valuation of the full complex of ecosystem services, and turn towards other value indicators.
28 Costanza (1991) argues for methodological pluralism, where the imperfect information gained
29 from multiple approaches is combined in an integrated informational structure. The
30 identified research agenda for valuing ecosystem service flows crosses traditional disciplinary
31 boundaries (Russell, 1993). Interdisciplinary dialogue, cooperation, and development of
December 1993 5_273 DRAFT-DO NOT QUOTE OR CITE
-------
1 shared language are necessary for successfully designing future valuation experiments
2 concerning ecosystem service flows, and determining the proper role for such valuation.
3
4 5.8.5 Summary
5 The 1986 Criteria Document contained a review of assessments of the economic
6 consequences of O3 on U.S. agriculture. This section has evaluated selected post-1986
7 literature on the same topic. In addition, the review has been expanded to include potential
8 economic effects on forests and ecosystems.
9 Based on economic assessments and scientific data available at the time, the previous
10 criteria document (U.S. Environmental Protection Agency, 1986) concluded that O3 at
11 ambient levels was imposing economic costs on society. The review of more recent (post-
12 1986) literature on agriculture corroborates that earlier conclusion. Specifically, the recent
13 literature, using the full set of NCLAN data and addressing some deficiencies in the pre-1986
14 assessments, confirms the finding of economic losses from ambient O3 concentrations.
15 The exact level of these economic effects is a function of cropping patterns,
16 O3 concentrations (both ambient and episodic), and the spatial and temporal characteristics of
17 projected or observed O3 levels. The current economic assessments represent improvements
18 in the scientific understanding of O3 effects on agriculture. However, the assessments of
19 economic effects initially incident on the agricultural sector remains incomplete.
20 Only a few assessments consider the economic effects of O3 on forest trees as well as
21 urban trees, shrubs, and ornamentals. These studies assess the economic effects of
22 hypothetical changes resulting from O3 or other stressors on forest productivity and aesthetics
23 and are best viewed as measures of the potential effect of O3 on these receptors.
24 Improvements linking O3 effects data to productivity and aesthetic effects will improve the
25 utility of such economic analyses.
26 The effects of O3 on ecosystems have not been addressed in the published literature.
27 There is, however, an emerging interest in applying economic concepts and methods to the
28 management of ecosystems. Ecological research is also addressing the challenging
29 conceptual and practical issues in understanding and managing ecosystem functions.
30 Economic research continues to develop, refine, and apply techniques for valuing market and
31 nonmarket products and services that will be of help in estimating the economic effects of O3
December 1993 5-274 DRAFT-DO NOT QUOTE OR CITE
-------
1 on ecosystems. Increased dialogue between the disciplines is needed before empirical
2 analyses of the economic consequences of ecosystem management are feasible.
3 In summary, the state of science concerning O3 economic effects on agricultural crops
4 is sufficient to conclude that O3 imposes costs on society. Conclusions regarding effects on
5 forests and ecosystems must await the acquisition of additional data and possible refinements
6 in ecological and economic methods.
