EPA-600/9-75-004 *
PROCEEDINGS OF THE FIRST US/USSR
SYMPOSIUM ON COMPREHENSIVE
ANALYSIS OF THE ENVIRONMENT
MARCH 25-29. 1974
-------
EPA-600/9-75-004
PROCEEDINGS OF THE FIRST US/USSR
SYMPOSIUM ON COMPREHENSIVE
ANALYSIS OF THE ENVIRONMENT
-------
Preface
These proceedings include the 24 papers presented at the first U.S./USSR
Symposium on Comprehensive Analysis of the Environment. The papers were
originally presented in English and Russian at Tbilisi, Georgia, USSR, between
March 25 and March 29, 1974. Their publication herein is part of the protocol
agreement signed on March 29, 1974, for simultaneous and independent publica-
tion in both countries after coordination of the manuscripts.
Acknowledgement
The success of the first symposium on Comprehensive Analysis of the Environ-
ment resulted from the efforts of many persons who planned the agenda, arranged
facilities, coordinated papers and their translation, and attended to the myriad of
details which occurred through the symposium. The productivity of the 1974 sym-
posium and its fruitful exchange of ideas were a vivid realization of the objectives
in the Agreement on Cooperation in the Field of Environmental Protection, signed
in Moscow on May 23, 1972.
Achievement of these plans is a special credit to the Soviet delegation under
their leader, Professor Yu. A. Izrael, First Deputy Chief of the Main Administra-
tion of the Hydrometeorological Service, Council of Ministers of the USSR. Since
the Tbilisi symposium, now Academician Izrael has succeeded Academician E. K.
Fedorov as Chairman of the Soviet Side of the Joint Committee on Cooperation
in the Field of Environmental Protection. The warm appreciation for all his help
has been duly noted by the head of the American delegation, Dr. Stanley M.
Greenfield, Assistant Administrator for Research and Development, U.S. Environ-
mental Protection Agency. The American contributions were compiled by Dr.
Leland D. Attaway, Deputy Assistant Administrator for Program Integration, U.S.
Environmental Protection Agency. Completion of these proceedings has been
directed by Roger S. Cortesi, Director of the Washington Environmental Research
Center, U.S. Environmental Protection Agency, who succeeded Dr. Attaway as
Chairman of this U.S. Working Group Project.
m
-------
Contents
Page
I. THEME-ANALYSIS TO SUPPORT ENVIRONMENTAL
MANAGEMENT
Comprehensive Environmental Management, an Overview—
S. M. Greenfield 1
II. COMPREHENSIVE MODELLING
Systems Analysis and Simulated Mathematical Modelling as a
Methodological Basis for the Standardization of Anthropo-
genic Approach—Yu. A. Izrael, Yu. A. Anokhin 5
A Comprehensive Environmental Analysis of the Upper Potomac
Estuary Eutrophication Control Requirements—N. A.
Jaworski 15
The Ecological System Stability Concept—V.D. Fedorov 27
Development of Maximum Permissible Environmental Loading-
Air—B. J. Steigerwald 35
III. MAXIMUM PERMISSIBLE LOADING
Estimating the Maximum Permissible Intensity of the Complex
Effect of Chemical Factors in the Production, Municipal, and
Domestic Environment on Man—N. F. Izmerov 45
The Concept of the Threshold Nature of Reaction of Living Sys-
tems to External Actions and its Consequences in the Problem
of Protecting the Biosphere against Chemicals—I. V. Sanot-
skiy 49
Questions of Extrapolating Data in Evaluating the Mutagenic
Action of Environmental Factors on Man—N. P. Bochkov .... 55
The Question of Genetic Danger from Environmental Pollu-
tants—L. M. Filippova 59
Hygienic Bases for Protecting the Environment—G. I. Sido-
renko, Ye. I. Korenevskaya, G. N. Krasovskiy 63
Hygienic Normalization under Scientific and Technical Progress
Conditions—A. P. Shitskova, M. I. Gusev, S. M. Pavlenko,
I. L. Karagodina 67
Maximum Permissible Human Stress—V. Newill 71
Maximum Permissible Human Stress with Multiple Pathways:
lonization Radiation, Asbestos and Lead—J. H. Knelson 81
Complex Analysis of the Environment—Approaches to a Deter-
mination of the Permissible Loads on the Natural Environ-
ment and to a Justification for Monitoring—Yu. A. Izrael .... 89
Defining "No Significant Deterioration" of Air Quality—R. A.
Luken 95
IV. ECONOMIC CRITERIA FOR ESTABLISHING ENVIRON-
MENTAL QUALITY GOALS
Determining Acceptable Levels of Health and Environmental
Damages—F. H. Abel 107
-------
Pat*
V. IMPLEMENTATION STRATEGIES FOR ACHIEVING AND
MAINTAINING ENVIRONMENTAL STANDARDS
Methods for Abatement of Water Pollution—R. B. Schaffer .... 117
Methods for Control of Air Pollution—K. E. Yeager 125
Development of Maximum Permissible Environmental Loading
(MPEL)-Water—M. A. Pisano 139
VI. IMPACT OF POLLUTION AND ENVIRONMENTAL CON-
DITIONS ON NON-HUMAN SPECIES
The Effect of Polluting a Biogeocoenosis with Strontium 90 on
a Population of Mammals and the Zoocoenosis—V. Ye.
Sokolov, A. I. Ilycnko, D. A. Krivolutskiy 149
Effects of Pollution on Species and Populations of Fish and
Birds—H. E. Johnson 159
The Influence of Environmental Factors on Population Dynamics
(Mathematical Models)—M. Ya. Antonovskiy 171
VII. POLLUTION IMPACTS ON CLIMATE
Global Climate and Human Activity—M. I. Budyko, I. L. Karol 179
VIII. SUMMARY AND RECOMMENDATIONS
Concluding Remarks for the U.S./USSR Symposium on Compre-
hensive Analysis of the Environment—L. D. Attaway 187
vi
-------
Attendees
AMERICAN PARTICIPANTS
Dr. Stanley M. Greenfield
Dr. Leland D. Attaway
Dr. Frederick H. Abel
Dr. Norbert A. Jaworski
Dr. Howard E. Johnson
Dr. John H. Knelson
Dr. Ralph A. Luken
Dr.VaunNewill
Dr. Mark A. Pisano
Mr. Robert B. Schaffer
Dr. Louis J. Schoen
Dr. Bernard J. Steigerwald
Mr. Kurt E. Yeager
SOVIET PARTICIPANTS
Prof. Yu. A. Izrael
Dr. Yu. A. Anokhin
Prof. M. Ya. Antonovsky
Head of Delegation
Assistant Administrator for Research and Devel-
opment, USEPA
Co-chairman, Comprehensive Analysis of the En-
vironment Project
Deputy Assistant Administrator for Program Inte-
gration, USEPA
Chief, Economic Analysis Branch, Washington En-
vironmental Research Center, USEPA
Director, Pacific Northwest Environmental Research
Laboratory, USEPA
Private Citizen, Associate Professor, Department of
Fisheries and Wildlife Michigan State University
Acting Director, Human Studies Laboratory, USEPA
Private Citizen, Consultant on Environmental Eco-
nomics
Special Assistant to the Administrator, USEPA
Director, Water Planning Division, USEPA
Acting Director, Permit Assistance and Evaluation
Division, USEPA
Special Assistant for International Research and
Development, USEPA
Director, Office of Air Quality, USEPA
Chemical Engineer, Office of Research and Devel-
opment, USEPA
Co-chairman, Comprehensive Analysis of the En-
vironment Project
First Deputy Chief of Main Administration of Hydro-
meteorological Service, Moscow
Institute of Applied Geophysics, Moscow
Central Institute of Economic Mathematics, Moscow
vu
-------
Prof. M. E. Berljand
Prof. M. I. Budyko
Prof. V. D. Fedorov
Dr. L. M. Filippova
Prof. M. I. Gusev
Prof. G. F. Hilmi
Prof. A. I. Ilyenko
Dr. N. F. Izmerov
Dr. V. S. Jurkov
Prof. I. L. Karol
Dr. Yu. E. Kazakov
Dr. G. N. Krasovsky
Dr. B. A. Kuvshinnikov
Prof. M. J. Lemeshev
Dr. V. P. Lominadze
Prof. A. M. Molchanov
Dr. I. M. Nazarov
Dr. V. G. Novozhilov
Prof. I. V. Sanotsky
Acad. S. S. Shvarts
Main Geophysical Observatory, Leningrad
Main Geophysical Observatory, Leningrad
Moscow State University, Moscow
Institute of Applied Geophysics, Moscow
Institute of Hygiene, Moscow
Institute of Applied Geophysics, Moscow
Institute of Evolutional Morphology and Ecology of
Animals, Moscow
Institute of Labour Hygiene and Professional Dis-
eases, Moscow
Institute of Medical Genetics, Moscow
Main Geophysical Observatory, Leningrad
Joint Soviet-American Committee, the Soviet Part
of the Project
Institute of General and Communal Hygiene, Mos-
cow
Joint Soviet-American Committee, the Soviet Part
of the Project
Central Institute of Economic Mathematics, Moscow
Director of Transcaucasian Hydrometeorological In-
stitute, Tbilisi
Computer Centre, Puschino-na-Oke
Institute of Applied Geophysics, Moscow
Assistant Chief of Main Administration of Hydro-
meteorological Service, Moscow
Institute of Labour Hygiene and Professional Dis-
eases, Moscow
Institute of General Biogeocenology and Ecology,
Sverdlovsk
vm
-------
Introduction
The Joint Committee on Cooperation in the Field of Environmental Protection
was established by the signing of an agreement in Moscow, May 23, 1972. This
was followed by the Second Session of the U.S./USSR Joint Committee in the
Field of Environmental Protection (Washington, D.C., November 16, 1973).