7
8
9 5.9 INTEGRATIVE SUMMARY AND CONCLUSIONS FOR
10 VEGETATION AND ECOSYSTEM EFFECTS
11 5.9.1 Introduction
12 Review of the post-1986 literature has not altered the conclusions of the earlier Criteria
13 Document (U.S. Environmental Protection Agency, 1986) or Supplement (U.S.
14 Environmental Protection Agency, 1992). In the 1986 criteria document, several general
15 conclusions were drawn from various experimental approaches: (1) current ambient ozone
16 concentrations in many areas of the country were sufficient to impair growth and yield of
17 plants; (2) effects occur with only a few hourly occurrences above 0.08 ppm; (3) growth and
18 yield data cited in the 1978 criteria document (U.S. Environmental Protection Agency, 1978)
19 indicate several species exhibited growth and yield effects when the mean ozone
20 concentration exceeded 0.05 ppm for 4 to 6 h/day for at least 2 weeks; and (4) regression
21 analyses of NCLAN data developing exposure-response functions for yield reductions
22 indicated that at least 50% of the crops were estimated to exhibit a 10% yield reduction at
23 7 h seasonal mean ozone concentrations of 0.05 ppm or less. These conclusions remain valid
24 today. The 1992 Supplement reviewed the literature concerning the appropriate exposure
25 index for expressing O3 effects on vegetation, including evaluation of the (1) role of
26 exposure duration, (2) the role of peak concentrations, (3) the 7- and 12 h mean
27 concentrations, and (4) comparison of many possible exposure indices to summarize seasonal
28 exposures and relate to yield loss. It was concluded that the 7- and 12 h seasonal mean was
29 not appropriate because of its treatment of all concentrations equally, and the lack of
30 consideration of exposure duration. Experimental studies had indicated the influential role of
31 episodic, peak concentrations and the length of the exposure. A comparison of possible
December 1993 5.275 DRAFT-DO NOT QUOTE OR CITE
-------
1 exposure indices concluded that the preferred indices cumulated all hourly concentrations
2 during the growing season and preferentially weighted the higher concentrations. These
3 conclusions remain valid today.
4 The review of the post-1986 literature has revealed additional analyses of the NCLAN
5 database as well as several European crop yield loss studies, substantiating the effects
6 observed in this country. There has been only limited publication of experimental studies
7 directly addressing the role of each of the exposure components to further our understanding
8 of what should be appropriate exposure indices relating ambient O3 concentration to effects.
9 There have been, however, several retrospective analyses of NCLAN data that increase
10 confidence in the use of the peak-weighted, cumulative indices. Studies of forest tree
11 seedlings have substantiated pre-1986 studies indicating the sensitivity of a number of
12 species, at least as seedlings. Data concerning the response of large mature trees as
13 individuals or in stands are limited. Seedling growth response of some species is altered at
14 O3 concentrations observed in many areas of the United States. Studies of the role various
15 biotic and abiotic environmental factors play in the response of plants to O3 indicate the
16 complexity of determining the response of plants in natural ecosystems wherein the
17 interaction of species, genotypes, and the multitude of environmental influences dictate the
18 eventual response of the species or community in question.
19
20 5.9.2 Species Response and Ecosystem Response
21 Ozone injury was first observed on crop plants. Therefore crop plants have been
22 studied more intensively than those plant species growing in natural habitats. As a result,
23 most of the information concerning species response, to O3 comes from the studies of crop
24 plants. The number of crop species/cultivars for which response data is available, however,
25 is a mere fraction of the total plants grown as crops. This is especially true for native
26 herbaceous plants and trees where the studies are far fewer in number. Because of the
27 known wide range of sensitivities to O3 among the species that have been studied and even
28 among cultivars of individual crops (see Section 5.4.2), it is not possible to estimate the
29 sensitivity of any given species and/or cultivar that has not been investigated directly except
30 in very general terms.
December 1993 5-276 DRAFT-DO NOT QUOTE OR CITE
-------
1 One attempt at using our fragmented knowledge to develop a general framework of
2 response covering a range of species has been presented by Reich (1987). He described a
3 conceptual model for conifers, hardwoods and agricultural crops, based on considerations of
4 the carbon balance and growth of individual plants in relation to O3 exposure or uptake
5 (estimated from O3 concentration and stomatal conductance). For equivalent exposures to
6 O3 during a single growing season, he ranked these groups in order of decreasing sensitivity
7 as: crops, deciduous hardwoods, evergreen conifers. A similar ranking results when the
8 published response data are reviewed. The concentration of ozone causing yield losses in
9 crop species is slightly less than that observed to cause biomass reduction in deciduous
10 species. A slightly greater concentration of ozone causes reductions in total biomass in some
11 coniferous species. However it is important to note that the variation of this biomass
12 response within these large groupings by growth strategy is as great or greater than the
13 variation in response between the groups. Caution is also needed with such generalizations
14 because these groupings compare annual versus perennial growth of long-lived species, as
15 well as the influence on growth response of a vast array of environmental factors at play on
16 these species in natural ecosystems.
17 The focus of research for developing quantitative relationships between O3 exposure
18 and biological effects has been on the response of individual species for three reasons:
19 (1) Single species studies are achievable experimentally, including ease of developing
20 adequate experimental design and exposure technology. (2) In many instances the plants are
21 grown in monoculture (e.g., most crop plants, ornamentals, fruit and nut species, plantation
22 forests), and the inter-specific competition and diversity, which typifies natural communities
23 is not an issue. The predicted response may also include the environmental influences of its
24 growing environment (e.g., drought) that modify the exposure-response relationship.
25 (3) In systems that are comprised of a multitude of species (e.g., mixed forest stands,
26 pastures or grasslands) it is important to understand the response of the individual
27 components so that behavior of the system might be analyzed in a systematic fashion. The
28 underlying assumption is that understanding how a forest stand responds to O3 requires
29 knowledge of the response of each species within that stand as a starting point. The
30 interactions that typify the community are subject to O3 effects as well and may manifest
31 themselves as a measurable effect some time later as a result of these interactions. Up to this
December 1993 5-277 DRAFT-DO NOT QUOTE OR CITE
-------
1 time, aside from the study of the San Bernardino forests, the holistic approach to
2 understanding O3 effects has not been pursued. Exposing large woody and herbaceous
3 species growing in natural species assemblages and subjected to a large array of climatic and
4 biotic environmental factors over decades has not been possible.
5 It is important to note that there may be a limited potential for using individual plant
6 (species) responses to an environmental stress (such as O3 exposure) to predict population
7 and community response (Woodward, 1992). As one progresses from knowledge of a tissue
8 or organ's stress-response to the whole plant, population, community and ecosystem levels,
9 various feedback mechanisms may modify the propagation of the effect. Important
10 components of such feedback are the mechanisms of homeostasis which involve injury repair
11 (at the metabolic level) or various types of compensation (Tingey and Taylor, 1982).
12 Compensation, which may occur at all levels of organization, from the subcellular to the
13 ecosystem,invokes processes that counteract the detrimental effects of the stress. At the
14 ecosystem level, an effect on the growth rate of a particular species may not be translated
15 into a comparable effect on the growth rate of a population of the species, because of
16 changes in the intensity of competition (Woodward, 1992).
17 At the present time, most of our knowledge of O3 exposure concerns the effects on
18 individual plants or their parts. Although we have some information about effects at the
19 population level with some agricultural crops and some forest tree populations, the
20 information regarding the propagation of effects upward through the different hierarchical
21 levels within natural and forest ecosystems is limited.
22
23 5.9.3 How Does Ozone Affect Plants?
24 Plant growth and yield are the product of a series of biochemical and physiological
25 processes related to uptake, assimilation and translocation of ozone in the individual plant.
26 Ozone exerts a phytotoxic effect only if a sufficient amount reaches the sensitive cellular sites
27 within the leaf. To do this, it must diffuse from the ambient air into the leaf through the
28 stomata, which exert control on O3 uptake. Ozone effects will not occur if (1) the rate of
29 ozone uptake is low enough that the plant can detoxify or metabolize O3 or its metabolites;
30 or (2) the plant is able to repair or compensate for the effects. Cellular disturbances that are
31 not repaired or compensated are ultimately expressed as visible injury to the leaf and/or
December 1993 5-278 DRAFT-DO NOT QUOTE OR CITE
-------
1 effects on growth, yield, or both (Tingey and Taylor, 1982; U.S. Environmental Protection
2 Agency, 1986). Ozone effect is cumulative, resulting in net reductions in photosynthesis,
3 changes in allocation of carbohydrate, and early senescence, which leads to reductions in
4 biomass and productivity, and alterations in susceptibility to abiotic and biotic stresses and/or
5 decreased reproduction.
6 Ozone is expected to reduce growth or yield only if (1) it directly impacts the plant
7 process that is limiting to plant growth (e.g. carbon fixed); or (2) it impacts another step
8 sufficiently so that it becomes the step limiting plant growth (e.g., allocation of
9 carbohydrates to roots and nutrient uptake becomes limiting to plant growth) (Tingey, 1977).