From these organizational meetings, eleven areas were specified for cooperative
investigation and exchanges. Within these broad areas a number of projects have
been identified. These proceedings are based on one of these projects: Project
VII-2.1, Comprehensive Analysis of the Environment, within Area VII—Bio-
logical and Genetic Ecects of Environmental Pollution. In accordance with pro-
visions of the Memorandum of the Second Session, the symposium at Tbilisi was
scheduled to bring together experts from both countries for presentation and dis-
cussions of papers in this area. This area is unique in the comprehensive perspec-
tive of its approach to environmental problems; consequently the contents or titles
of some of the papers could suggest that they are more germane to other program
areas.
The proceedings are arranged to present papers in topical clusters, and do not
necessarily reflect the original sequence of presentations. Papers from each dele-
gation are intermingled to emphasize the contrast in approaches, and suggest the
lively discussions which they engendered.
Primary emphasis was given to the determination of scientifically permissible
loading on the environment. This emphasis considered the comprehensive analysis
of the pollution sources and receptors as well as approaches to the development
of criteria for scientifically permissible stress to the environmental and identification
of critical links in this development.
The Tbilisi symposium identified the following problems which will require at-
tention:
(1) More complete accounting of pollution sources and receptors, types of im-
pacts and biological reactions, as well as peculiarities of the area or region where
such impact occurs;
(2) Investigation of dose/response relationships in lexicological, genetic, and
other experiments;
(3) Scientific substantiation of environmental monitoring techniques with due
consideration of critical links in analyzed systems;
(4) Studies of the fate and effects of multimedia pollutants, supplemented by
development of the required standards;
(5) Assessment of the needs for separate standards limiting long- and short-
term dosages of selected environmental pollutants;
(6) Methods for evaluaing the response of ecosysems to varying environmental
conditions;
(7) Development of criteria for scientifically permissible environmental loading
and identification of critical links;
(8) Development of practical quantitative methods which will permit compre-
hensive environmental analysis to be applied effectively to environmental problems;
(9) Development of methodology for assessment of possible economic damage
resulting from man's impact on the environment; and
(10) Feasibility of a joint U.S.-USSR monitoring system.
The next meeting on the Comprehensive Analysis of the Environment, or Sym-
posium II, is scheduled for the Fall of 1975, and will have as its theme:
Maximum Acceptable Environmental Stress on Organisms, Populations, Eco-
systems and the Biosphere as a Whole.
ix
-------
Comprehensive Environmental Management—An Overview
Stanley M. Greenfield *
From a regulatory sense, it is possible to sepa-
rately consider individual pollutants emerging from
single sources into a single medium. Experience has
shown, however, that to derive a strategy that will
ultimately result in environmental quality, one must
comprehensively consider the entire complex inter-
acting environmental system within the context of
operative socio-economic constraints. Such consider-
ation must take into account the distribution of pol-
lutant sources throughout the spatial region in ques-
tion, and their intensity variation with time. What
then must be derived is a mechanism for specifying
the required temporal and spatial variation in effluent
so as to produce the desired environmental quality.
Implicit in such a determination is the recognition
that ultimately there must be a balance in the cost
of achieving this level of quality and the benefit that
accrues to society in its achievement. All of human
activity, in one way or another stresses the environ-
ment. The question is, how much stress is acceptable,
and how is such a limit achieved and maintained?
Obviously, within limits, a society could choose the
degree of environmental management it desires, the
limits on the choice of degree being when it permits
those levels of stress to occur which produce irre-
versible, deleterious impacts on the society itself. The
basic question at issue, however, is how does the
society properly make such a choice? And, within
the context of that question, what information is re-
quired to make such a decision? What trade-offs are
to be considered? What are the potential methods of
implementation of any given management strategy?
What analytical tools are available and/or required?
What is the proper relationship between short-term
goals to accomplish an immediate end, and long-
term goals to find a lasting solution?
First on any list of information required for such
a decision process must be some determination of the
maximum permissible species stress. This concern is
conveniently divided into two basic areas, first the
effects associated with human health, and second,
those associated with the other segments of the eco-
logical system.
In the case of health effects, we immediately face
the dilemma of social risk and its acceptable level.
Are we to consider the economic questions associated
with protecting populations, and are these to enter
into any determination of maximum permissible
stress? The first inclination is to dismiss economic
and social risk as an inappropriate consideration
within any discussion of human health. This is par-
ticularly easy to do when one is faced with a health
effect threshold clearly defined and applicable to the
population as a whole. In general, however, this
simple case is seldom realized. What we are normally
faced with is either a dose-response curve with con-
siderable uncertainty near zero dose, or a threshold
that varies with different subgroups of the population
(e.g., cardiovascular sufferers, etc.). The dilemma is
clear in both cases. It is possible or desirable to pro-
vide a level of maximum permissible risk that is not
protective of the entire population? Conversely, is it
desirable or necessary to protect down to the last
noticeable effect, when to do so may require literally
the forbidding of the flow of any quantity of a par-
ticular pollutant into the environment over and above
the natural background?
In addition to the relatively simple case of the
single pollutant as a unique stressor in a unique path-
way, we must consider the implication of the syner-
gistic and nonsynergistic effects of multiple stressors,
and the equally complex multiple pathway single pol-
lutant case. In the rst case, one is possibly implying
a lowest common denominator maximum permissible
stress that depends on what other pollutants are pres-
ent in the environment. In the latter case, such as
lead and its many pathways to man, we are faced with
a maximum permissible stress depending on the
mechanism of uptake and some consideration of the
relative intensity of various intake models available.
The data that underlie the determination of the
maximum permissible stress for a given pollutant are
not particularly impressive. They become even less
impressive when one considers multiple pathway
and/or multiple pollutants. Epidemiologic experi-
ments are difficult to carry out and when stratified to
reveal the single or specific multiple pollutant effect,
quickly reduce even a large body of data to an in-
sufficient sample. Animal lexicological experiments
are difficult to carry out successfully for the long
•AMittaat Administrator for Research and Development, U.S. Environmental Protection Agency
-------
periods of time required, and suffer in their transla-
tion to man. Yet, the ability to state the maximum
permissible stress is the cornerstone to being able to
describe an adequate control plan, and to provide
the input options for an adequate environmental
management system. In addition, adequate ability to
quantify the stressor-effect relationship is mandatory
if we are to ultimately describe the social conse-
quences of various levels of pollutant concentrations.
Of particular interest here is the question of non-
degradation, or the decision to not permit a region
that is currently below the maximum permissible
stress to degrade so as to reach that limit. This is an
extremely difficult question to deal with from both
a philosophical and social standpoint. It starts with
the assumption that we have established a maximum
permissible stress. Philosophically or pragmatically,
do we now allow all areas to be polluted up to that
point or are we determined to hold the environmental
quality of relatively unimpacted areas? The former
implies that the maximum permissible stress repre-
sents a level to which we will allow all levels to rise.