10 Conversely, O3 will not limit plant growth if the process impacted by O3 is not growth-
11 limiting. This implies that not all effects of O3 on plants are reflected in growth or yield
12 reductions. These conditions also suggest that there are combinations of O3 concentration
13 and exposure duration that the plant can experience that may not result in visible injury or
14 reduced plant growth and yield (U.S. Environmental Protection Agency, 1986). However,
15 subtle physiological effects which may not result in immediate growth reductions may result
16 in increased plant susceptibility to other environmental factors and competition.
17 The mode of action of O3 on plant species described in the 1986 criteria document
18 (U.S. Environmental Protection Agency, 1986) still holds true. The plant leaf is the site
19 O3 action and the critical effect is on the plant's carbon budget (the amount of carbohydrate
20 produced). Inhibition of photosynthesis limits carbohydrate production and allocation
21 resulting in reduced biomass, growth and yield. Studies since 1986 corroborate this
22 understanding, adding information on the effect of O3 on photosynthetic capacity, respiration,
23 leaf dynamics, and on the detoxification and compensatory processes. In particular, exposure
24 to O3 concentrations at or near current ambient levels (see Table 5-17 Section 5.6) has an
25 effect on photosynthesis, but a longer exposure duration is necessary to produce a growth
26 response, taking days to weeks rather than hours as in earlier studies with higher
27 concentrations (0.25 ppm or greater). The loss of leaves prematurely as a result of
28 O3 exposure has been observed in several species and is particularly important in coniferous
29 tree species. This O3-induced loss in leaf area can be a significant factor in the reducing the
30 amount of carbohydrate produced by the plant. Both a reduced photosynthetic capacity and a
31 reduced leaf area contribute to the reduction in carbohydrate production by plants. However,
December 1993 5.279 DRAFT-DO NOT QUOTE OR CITE
-------
1 the mechanism of premature senescence is not understood. In addition to leaf loss, reports of
2 stimulation of production of new leaves and higher photosynthetic capacity of new leaves
3 represent compensation processes that operate in some species of trees. These mechanisms
4 may counteract the reduction in canopy carbohydrate production resulting from O3. Some
5 quantitative understanding of these processes is needed to be able to predict long-term effects
6 of O3 on tree species. More information is needed as well in understanding O3 uptake at the
7 canopy level and how the plant integrates the effects of 63 to enable prediction of long-term
8 effects of O3 exposure in ecosystems, i.e., species response as a function of interactions with
9 other species, abiotic and biotic environmental factors. Unfortunately, there is little
10 experimental evidence to date regarding effects of long-term O3 exposure on perennial plants.
11 Few experimental studies have extended exposures beyond one season and only in a limited
12 number of studies have observations of growth been extended into the following year, thus
13 observing "carry-over" effects in several tree species. These carry-over effects are
14 significant to long-lived species such as trees since they affect the elongation of new spring
15 shoots or root growth in the year following exposure to O3. In at least one instance this has
16 been correlated with reduced storage carbohydrate in roots. The implication of these effects
17 on long-lived species is significant. Reduction in growth and productivity, a result of altered
18 carbohydrate produced and allocated, may appear only after a number of years or when
19 carbohydrate reserves in the tree are depleted below some threshold concentration.
20
21 5.9.4 Factors That Modify Plant Response to Ozone
22 Plant response to O3 exposure is modified by factors within and external to the plant
23 species; cultivars and individuals within populations display variable response to O3. The
24 plant's response and the variation of that response is dictated by genetics and the plant's
25 present and past environmental milieu. The environment includes biotic and abiotic factors
26 of the species' growing environment, the temporal pattern of exposure concentrations, and
27 the plant's phenotypic stage during exposure.