The cost of nondegradation must be very carefully
weighed against the social benefits that accrue.
All of the issues raised in describing the process
of obtaining a maximum permissible environmental
stress for health effects are applicable to the case of
ecological effects as well: threshhold versus non-
threshold pollutants, benefits of given levels of maxi-
mum permissible stress, defense of especially sensi-
tive subpopulations, simple pollutants, multiple
stresses, both synergistic and nonsynergistic, and
multiple pathways. In general, however, the eco-
logical effects require an even greater consideration
of the social benefits that accrue to a given level of
permitted stress or control. Obviously, we are deal-
ing with a larger set of intangibles as opposed to the
relatively more easily defined area of health protec-
tion. The result is that a society which can under-
stand and accept costs attached to protecting human
health has a great deal of trouble justifying large costs
to protect ecological systems. The interrelated char-
acter of our environmental system is little under-
stood outside of the scientific disciplines. The idea
that damaged species in an ecological system can
produce irreversible effects in the entire system is
difficult to grasp and hence, demands an overt at-
tempt to trace the relationship through to a clearly
denned societal cost and benefit. The one area where
this is possibly not true in the consideration of eco-
logical effects is that involving impacts on the food
chain that can more easily be visualized by the lay-
man as having a definable effect on man. These ef-
fects, of course, are even more difficult to deal with
from a technical sense and involve an in-depth con-
sideration of the complex multiple pathways, multiple
stressor interactions. However, the ability to demon-
strate that man is destroying his food web has a pro-
found effect that leads to a recognition in society of
the benefits that accrue and a willingness to accept
the high "cost" of prevention.
To this point, we have been discussing the pol-
lutants, the pathways, and the maximum stress allow-
able through these pathways so as to protect various
species. The information obtained must provide a
sense of the levels of control necessary in these path-
ways to provide the required protection, and a sense
of the population at risk as a function of the applied
controls. In essence, this represents the first step in a
decision making process leading to effective manage-
ment of the environment. The decision maker, how-
ever, must have the ability to address the control
options available to him and the cost, both social and
monetary, that apply to the various levels.
There are two basic control options available—
technological, and process or operations manage-
ment. In the technological case, we are referring to
specific waste stream controls that can be applied to
specific sources. Typically, these are stack gas clean-
ers, biological treatment plants, etc., hardware solu-
tions for clearly defined waste streams. The process
or operational option refers to changes in the way the
pollutant is produced so as to limit the effluent. In
the case of a specific plant, this could involve a
change in an internal chemical process (e.g., substitu-
tion for the mercury cell in a chlor-alkali plant). In
the case of a widespread source, such as transporta-
tion, it could involve a management process that re-
sults in a change in use pattern. Sources considered
can be either stationary or mobile, located at a point
as in the case of a power plant or widespread as hi
the case of agriculture. Pollutants can be those that
are emitted in final deleterious form (e.g., radioactive
gases, carbon monoxide, etc.) or take form after
emission (e.g., photochemical smog, sulfates, etc.).
The acceptability of a given control option is de-
pendent on both its dollar and social cost. A typical
example of a dollar and social cost is the combined
cost of restricting the use of automobiles and the
social acceptability or unacceptability of such con-
straints. These must 'depend on the level of stress
limitation desired. Yet, this is not an explicit factor
in the sense that to some undetermined degree the
maximum permitted stress is dependent on the cost
that society must pay. The decision maker in man-
aging the environment requires, therefore, a deter-
mination of the economic costs of various options as
a function of desired control level as well as some
estimate of the variation in the institutional impacts
that result. This then provides him with the ability
to take the second step in the decision making proc-
ess, the information that will permit option trade-offs
to be examined.
The final step in this process of developing an en-
vironmental management plan involves the develop-
ment of an operational strategy within which the
various trade-offs can be examined. In essence, these
sets of strategies represent control plans involving the
-------
effluent quality and its variation with time for all
sources distributed over the entire region. The
strategy, therefore, represents a plan for limiting the
stress to the maximum permissible by limiting the
maximum permissible loading, and distributing this
over the region by assigned source effluent limits.
Obviously, adequate knowledge of control costs,
stress impact, and source distribution in time and
space will permit the construction of management
strategies that can be optimized either singly or in
combination around various conditions (i.e., cost,
maximum protection, etc.). The independent variable
in all cases is the date by which a given level of pro-
tection will be achieved.
Strategies for managing the environment can be
subdivided into two basic classes, the first being the
case described previously of preventing an area that
is currently clean from ever reaching the maximum
allowed stress and the second being the implementa-
tion of strategies that force a currently polluted area
to recover a quality environment at least equal to
the maximum permissible stress.
In the first case, the point at issue concerns the
price that society is willing to pay to keep a clean
area clean. Within this issue is the requirement to
define what we mean by clean. We presumably have
already established a level of maximum permissible
stress. Further, we have already indicated that level
in all probability is not an absolute protection for all
affected species. Does this imply that nondegradation
means zero increase above present levels? Or con-
versely, does it imply an allowable increase that will
not permit the stress to exceed a stated fraction of
that which is the maximum permissible? The first
case will demand a cost which will probably make it
impossible for any additional development to occur.
The second might permit some limited development,
but at a stated cost that may or may not be unac-
ceptably too high. It is clear that with the approval
of society, the description of the acceptable nonde-
gradation level can be used as a mechanism for pro-
tecting designated areas from development of a
certain character, and restrict development to speci-
fied areas. It becomes, in effect a deliberate tool for
land use management.
It is certainly attractive to consider the ultimate
protection of pristine areas and the demand for an
absolute nondegradation. It must be pointed out
however, that too restrictive a demand can be coun-
terproductive to the environmental needs of society.
Imagine, for example, a nondegradation level that is
so restrictive as to permit development to occur only
in the already polluted and degraded areas. It is clear
that such an approach would condemn the highly
developed, most populated areas to ever increasing
development and stress. Once again, it is necessary
to carefully balance the costs and benefits to society
of the management actions to be taken in the sense
that any management tool is a two-edged sword that
badly used can be more damaging than inaction.
Used intelligently, the nondegradation issue can help
control and alleviate the environmental stress in the
more polluted areas, can protect the pristine areas,
and can provide for an orderly growth in the desig-
nated regions.
In the second case, the case of environmental re-
covery, we are attempting to work with an area that
is normally heavily populated, heavily developed, and
most probably of strong economic importance. The
problem the environmental manager faces is how to
design a program of distributed control for the region
that takes into account the temporal and spatial dis-
tribution of sources and further, that this program of
control provides the region with an environmental
quality that is within the maximum permissible stress
without destroying its socio-economic viability. In
addition, a decision must be made as to the desired
date of achievement of this level of control. The con-
trol plan evolved and the attainment schedule chosen
are obviously related when considered in the context
of cost-benefit balancing. This is true both in the sense
of the stipulated maximum permitted stress as indi-
cated previously, and in the distribution of control
level and its designated time of application. The abil-
ity to do so, however, implies a capability to assess
the social risks of a given stress level and its period
of maintenance, and to accurately assign a stress
level contribution to each source. Under the influence
of total cost as related to benefits, one might stipulate
a decreasing maximum permissible stress with in-
creasing time, paced to match into the achievement
ultimately of a more effective control capability with
a smaller societal cost. Obviously, this must also be
paced with the cumulative social disbenefits that may
increase with the delay in the achievement of the ulti-
mate maximum permissible stress. What we are
searching for then is the ability to minimize the sum
of direct abatement costs (whether actual economic
costs for control equipment or social costs due to
operational changes, such as reduced automobile
use), and the social costs of delaying the achieve-
ment of the most desirable maximum permissible
stress.
One important note of caution must be sounded in
the above approach. Unless considered directly, and
included in the management plan, one may find that
this approach contains no incentive for ever acquir-
ing the techniques for achieving the most desired
maximum permissible stress. The approach utilized
to date, particularly in the United States, has been
directed primarily towards setting maximum permis-
sible stress on the more restrictive grounds of en-
vironmental effects as determined, and the achieve-
ment date as a legislated mandate. What cost-benefit
analysis is done, is normally done after the fact of
setting the values of these parameters. In defense of
-------
this system: First, it does not require knowledge, as
described above, that is not currently available to
the extent needed; hence, it does not delay the regula-
tory process. Second, it does force the development of
techniques to achieve the desired levels. The unde-
sirable feature of this system is that it may be
counterproductive in its attempt to protect society
and at the very least, it is an abrasive system that
puts a premium and a relatively low penalty on non-
compliance by individual sources.
The inadequacy of the current regulatory system
is rather widely recognized, and gains in recognition
with each further thrust towards comprehensive en-
vironmental management. This is not to say that a
great deal has not been accomplished with the current
approach. This is particularly true in the sense of
bringing about a problem awareness, underscoring
the government's resolve to achieve an improvement
in the environmental quality, and setting the research
program on the proper course. As the problems be-
come more complex, however, as the societal prob-
lems become more massive and evident, the realiza-
tion of the need for a more analytical comprehensive
approach grows. As complexities grow, so grow the
questions that must be addressed. Should there be
ambient standards, effluent standards, or both?
Should health effects carry the same weight as other
ecological effects, can and should cost-benefit analy-
sis be used to set the standards, the date of achieve-
ment, or both? Should standards be expressed as a
maximum allowable concentration over a specified
period, and if so with what frequency? Are multiple
pollutant standards required, and how will goal
achievement be determined?
Along with the realization of the inadequacy of the
simplistic direct enforcement approach to deal with
complex environmental management problems goes
an awareness of the accompanying inadequacy of the
needed analytical tools. In order to specify a control
plan and an achievement schedule which truly mini-
mizes socio-economic costs (e.g., minimizes the
societal cost-benefit ratio), it must be possible to
relate that plan to species stress and, therefore, to
ambient quality. The preferred method of doing this
would be through the use of stochastic physical-bio-
chemical models for both water and air which ade-
quately represent the probability aspects of local
through regional transport and bio-chemical pro-
cesses. Although much attention has been given and
is being given to the development of such models
and the required input data, neither is adequate at
this time. As a result, interim methodologies and ap-
proaches are utilized, which fall far short of being
able to adequately address the questions raised. It is
certainly clear that a substantial effort is needed in
the near term to address the highly impacted areas.
A major concern is felt, however, lest actions taken
now utilizing gross data and gross techniques pre-
clude more correct solutions being implementable at
a later date.
A final point of concern to the environmental
manager is involved in the achievement of a regional
long-term plan of environmental control that meshes
well with the pressing short-term needs. Short-term
programs can be expected to have less stringent goals
than the long-term ultimates, and rely more on im-
mediately available controls than on longer range
technological and operational fixes. Environmental
pollution, however, is a symptom of the intensity
and scale with which human activities are pursued in
the region. Each region has a natural "carrrying ca-
pacity" for these activities, which can be altered
significantly only over a long period of time. For ex-
ample, the carrying capacity can be altered by tech-
nology (e.g., mass-transit to off-set the use of auto-
mobiles; or pollution control technology) by urban
design and land-use planning, and by resource
conservation measures, such as improved building
insulation and design. Long-terd environmental pro-
grams in effect are increasing this regional carrying
capacity by altering the basic nature of the region
and its socio-economic activities. On the other hand,
much of a short-term program (e.g., transportation
controls) is simply reducing the intensity or scale of
human activities, within a fixed carrying capacity.
Ultimately, every region must live within, and plan
for, a finite carrying capacity. In particular, this
means that social and economic goals must be
matched to the natural characteristics of the region
involved and the interaction and impact of one region
on another.
In summary, then, the ability to effectively man-
age the environment comprehensively involves many
factors. First it must be managed over a region large
enough to permit the effective control of the opera-
tive pollutants. Second, it is necessary to determine
the maximum permissible environmental stress to
protect the effective health of the population and the
underlying ecological system. In this connection,
questions of social risks should be addressed, and the
concept of total benefit considered. Third, the cost of
control plans must be adequately known. Fourth, a
maximum loading plan must be established which
permits a consideration of source distribution and
intensity, as well as the social costs of control. Fifth,
a long-term plan compatible with short-term needs
must be evolved which, if necessary, can allow the
region to reach the limits of its natural carrying
capacity. These parameters, adequately treated, will
permit a management plan to be developed which
addresses the ability to comprehensively protect
against the environmental impact balanced with the
necessity for society to remain viable.
This symposium obviously will not provide an-
swers to all of the questions raised, but will allow
both countries to share the knowledge we have both
acquired and explore what has yet to be done—it is
a beginning.
-------
Systems Analysis and Simulated Mathematical Modeling as a
Methodological Basis for the Standardization of Anthropogenic
Pollution of the Environment: A Regional Approach
Yu. A. Anokhin, Yu. A. Izrael
1. One of the serious negative consequences of
man's activities (primarily economic), as is known,
is the pollution of the environment, the scales of
which grow rapidly, but which, nevertheless, ought
not be considered a necessary satellite of develop-
ment [1]. Anthropogenic pollution of the environ-
ment occurred in the course of historical develop-
ment, and is a unique (albeit far from the best) type
of "exploitation" of natural resources—the air reser-
voir, natural waters, and soils.
Scientists and specialists, while acknowledging the
social origin of the problem of environmental pollu-
tion ought not, however, limit themselves simply to
calling for new ethical standards to apply to man's
relationships with nature. Our task is to develop
effective, scientifically valid, measures for dealing
with pollution, and preventing the occurrence of pos-
sible crisis phenomena. By "pollution" let us under-
stand that we mean the input into the environment
of toxic substances that can be harmful to people's
health, as well as of any other substances that could
have undesirable effects on biogeocenosis and on
materials, and this includes excess heat, noise, and
radiation in the environment. In truth, impurities are
substances (and radiation fields) found where they
are not supposed to exist (this is a modified version
of an old chemical rule). It would only be natural to
include here as well anthropogenic changes in land-
scapes deemed unfavorable, not only from an eco-
nomic, but from an esthetic point of view as well.
The possible effects mentioned here are of great
social, cultural, and economic importance, and all
must be taken into consideration when setting pollu-
tion standards, either in terms of the conditions the
role of the "leading" object must be in so far as man's
health is concerned, the effect it will have on pre-
determined natural systems (biogeocenosis), on
certain forms of flaura and fauna, or even on certain
materials.
Moreover, and this is of utmost importance, an
understanding such as this of the pollution problem
will make it possible to deal with it as one of the
facets of the wrong, irrational use of natural re-
sources, so that the approach to solving the problem
follows as the antithesis: find ways to make intelli-
gent, "optimum," use of natural resources.
2. A characteristic property of the "optimization"
problem under consideration is its internal contradic-
tion, its "multidimensionality," so to speak, some-
thing a goal-oriented function takes on when an at-
tempt is made to assign the corresponding extremal
problem.
The overall goal, for example, is that of rationaliz-
ing the interaction between industry and the environ-
ment, and comprises the following sub-goals, the
latter in many respects contradictory: maximization
of physical production and of the productivity of
labor, and minimization of waste products and re-
jected products. While maximum physical produc-
tion and consumption could be considered the opti-
mum so far as satisfying human demands is con-
cerned, this approach is not optimum, however, from
the standpoint of intelligent use of nature (nor, inci-
dentally, from the standpoint of widely understood
social requirements). Natural resources are suscept-
ible to multiple use, so the unresolved problem is
"how to make a quantitative comparison of the
different social requirements because of the difficulty
in comparing the need for clean air, and the esthetic
enjoyment of the beauty of nature, with the needs
for electric power, metal, and fish, and because of
the difficulty, in turn, of comparing the latter each
with the other" [2].
It is obvious, therefore, that any attempt to resolve
the problem of how to use nature more efficiently
"in one fell swoop," by resort to optimization proced-
ures of the mathematical programming type, let us
say, is an illusion. An informal understanding of the
concepts of "optimum" and "optimization" must be
reached by the use of intelligent, "meta-model" con-
siderations, and this approach has a decisive effect
on methodology selection.
3. The development of the correct methodology
for standardization of anthropogenic pollution is the
-------
most important condition, and means, for dealing
successfully with pollution. Systems analysis, and
simulated mathematical modeling, as applicable to
relatively closed territorial-economic systems, can, in
our view, become the instrumentality of the method-
ology for standardization on a regional scale.