28
29 5.9.4.1 Genetics
30 An important component of this variation is genetically controlled. The specific genes
31 controlling O3 response and involved in mechanisms of O3 tolerance are as yet largely
December 1993 5-280 DRAFT-DO NOT QUOTE OR CITE
-------
1 unknown. However, control of stomatal conductance and internal biochemical defense
2 systems are among the most commonly described tolerance mechanisms. Ozone tolerance is
3 generally thought to be controlled by multiple genes. The implications of genetic variation
4 for managed and natural ecosystems are several-fold: (1) The potential for natural selection
5 for O3 tolerance and associated loss of sensitive genotypes is regional in nature, unlike well
6 known point-source pollution impacts which occur on local plant populations. However, the
7 intensity of O3 selection is generally thought to be quite low, 0.3 or less (Taylor and Pitelka,
8 1992) across most U.S. areas. (2) While it is known that individual plants within a species
9 vary in their O3 tolerance, the physiological costs to tolerant plants are not known in terms
10 of carbohydrate assimilation (energy production) and allocation. Tolerance mechanisms
11 based on reduced stomatal conductivity in the presence of O3 would likely reduce the growth
12 of tolerant plants. Similarly, tolerance mechanisms based on the productivity of antioxidant
13 compounds will likely shunt plant resources away from growth to the production of the
14 defense compounds. (3) Exposure-response equations and yield-loss equations developed for
15 a single or small number of cultivars, genotypes, families or populations may not adequately
16 represent the response of the species as a whole. As a corollary to this, the sensitivity of
17 responder genotypes can not be determined by measuring effects just in relation to mean
18 O3 concentrations.
19
20 5.9.4.2 Environmental Factors
21 Since the previous criteria document (U.S. Environmental Protection Agency, 1986),
22 additional studies have been published on a wide range of biological, physical, and chemical
23 factors in the environment that interact with plant response to O3. While understanding the
24 plant's response to O3 requires an understanding of the role of environmental factors that
25 modify that response (primarily to reduce uncertainty in the estimation of species' exposure-
26 response), the corollary is also important to understand, and in fact has not received as much
27 attention, i.e., the exposure to O3 can modify the plant's ability to integrate its environment.
28 For example, exposure to O3 reduces the tree's ability to withstand winter injury caused by
29 exposure to freezing temperatures and influences the success of pest infestations.
30 Biological components of the environment of individual plants include pests, pathogens
31 and plants of the same or other species in competition. With regard to insect pests, although
December 1993 5_281 DRAFT-DO NOT QUOTE OR CITE
-------
1 only a very limited number of plant-insect systems have been studied, there is a general trend
2 in the observations which suggests that some pests have a preference for and grow better
3 when feeding on plants that have been impacted by O3, but there is no evidence to suggest
4 that O3 may trigger pest outbreaks. Unfortunately, because we have no knowledge of how
5 the vast majority of plant-insect systems will be affected by O3, it is not possible to offer any
6 quantitative overall assessment of the consequences of such interactions on the growth of
7 crops and natural vegetation. At best, we may conclude that there is a reasonable likelihood
8 that some insect pest problems will increase as a result of increased ambient O3 levels.
9 Indeed, this was seen in the San Bernardino Mountains where ponderosa pine experienced
10 bark beetle infestations with higher O3 exposures (U.S. Environmental Protection Agency,
11 1986).
12 Plant-pathogen interactions also appear to be affected by O3. The suggestion that
13 diseases caused by obligate pathogens tend to be diminished by O3 while those caused by
14 facultative pathogens tend to be favored (Dowding, 1988), is generally supported by the
15 limited evidence available. In terms of its broader implications, this suggests that continued
16 exposure to O3 may lead to a change in the overall pattern of the incidence and severity of
17 specific plant diseases affecting crops and forest trees.
18 With regard to the physical environment, the combination of light, temperature and
19 water availability largely determines the success of plant growth because of the influence of
20 these factors on the processes of photosynthesis, respiration and transpiration. For
21 agricultural crops, perhaps the most important of these potential interactions with
22 O3 concerns water availability and use. There is consistent evidence that severe drought
23 conditions tend to reduce the direct adverse effects of O3 on growth and yield. Conversely,
24 the ready availability of soil water tends to increase the susceptibility of plants to O3 injury.
25 However, a lack of water should not be viewed as a potentially protective condition, because
26 of the adverse effects of drought per se. With perennial trees, there is some evidence that
27 prolonged exposures to O3 may lead to greater water use efficiency which may enable such
28 trees to be better able to survive drought conditions.