The very fact that environmental pollution is most
intensive within the limits of existing territorial-eco-
nomic regions is why regional introduction of stand-
ardization of anthropogenic pollution in accordance
with accepted criteria for rational use of nature is so
important. It seems clear that the only policy that
makes any sense at this time when developing these
criteria is that of a differentiated approach to our
country's various natural systems (to the biogeo-
cenoses, for example). Some natural communities
should be completely cut off from any interference by
man, and should be preserved in their natural form.
There are others, however, that should be adapted
to human needs more decisively and intensively than
is the case at the present time. It is possible that
there are intermediate types of communities, and
here we can permit ourselves to experiment, bearing
in mind at all times, however, the possibility of un-
foreseen, negative, consequences. A policy such as
this is optimum in the sense that it guarantees "maxi-
mum freedom of choice for our descendants" [3].
4. We shall, in this report, deal first (paras. 5 and
6) with some of the general properties of the systems
in the class under consideration, regardless of con-
crete realization. We then (paras. 7-9) will take up
in some detail certain methodological questions con-
cerned with the standardization of pollution on a
regional scale, and will illustrate these considerations
by the example of the Lake Baykal region (paras.
10-23).
5. As is known, the general properties of such
systems, because they are "organized complexity"
systems [4], are given theoretical consideration in
general systems theory. The most important of these
properties are openness (interaction with the en-
vironment), the hierarchical nature of the structure,
temporal changeability, and the attribute of control
inherent in these systems [5-8]. Of significance to
us is the fact that these general properties have made
it necessary to develop a new methodology for study-
ing complex systems, the basis of which is systems
analysis and mathematical modeling, with its general
directionality that from analysis (understood to mean
the breaking down of the whole into all of its smaller
parts and studying each of the latter separately) to
synthesis, and the elements of inadequacy and lan-
guage polymorphism inherent in this method [8, 9].
6. This approach is neither a narrow, special dis-
cipline, nor some sort of mystery. It is, rather, a
multidiscipline, quantitative method for studying
complex systems when knowledge of them is incom-
plete, the purpose of which is to predict the behavior
of the system under changing conditions, and, in a
certain sense, to gain optimum control of the system.
System behavior is determined by its structure, and
by the interrelationships of its elements. The follow-
ing properties of systems of this class are intrinsic
[10]:
(1) they have a great many variables (of the
order of several hundred, or more);
(2) the variables often have feedbacks;
(3) the system often has time lags, nonlinearities,
and even discontinuities.
Equations describing a system such as this usually
are without analytical solution, and this is why it
becomes necessary to turn to experimental investiga-
tions with mathematical models and resort to an
electronic computer. As is known, the Forrester
school has gathered a great deal of experience in this
field [11-13]. Forrester [11] has made a clear
analogy between controlling a complex system and a
unique method of driving an automobile, one in
which the driver is blindfolded and follows instruc-
tion given him by an instructor looking out the rear
window (that is, on the basis of information on the
road already traveled). Added to this is the fact that
the instructions, as given, contain time lags and
errors. Given conditions such as these, it is difficult
to overestimate the importance of an approach that
is based on systems analysis and simulated mathe-
matical modeling, primarily because this approach
makes it possible to predict possible system states.
This, in turn, serves to make system control more
effective, and behavior of the controlled system more
regular. This approach can, in its general outlines,
be said to comprise the following main stages [ll[
14]:
(1) the concrete question in need of an answer is
formulated. This is basic. Successful formulation of
the question is possible only if the situation, the
problem, and the purpose of the research, are clearly
understood;
(2) a verbal (verbal-causal) model, that is, the
fundamental internal and external relationships, is
formulated;
(3) the verbal model is graphed, and then is re-
flected as a mathematical model;
(4) system behavior is computer simulated, and
the results are compared with available data on
similar real processes;
(5) iterativeness of the process is determined.
Return to stage (2) if the representation in the model
of real processes is unsatisfactory.
We can obtain the required degree of adequacy of
the model to the real system by repeatedly "fitting"
the model as we perfect our knowledge to the sys-
tem's arrangement. And the statement of the problem
[(1), above] can be modified as a result of such
computer "experimentation."
-------
7. Let us now turn to the methodological ques-
tions of pollution standardization. There can be no
question of the fact that any program for dealing
with environmental pollution must include a complex
of sanitary-hygienic, biological, recreational-esthetic
requirements, etc., for environmental quality. Nor is
there any question of the fact that realization of a
program such as this has to depend on social-eco-
nomic, cultural, and other factors, as well as on the
"level of lack of quality" of the environment that in
fact exists. The present need for an optimum solution
of complicated problems concerned with pollution
standardization therefore should be obvious.
Let us note that existing economic and mathe-
matical methods of optimizing the use of production
resources cannot be applied directly to solving prob-
lems concerned with optimizing the use of natural
resources, and this is particularly true when it comes
to optimum pollution standardization, because these
methods fail to take into account the qualitative
specifics of natural resources, which are production
factors, as well as the "bodies and forces of nature"
[15], "the behavior" of which is determined by
natural laws that we know nothing about all too
often, unfortunately.
8. Nevertheless, the economic approach, despite
all its very serious and well-known limitations, must
and can be a powerful lever to use in solving the
problems of how to make optimum use of nature.
This approach is based on the concept of the repro-
duction of natural resources as one of the main
spheres of physical production [16]. At the same
time, the central economic problem is one of a
correct economic assessment of natural resources.
The concept now used by Soviet economists to make
this assessment is that of "closing costs," so-called
[17, 18].
9. One of the general laws of the behavior of en-
vironmental pollutants is that they tend to disperse.
Moreover, substances that are reciprocally con-
vertible, particularly when the resultant substance is
more toxic than the original one, are of the utmost
importance. It is only natural, therefore, that the
general principle for the standardization and moni-
toring of pollutants should be that of localization;
that is, the content of pollutants in the environment,
and particularly their source, must be standardized
and monitored. What follows from this almost self-
evident assertion is the practically important con-
clusion that the presently universally accepted prac-
tice of simply standardizing and monitoring the
threshold limit value (TLV) is not effective enough.
We must clearly realize that there are two closely
interrelated aspects of the pollution standardization
and monitoring problem. First of all, there is the
need to establish the corresponding criteria, or stand-
ards, for environmental quality (TLV types, or values
similar to them). Absolute conformity of environ-
mental quality with the established standard should
be the law, and should not be violated. Realistically,
however, the attainability of this state of affairs very
definitely depends on standardization and monitoring
the very sources of pollutants. This is why the stand-
ardization and monitoring of environmental quality
on the one hand, and of the sources of pollutants on
the other, are mutually complementary.
10. We shall, in this report, for purposes of greater
concretization, limit ourselves to the consideration of
just one region, that of the Lake Baykal basin
(Figure 1). The region is one with quite high levels
of economic activity, yet has a special feature, the
unique cultural and historical value of the lake in
addition to its high economic value. The need to
foster the future economic development of this re-
gion and to utilize its riches, on the one hand, and the
unquestioned need to preserve the lake, together with
its fauna and flora in their original form on the other,
therefore does much to explain the important role
this region has as a unique model of possible
judicious use of natural resources.
There are definite methodological difficulties in
depicting the region as a relatively closed economic-
ecological system. Even the statement of the problem
of depicting the region is contradictory because, as
has been pointed out in the foregoing, an intrinsic
property of this type of system is openness.
On the other hand, as we understand it, the
regional determinant is a known homogeneity of
certain criteria within the limits of the region, such
as some combination of physical-geographical, eco-
nomic, administrative, social, and like properties, so
that the concept of "boundary of a region" takes on
a definite meaning. In our case the problem lends
itself to a relatively simple solution. The Lake Baykal
basin is a well-defined territory with accurately estab-
lished boundary, hydrologically speaking. The limits
of this territory contain clearly distinguishable land-
scapes, composed in turn of clearly distinguishable
systems of biogeocenoses. And although the land-
scape and the biogeocenosis are essentially open
systems from the energy flow standpoint (sunlight,
heat), as well as from the standpoint of their inter-
relationships with adjacent regions (for, in the final
analysis, this is what the evolution of the biosphere
is all about), it is important to remember the con-
ception of the unity of biogeocenoses as elementary
"atoms" of the biosphere developed by V. I. Ver-
nadskiy and V. N. Sukachev. The fundamental bio-
logical flows of matter and energy are locked within
the limits of the biogeocenosis, so the biogeocenosis
boundary determinant is the condition that the mag-
nitudes of the flows across the boundary be substan-
tially smaller than the internal flows. Thus it is that
the colossal Lake Baykal region "ecosystem", which
is a system of relatively closed biogeocenoses, is it-
self relatively closed.
11. The economic subsystem of the region under
review is extraordinarily "open." It cannot be con-
sidered closed, even as an approximation, because it
-------
STATE BOUNDARY BETWEEN
THE USSR AND THE IV
BOUNDARY OF THE
LAKE BAYKAL BASIN
USSR
/
MPR C
}
r
j
J
\
FIGURE 1. THE LAKE BAYKAL BASIN AND SELECTED TERRITORIAL SUBSYSTEMS
is obvious the level of economic activity, as well as
the region's demographic features (population and
its structure), depend most heavily on exogenous
factors, while the flows of materials and goods, let
us say, across the regional boundary are of the same
order of magnitude as are the internal flows.
However, disregarding analysis of the relationships
between the region selected and other regions, the
exogenous factors mentioned can be presented in the
form of deterministic "forces" controlling the de-
velopment of the economics, and thereby formally
close the system.
12. So the region is a relatively closed system
consisting of interacting economic and ecological
subsystems. Figure 2 lists the main groups of factors
describing the behavior of these subsystems (paras.
14-17 will contain brief comments on these sub-
systems).
We currently are in the process of developing a
simulated mathematical model of the functioning of
the Lake Baykal region as an economic-ecological
8
-------
Economic subsystem
1. Exogenous factors
2. Population
3. Industry
4. Agriculture
5. Lumbering
6. Exploitation of the lake
7. Other factors
Ecological subsystem
1. Exogenous factors
2. Pollutants field
3. Biogeocenoses models
(stability, resistability)
4. Reaction of biogeocenoses
5. Ecological damage
6. Other factors
FIGURE 2. TWO OF THE REGION'S SUBSYSTEMS.
system so we can predict the effect economic activity
will have on the region's ecological system. We firmly
believe that this model, initially of necessity very
rough, will be a tool, as well as a powerful stimulus,
for future research, because it will help show up the
"white spots" (gaps) in our understanding of the
functioning of this very complicated system, and thus
serve to formulate requirements for the type, quan-
tity, and quality of the needed data, the obtaining of
which must be organized. Participation in this work
of specialists in the most varied of disciplines who
seldom meet as a rule is, in our opinion, extremely
valuable. This would include chemists, toxicologists,
hydrometeorologists, medical personnel, hygienists,
biogeocenologists, industrial workers, and economic
geographers.
The development of the model, and full-scale re-
search, forming two continuous, interrelated flows,
will become the scientific base for valid decisions on
how to standardize anthropogenic pollution on a
regional scale.
We cannot provide any details on the work we are
doing since it has only just begun, but would, never-
theless, like to share with you some fundamental
considerations, and to discuss some of the concrete
approaches we have planned for solving the problem.
13. A source of pollution in the regional approach
to pollution standardization can be an individual
chimney, a plant, a city, or even a whole economic
region, depending on how detailed one wants to get,
that is to say, what "degree of resolution" of the
problem is desired. The sources of pollution in the
production-branch approach are individual types of
production, or even a branch as a whole. The im-
mediate task confronting us is investigatory in nature,
and involves the use of the simulated mathematical
model of how the region functions to make a com-
parison of the different anthropogenic sources of pol-
lution in terms of the degree to which they are harm-
ful to the environment. Our sources of "pollution"
will be:
(1) large cities;
(2) industrial plants;
(3) individual, nonlocalized types of economic
activity (animal husbandry, fishing, shipping, lumber-
ing, and the like).
A model such as this must be dynamic, naturally,
so the comparison will result in approaching a solu-
tion to a task of higher order, that of comparing
various alternative programs for the economic de-
velopment of the territorial economic system under
review, and making recommendations for the selec-
tion of what would, in some way, be an optimum
variant. It seems clear that the solution of the prob-
lem will involve the need to combine the regional
and production-branch approaches if the solution is
to be responsible and concrete. The general arrange-
ment of the solution of this problem can be repre-
sented in the form shown below (Figure 3).
14. Now let us make a few brief comments on the
economic and ecological subsystems in the Lake
Baykal region.
The Lake Baykal region's economic subsystem is
growing rapidly and is extraordinarily dynamic. We
have selected two sectors in the subsystem, the demo-
graphy and the national economic, for discussion.
The group of exogenic factors describes all those
external planning and economic "forces," inclusion
of which makes it possible to close the system (para.
11).
-------
Sources of
pollution and
environmental
effects
L
Transport of
pollutants
(atmosphere,
soil, natural
waters)
Reactions of
ecosystems
to pollution
Evaluation in
terms of
gain/loss
FIGURE 3. CLOSED CHAIN PROBLEM SOLUTION
The demographic sector of the economic sub-
system can be described in broad outline as follows.
Regional development and population are very ir-
regular, and overall are well down on the scale. Mean
population density is well below that for the Soviet
Union. Immigration from other regions in the coun-
try, as well as natural growth, are increasing regional
population. The distribution and dynamics of the
in ban and rural populations are not the same
throughout the region, and in fact are determined by
the special features of economic development.
The national economic sector describes the
region's own economy, the basis of which is well-
developed industry (with its timber felling and saw-
mill branches, as well as branches of heavy and light
industry, and power engineering, having an im-
portant place in the structure), and diversified agri-
culture (different branches of livestock breeding,
animal husbandry, and plant production).
Hunting and fishing have an important place in the
region's economy.
The group of factors combined under the rubric
"other factors" includes tourism, and other forms of
recreation.
15. So, to deal with the problem we are interested
in, that of comparing the different anthropogenic
sources of pollution in terms of the damage they do
to the environment (para. 13), let us look at the
economic subsystem as a whole as a source of pol-
lutants and at its effects on the ecological subsystem,
and, in turn, look at the ecological subsystem in
terms of its effects on the economic subsystem
(Figure 2).
The ecological subsystem is a hierarchy of rela-
tively closed land and water ecosystems (para. 10).
As we know, Lake Baykal is one of the largest
fresh-water lakes in the world. Its exceptional age,
vast depths, uniqueness of its fauna and flora, and
the exceptional purity of its water, all make Lake
Baykal stand out sharply from other lakes, and give
it its special, unique, character [19].
This very uniqueness, as well as the fact that a
huge number (some 400) of rivers empty into the
lake (and bring with their waters the pollutants they
contain), makes Lake Baykal's biome one of the
most critical links within the region. (We should
point out that a great many of the natural and cul-
tural values in any region obviously are associated in
some way with the natural water systems [3], so that
in this respect the Baykal region is a typical ex-
ample.)
The Selenga is the largest river in the basin. Its
drainage basin covers 60% of the total area occupied
by the lake's basin. The river is 1024 km in length,
with 409 km within the boundaries of the Soviet
Union. The upper reaches of the river are in the
Mongolian People's Republic.
Only one river, the Angara, flows out of the lake.
16. Irrational management of fishing and sealing
(particularly the taking of too many seals, and over-
fishing the lake without taking reproduction into
consideration), and insufficient attention to fish pro-
pagation, to say nothing of polluting the rivers
with the wastes from the timber industry and with
industrial runoffs that cause breeding grounds to
deteriorate, can do definite damage to Lake Baykal's
biome, and in particular to the fish stocks and to the
Baykal seal (ringed seal) population [20].
Change in the lake's chemistry, the result of
change in the hydrological and chemical conditions
of the lake's most important tributaries, the Selenga
and Barguzin rivers, "caused by the development of
agriculture and livestock breeding, and by intensifi-
cation of the feeling of trees, and more recently by
the dumping into the river of industrial and domestic
sewage," also can harm Lake Baykal's biome [21].
"Change in the composition of organic matter of
allochthonous origin attributable to human economic
activity such as to increase compounds resistant to
oxidation, as well as various organic toxicants that
can be the reason for the shift in the equilibrium
state of the balance of organic matter in the lake so
that bacteria accumulate because their decay is sup-
pressed," is cause for alarm [22].