29 The plant's environment also contains numerous chemical components, ranging from
30 soil nutrients and other air pollutants to agricultural chemicals used for pest, disease and
31 weed control.
December 1993 5-282 DRAFT-DO NOT QUOTE OR CITE
-------
1 A large number of studies have been conducted on the effects of O3 in conjunction
2 with other gaseous air pollutants such as SO2 and NO2, although the information obtained in
3 several of the studies is of no more than academic interest because of the unrealistic exposure
4 conditions used. Preliminary evidence suggests that nitrate deposition in the soil could
5 influence plant response to O3.
6 Since increased tropospheric O3 is a component of global climate change, results from
7 studies on the interactions of O3 with increased levels of CO2 and UV-B radiation are
8 beginning to appear. Initial work with CO2 suggests that increased CO2 levels may
9 ameliorate the effects of O3. However, it is too soon to be able to generalize on the outcome
10 of this interaction. At the present time, no investigations of the compound interactions
11 involving O3, CO2, UV-B, increased temperature and changed soil moisture status have been
12 reported.
13 In conclusion, in spite of the amount of work carried out to date on the interactions of
14 O3 with environmental factors, there is not a strong quantitative database to generalize and
15 extend the effects of O3 on species across environments. Uncertainty about the extent of
16 modification of the exposure-response relationship as a result of different growing
17 environments remains, as well as the uncertainty associated with understanding O3's role in
18 altering a species' ability to integrate its environment.
19
20 5.9.5 Exposure Dynamics
21 The effects of O3 on individual plants is not only a function of inter- and intra-
22 specific differences in tolerance (genetics), and growing environment affecting the uptake of
23 O3 (e.g., soil water, nutrient), but also the characteristics of the exposure itself which affect
24 the uptake of O3 from ambient air, and thus the ultimate growth response. From the studies
25 prior to 1986 and after 1986, evidence indicates that the components of exposure, i.e. peak
26 concentration, frequency of occurrence and duration, play various roles in the plant response.
27 The temporal pattern of hourly concentrations in the exposure influence the response. There
28 is a reported greater influence on growth in annual and perennial crop species (bush beans
29 and alfalfa), and in tree seedlings (ponderosa pine and aspen) of episodic peak exposures
30 compared to either daily peak occurrences in the case of crops or non-diurnal, continuously
31 elevated exposures typical of remote regions in the case of tree seedlings. The results
December 1993 5_283 DRAFT-DO NOT QUOTE OR CITE
-------
1 suggest a variation in the growth response of species as a function of the dynamics of the
2 exposure. In addition to the temporal distribution of concentration during a day and during a
3 season, there is also the distribution of the exposure during the growing season and the
4 phenology of the plant. Some phenotypic stages of growth are more sensitive to O3 than
5 others. In most instances of crop exposure studies, in particular the NCLAN database, the
6 exposure treatments used in developing response functions have used ambient concentrations
7 as a base. Thus, the variation in the response as the result of a variation in regime type is
8 not known. The same reservation is applicable to the growth response reported for tree
9 seedlings. It is not possible at this time to quantify the alterations in the growth or yield
10 response as a result of different exposure patterns. However, it is possible to suggest which
11 components of an exposure play a relatively greater influence in growth alterations than do
12 other components of the exposure. For example, peak concentrations are more effective than
13 lower concentrations in altering growth; and the episodic occurrence of peak events seems to
14 be more damaging than the daily occurrence of the same peak value or a continually elevated
15 concentration over a growing season. The variation in species' response as a function of
16 exposure concentration and duration suggests the importance of a measure of the ambient air
17 exposure to relate to the biological effects.
18
19 5.9.6 What Measure of Exposure Characterizes Species Effects?