Efforts have been made in recent years to restock
Lake Baykal with fish. A number of farms have been
built for the artificial breeding of Baykal omul (C.
autumnalis migratorius), and steps have been taken
to improve natural spawning grounds (timber rafting
in free form was prohibited along the Selenga and
its tributaries in 1972, for example). Studies of the
links between the different ecological and morpho-
10
-------
logical omul groups and the heat and water condi-
tions in the lake, and the unacceptability of estab-
lishing strict limits of any sort on fishing for a long
period of time, thus become highly important to near-
term planning for the reorganization of omul fishing
in Lake Baykal, to transferring it along the path of
spawning migrations [23, 24].
So far as the ringed seal is concerned (the popula-
tion of which has been restored and now numbers
some 68,000 [25]), it is worth noting that the
world has very few lakes with ecosystems similar to
the marine ecosystem, those in which highly-
organized mammals complete the food pyramid.
"The biocenotic and economic importance of the
ring seal is that it transforms the huge quantities of
small, pelagic, golomanka [Comephorus—Baykal
fishes related to sculpins] and gobies not now avail-
able to man (because they are scattered throughout
the body of water and do not concentrate in groups
large enough to make their fishing economically ad-
vantageous), thereby converting them into a product
available in the form of edible and commercial fats,
meat, and, in particular, fur skins" [25].
17. Another example of an important ecological
link in the region under consideration is the unfavor-
able change that has taken place in landscapes as a
result of economic activity. The problem we inherited
from the past, that of moving (eolian) sands and
erosion (deflation) of soils in some regions of the
Transbaykae is particularly acute, leading as it does
to the blocking of populated areas, rivers, roads,
to the filling up of lakes and irrigation canals, and
covering fields, meadows, and forests. The problem
arose many decades ago, and is the result of the de-
struction of vegetation, brought about by continuous
concentrated logging, frequent fires, overgrazing of
cattle, as well as ploughing up the light soils of the
broad steppes (and thus exposing earlier anchored
sands).
Movement of the sands, and the deflation proc-
esses, have intensified sharply in recent decades, and
this has increased the area of moving sands in the
south of the Buryat ASSR several times over. 'This
outbreak of deflation is merely the reflection of the
increasing changes that are taking place in the land-
scape in the area because of anthropogenic factors"
[26].
18. The examples of critical ecological links cited
confirm the obvious fact (para. 1) that dealing with
industrial and domestic types of pollution is closely
interwoven with the broader questions of the ra-
tional use of natural resources (land and forest,
in particular). The task of comparing different an-
thropogenic sources of pollution in terms of their
"harmfulness," therefore, can be performed only
within the framework of a model broad enough to
include the questions mentioned. Moreover, the
model should contain the whole approach to eco-
systems, something that is possible only when prob-
lems of a qualitative understanding and of quanti-
tative modeling of the behavior of ecosystems are
solved in terms of their common characteristics (sta-
bility, capacity to resist near the stability limits, and
the like). Then too, there may be individual elements
of the ecosystems, such as the omul population, the
ringed seal, and others, that could be of interest to
us.
The development of this model is a complicated
business, and that is why the correct structural
organization of the model is so important. This is
particularly true of the modular principle of its con-
struction, when the model consists of sectors and
subsectors that are relatively independent, and only
loosely bound to each other, so that individual parts
of the model can be modified without the need for
substantial alterations to the rest of the parts.
19. The remainder of our discussion will be purely
illustrative in nature because its purpose is to ex-
plain the approach, without in any way pretending
that the description of that approach is complete, or
accurate.
Anthropogenic effects on the Lake Baykal eco-
system reduce to the following types (or combina-
tions of them):
(a) the addition of organic and mineral substances
peculiar to the lake (particularly in quantities ex-
ceeding their natural fluctuations);
(b) the addition of toxic substances (not peculiar
to the lake);
(c) the removal of organic substances (fishing,
for example);
(d) change in the makeup of the ecosystem.
Types (a) and (b) can be combined under the ru-
bric of "pollution," and in the main are the side
effects of economic activity. Type (c) is one of the
forms of economic activity. Type (d) often is the
consequence of types (a), (b), and (c), but can be
independently important (when new forms are intro-
duced into the ecosystem, for example). Figure 4
shows the paths over which pollution can be carried
into Lake Baykal.
20. Elaboration of the scheme shown in Figure
4 requires that individual territorial subsystems
within the limits of the region under consideration
be isolated. In our case, the criterion for isolating
subsystems is known homogeneity of the subsystems
by degree of pollution and effect on Lake BaykaTs
ecosystem.
We have isolated 22 territorial subsystems, in-
cluding Lake Baykal, within the limits of the region
on the basis of the available description of the
region as an economic-ecological system (Figure 1
shows the numbers of these territorial subsystems).
Lack of space precludes describing these terri-
torial subsystems in detail. We shall simply point
11
-------
Sources of
Pollution
Air Basin
II
Ground Surface
1
Lake
Baykal
Rivers
J
FIGURE 4. PATHS OVER WHICH POLLUTION CAN BE CARRIED INTO
LAKE BAYKAL.
out the fact that these subsystems include regions
that are the largest industrial centers, various hydro-
logical sections (sections of the Selenga River), and
various regions around Lake Baykal, and that they
are in addition to areas external to the region, and
to Lake Baykal itself.
21. Figure 5 shows the interaction between the
isolated subsystems as they affect Lake Baykal's eco-
system. Let us explain the subsystems and their
interaction (the numbers, are those shown in Figure
1):
0. Lake Baykal, the object of the anthropogenic
effects in the model under consideration.
1. external areas. The subsystem reflects the over-
all geophysical transport of pollutants, and transport
from adjacent regions.
Fallout directly into the lake, fallout on the water-
sheds of the rivers in the Lake Baykal basin and
subsequent transport into the Lake, and pollutants
coming from the upper reaches of the Selenga River
(the Mongolian People's Republic), all must be
taken into consideration;
2-9. isolated sections of the Selenga River;
10-12. these subsystems describe the effect on
Lake Baykal's ecosystems of human settlements and
industry in the region. The outlines of regions 10,
11, and 12 (the Baykal Cellulose Plant, the Selenga
Cellulose-Cardboard Combine, and the city of Ulan-
Ude, respectively) have been drawn to correspond
to the transport of pollutants as depicted in Figure 4;
13-21. these are the land areas in the Lake
Baykal basin, and were isolated because of a hydro-
logical tie with some particular section of the Selenga
River, and other rivers in the basin, as well as be-
cause of the relative homogeneity of the type of land
use (arable land, haymaking, pasture land) and the
extent of forest utilization.
22. Let us, in addition to the territorial break-
down, isolate within the economic-ecological system
of the region, certain functional subsystems of in-
terest to us:
(1) human settlements, industry, transportation;
(2) resource utilization (the biome of Lake Bay-
kal, forest and land resources);
(3) recreation;
(4) hydrology, and quality of surface waters;
(5) ecological and economic aspects.
Let us now present the factors that describe quali-
tatively the interaction of these subsystems in the
form of a matrix (Table 1). The matrix element
with subscripts i, j (i is the row number, j is the
column number) describes the output factors of the
i'" subsystem with i = j, and these are simultane-
ously the input factors of the j'" subsystem (when
i = j, the element presents the concise content of
the corresponding subsystem).
As may be seen from Table 1, the presence of
diverse (including reverse) interrelationships be-
tween these subsystems makes the model extraordi-
narily dynamic (so the important role played by a
methodology that simulates modeling when investi-
gating similar systems becomes quite obvious, as we
pointed out in paras. 7, 13).
A model such as this should be "rotated" suc-
cessively for the individual territorial subsystems (in
accordance with Figure 5). The result is an assess-
ment of the effect the various anthropogenic sources
of pollution have on Lake Baykal's ecosystem.
23. In conclusion, let us point out that the degree
of spatial and/or functional disaggregation we have
adopted on a preliminary basis in the model can be
insufficient (or excessive, perhaps ?) for reaching
substantive conclusions. We are assuming, however,
that the modular principle we have adopted for
building the model is such that the necessary degree
of disaggregation, or aggregation, can be introduced
in the future without significant rearrangement of
the structure of the model.