20 A measure is needed to relate ambient exposure to the observed biological effect(s).
21 The amount of O3 taken up from the atmosphere by the plant (either rate of uptake or
22 cumulative seasonal uptake) is the ultimate measurement of O3 exposure, but this is not a
23 practical measure to make directly. The uptake rate and total cumulative uptake can be
24 calculated from ambient concentrations and species' gas exchange characteristics. A measure
25 of uptake integrates all those environmental influences on stomatal conductance controlling
26 the effect, e.g., temperature, humidity, soil water status, etc., but it is not practical for most
27 experimental approaches used in determining what ambient concentrations of O3 can be
28 tolerated by vegetation. Thus an index of ambient exposure is required as a surrogate of
29 uptake. Any index that relates well to plant response should incorporate these environmental
30 and exposure dynamic factors, directly or indirectly, in weighting the hourly
31 O3 concentrations differentially.
December 1993 5-284 DRAFT-DO NOT QUOTE OR CITE
-------
1 Given the evidence suggesting a relatively greater contribution to effects of higher
2 concentrations, episodic occurrence of peak concentrations, and the duration of the exposure,
3 an index that cumulates hourly concentrations during the season and gives greater weight to
4 the higher concentrations appears to be an appropriate index for relating ambient exposure to
5 growth or yield effects. A number of different forms of a peak-weighted, cumulative index
6 have been examined, e.g., SUM06, SIGMOID, W126, for their ability to properly order
7 yield response in the large number of crop yield studies of NCLAN. All perform equally
8 well and it is not possible to distinguish between them on the basis of statistical fits of the
9 data. In retrospective analyses when O3 is the primary source of variation in response, year
10 to year variations in plant response are minimized by the peak-weighted, cumulative exposure
11 indices. No experimental studies, however, have been designed specifically to evaluate the
12 adequacy of the various peak-weighted indices.
13 This document's conclusions are no different than those of the 1986 criteria document
14 or the 1992 Supplement (U.S. Environmental Protection Agency, 1986, 1992): (1) mean
15 indices are not among the best exposure indices; and (2) the preferred indices are those
16 cumulating hourly values over the growing season and preferentially weighting the peak or
17 higher concentrations.
18 The 7- and 12-h seasonal means were earlier indices used to describe plant response.
19 The mean indices have been shown to be inadequate in ordering growth response in
20 experimental studies as well as retrospective analysis of NCLAN data (U.S. Environmental
21 Protection Agency, 1986). This is not surprising since a mean index implies that all
22 concentrations are equal in their effect on plants, and does not consider duration of exposure.
23 Exposure duration is influential in the response, as well as the timing of exposure during a
24 growing season coinciding with sensitive plant growth stages. Phenology is one factor, other
25 than peak events and duration, that is influential in response. Others such as respite time,
26 canopy structure, environmental conditions, e.g., soil and nutrient conditions, are considered
27 important but not well understood. All these factors interact with concentration and duration
28 in different fashions depending on species. The peak-weighted cumulative indices do a better
29 job of considering at least a part of this information on environment, phenology, etc., than
30 the mean or single-event peak indices, because of their inclusion of concentration weighting
31 and duration. Recently, the mean O3 flux was used as an index and it minimized the year to
December 1993 5_2g5 DRAFT-DO NOT QUOTE OR CITE
-------
1 year variation in response when combining replicate studies, indicating the importance of
2 environmental conditions in quantifying the relationship between O3 exposure and plant
3 response.
4 5.9.7 What Is the Estimated Crop Yield or Biomass Change with Ozone
5 Exposure?
6 In the 1986 criteria document (U.S. Environmental Protection Agency, 1986) and its
7 supplement (1992) a distinction was made between foliar injury and damage. Yield loss was
8 defined as an impairment in the intended use of the plant. This concept included reductions
9 in aesthetic values, foliar injury (changes in plant appearance), and losses in terms of weight,
10 number, or size of the harvested plant part. Yield loss may also include crop quality.
11 Losses in aesthetic values are difficult to quantify. Foliar injury symptoms can substantially
12 reduce the marketability of ornamental plants or crops where foliage is the plant part (e.g.,
13 spinach, lettuce, cabbage), and they constitute yield loss with or without concomitant growth
14 reductions. It should also be recognized that foliar injury to vegetation in national parks
15 constitutes reduction in aesthetic value. At that time (1986), most studies of the relationship
16 between yield loss and O3 concentration focused on yields as measured by weight of the
17 marketable organ of the plant. From this perspective, damage can be segregated into several
18 categories: (1) economic; (2) ecological value of structure and function; (3) genetic
19 resources; and (4) cultural values (Tingey et al., 1990). In the instance of crop species, the
20 effect of interest was yield, at least in the NCLAN research program. In the instance of tree
21 species, much of the research has sought to measure changes in biomass or productivity,
22 which could relate to all four of the above effect categories.
23 Diverse experimental procedures, ranging from field exposures without chambers to
24 open-top chambers and to exposures conducted in chambers under highly controlled climates,
25 have been used to study effects on crops and tree seedlings. In general, the more controlled
26 conditions are most appropriate for investigating specific responses and for providing the
27 scientific basis for interpreting and extrapolating results. These systems are powerful tools
28 for adding to an understanding of the biological effect of air pollutants. However, to assess
29 the economic impact of O3 on crop yield or biomass partitioning in tree species, exposures
30 should have minimum deviations from the typical field environment in which the plant is
31 grown. Much of the data reviewed in this document and in the 1986 ozone criteria document
December 1993 5-286 DRAFT-DO NOT QUOTE OR CITE
-------
1 (U.S. Environmental Protection Agency, 1986) utilized the open-top chamber methodology.
2 This approach has been the primary methodology for developing the empirical database of
3 O3 effects on crop yield during the last 15 years. These are used because of the control they
4 offer over exposures and still offer some semblance of relevance to field condition, and the
5 ability to replicate studies from year to year. The NCLAN studies used sufficient treatments
6 to allow development of exposure-response functions and became the largest database
7 available for establishing a quantitative relationship between exposure and biological effect in
8 crop and tree species (as seedlings). The response functions allow estimations and
9 generalizations of biological response to O3 unlike information from multiple comparisons.
10 Two different approaches to re-analysis of NCLAN data have occurred. The National
11 Crop Loss Assessment Network studied the major agronomic crop species,including corn,
12 soybean, wheat, cotton, bean, and alfalfa, as well as several other regionally important
13 species accounting for 70% of all crop land in the United States and 73% of the agricultural
14 receipts. One approach estimated yield reduction of up to 20% at 12 h seasonal mean of
15 0.06 ppm and a 10% reduction at 0.045 ppm. In another approach, they compared estimated
16 yield reduction in 54 studies and 12 crop species with 3 measures of exposure. The average
17 duration of the studies was 74 days. It was concluded that 50% of the crops would
18 experience 10% yield reduction at a 3 mo Sum06 concentration of 26.4 ppm-h, 7 h seasonal
19 mean of 0.049 ppm, or a 2HDM of 0.094 ppm. These are averaged yield losses for all
20 species; more sensitive species would experience greater yield losses at these concentrations.
21 The SUM06 has been examined in many papers; therefore, it has been used as an example in
22 the discussions of O3 effects that follow.
23 Similar results have been reported from European crop studies. Wheat yields were
24 reduced up to 29% depending on the exposure and cultivars, but in no instance were
25 O3 concentrations greater than a 0.062 ppm 7 h seasonal mean. Spring rape yields were
26 reduced 9 to 26% at 8 h seasonal means of 0.03 to 0.06 ppm. Seasonal 7 h means of
27 0.045 ppm reduced bean yields 17%.
28 5.9.8 Ozone Concentration Across the United States
29 The O3 concentrations causing 10% or greater yield reductions in crop species are
30 concentrations that are not atypical of O3 air quality during the past 10 years across the U.S.
31
December 1993 5_287 DRAFT-DO NOT QUOTE OR CITE
-------
1 - Eastern Seaboard
2 - Appalachian Highlands
3 - Greatiakes and Ohio Valley
4 - Interior Plains
5 - South
6 - Intermountain
7 • Southern California
8-North Pacific Coast
Figure 5-27. Regions of the United States for analysis of trends of ozone concentration.
1.E«m«nS««bo |