12
-------
Table 1. Interrelationships Between Functional Subsystems
SUBSYSTEMS
1. Human settle-
ments, industry,
transportation.
2. Resource utiliza-
tion (biome of Lake
Baykal, forest and
land resources).
3. Recreation.
4. Hydrology, and
quality of surface
waters.
5. Ecological and
economic aspects.
1. Human settle-
ments, industry,
transportation.
2. Resource utili-
zation (biome of
L. Baykal, forest
and land re-
sources).
3. Recreation.
4. Hydrology and
quality of surface
waters.
Population, level of
industrial activity,
intensity of trans-
portation.
Possibility of unfa-
vorable consequen-
ces (blocking pop-
ulated areas and
roads, shoaling of
waterways).
Development o f
sanitation institu-
tions, road building.
Possibilities of de-
veloping the paper
and pulp industry.
Logging, plowing,
laying roads.
Use of fish re-
sources, sealing,
lumbering, land use
(hay-making, p a s -
ture, arable lands).
Man-hours on fish-
ing, hunting; tour-
ism, and the like.
Field of concentra-
tion of ingredients.
Development of
suburban rest
areas.
Perceived qual-
ity of the envi-
ronment. Sport
fishing.
Space-time dis-
tribution of rec-
reational activi-
ties.
Field of concen-
tration of ingre-
dients.
Sources of pollution
(domestic, indus-
trial run-offs, pollu-
tion by shipping).
Pollution (eutrophi-
cation), soil ero-
sion, clogging of
rivers, regulatory ef-
fect of forests.
Sources of pollution
(coliform bacteria).
Transport of water
masses and pollu-
tants, self-purifica-
tion.
Public health, fu-
ture development
of industry, and
transportation.
Fishing, shooting of
seals, scope of lum-
bering, pasture and
arable land areas.
Economic advan-
tage, capacity of
sanitation institu-
tions.
Development of the
paper and pulp in-
dustry, field of con-
centration of pollu-
tants.
5. Ecological and
economic.
Regulating effect of the ecological and economic aspects on the functional sub-
system.
Assessment of the
ecological and eco-
nomic aspects.
Finally, even the coarse degree of disaggregation
we have adopted will, we hope, provide the answer
to questions of practical interest.
BIBLIOGRAPHY
1. Ye. K. Fedorpv. Vzaimodeystviya obshchestva i prirody
[The Interaction Between Society and Nature]. Gidro-
metizdat., Leningrad, 1972.
2. A. A. Mints. "Geographic Approaches to an Economic
Assessment of Natural Resources." V kn: Ekonomiches-
kiye problemy optimizatsii prirodopol'zovaniya [In
book: Economic Problems of the Optimization of the
Use of Nature]. "Nauka" Publishing House, Moscow,
1973.
3. D. Ehrenfel'd. Priroda i, lyudi [Nature and People].
"Mir" Publishing House, Moscow, 1973. David W.
Ehrenfeld, Biological Conservation, New York, 1970.
4. R. Waterman. Teoriya sistem i biologiya [Systems The-
ory and Biology]. "Mir" Publishing House, Moscow,
1971.
5. O. Bertalanfi. Issledovaniya po obshchey teorii sistem
[Research in General Systems Theory]. "Mir" Publish-
ing House, Moscow, 1969.
6. F. Mesarovich. Issledovaniya po obshchey teorii sistem
[Research in General Systems Theory]. "Mir" Publish-
ing House, Moscow, 1969.
7. B. S. Fleyshman. Elementy teorii potentsial'noy effek-
tivnosti slozhnykh sistem [Elements of the Theory of
the Potential Efficiency of Complex Systems]. "Sovets-
koye Radio" Publishing House, Moscow, 1971.
8. V. V. Nalimov. Logicheskiye osnovaniya prikladnoy
matematiki [Logical Bases of Applied Mathematics],
Moscow State University Publishing House, Moscow,
1971.
9. I. M. Gel'fand. "Interaction in Biological Systems."
Priroda, No. 6, 1969.
10. Hamilton et al. Systems Simulation for Regional Anal-
ysis. MIT Press, 1969.
11. J. Forrester. Industrial Dynamics. Cambridge (Mass.),
MIT, New York-London, 1961.
12. J. Forrester. Urban Dynamics. Cambridge (Mass.),
MIT, 1969.
13. J. Forrester. World Dynamics. Cambridge (Mass.),
MIT, 1972.
14. A. A. Lyapunov. Problemy kibernetiki [Problems of
Cybernetics], 25. "Nauka" Publishing House, Moscow,
1972.
15. A. A. Mints. Mathematicheskiye metody v geografii
[Mathematical Methods in Geography]. Kazan', 1971.
16. M. P. Fedorenko. "Economic Problems of the Optimi-
zation of the Use of Nature." In the book of the same
name, "Nauka" Publishing House, Moscow, 1973.
17. K. G. Gofman. "Methodological Bases for an Economic
Assessment of Natural Resources." V kn. Ekonomiche-
skiye problemy optimizatsii prirodopol'zovaniya [In
book: Economic Problems of the Optimization of the
Use of Nature]. "Nauka" Publishing House, Moscow,
1973.
18. "Basic Assumptions in the Procedure for an Economic
Assessment of Natural Resources in Mass Planning and
Design Calculations" (project). V kn. Ekonomicheskiye
problemy optimizatsii prirodopol'zovaniya [In book:
Economic Problems of the Optimization of the Use of
Nature]. "Nauka" Publishing House, Moscow, 1973.
19. K. K. Votintsev. Gidrokhimiya ozera Baykal [The Hy-
drochemistry of Lake Baykal]. Irkutsk, 1963.
20. B, Buyantuyev. K narodokhozyaystvennym problemam
Baykala [National Economy Problems of Baykal]. Ulan-
Ude, 1960.
21. A. I. Metseryakova. "The Chemical Runoff of Rivers
in the Lake Baykal Basin." V kn. Krugovorot vesh-
chestva i energii v ozerakh i vodokhranilishchakh [Cir-
culation of Matter and Energy in Lakes and Reser-
voirs]. Listvenichnoye naBaykale, 1973.
22. K. K. Votintsev, G. I. Popovskaya. "The Circulation of
Organic Matter in Lake Baykal." V kn. Krugovorot
veshchestva i energii v ozerakh i vodokhranilishchakh
[In book: Circulation of Matter and Energy in Lakes
and Reservoirs]. Listvenichnoye na Baykale, 1973.
23. I. P. Shumilov. "A Procedure for Calculating the Num-
ber of Omul That Are Spawned." V kn. Krugovorot
veshchestva i energii v ozerakh i vodokhranilishchakh
[In book: Circulation of Matter and Energy in Lakes
and Reservoirs]. Listvenichnoye na Baykale, 1973.
24. N. S. Smirnova-Zalumi, V. V. Smirnov. "The Omul
Population in the Ecosystem in Lake Baykal." V kn.
13
-------
Krugovorot veshchestva i energii v ozerakh I vodo-
khranilishchakh [In book: Circulation of Matter and
Energy in Lakes and Reservoirs], Listvenichnoye na
Baykale, 1973.
25. V. D. Pastukhov. "The Role of Seals in the Ecosystems
of Continental Water Expanses (the Lake Baykal Ex-
ample)." V kn. Krugovorot veshchestva i energii v
ozerakh i vodokhranilishchakh [In book: Circulation of
Matter and Energy in Lakes and Reservoirs]. List-
venichnoye na Baykale, 1973.
26. "Wind Erosion of Soils and Measures to Deal With It."
Trudy Buryatskogo instituta yestestvennytkh nauk, No.
9, 1971. Ulan-Ude.
FIGURES. TERRITORIAL SUBSYSTEMS
AND TRANSPORT OF POL-
LUTANTS INTO LAKE BAYKAL.
14
-------
A Comprehensive Environmental Analysis of the Upper
Potomac Estuary Eutrophication Control Requirements
N. A. Jaworski *
I. INTRODUCTION
In the United States of America, lakes, rivers, and
estuaries are being subjected to increasing threat of
eutrophication. Some examples of ecosystems in the
U.S.A., that are currently experiencing euthrophica-
tion problems are Lake Erie, Lake Mendota, Lake
Washington and the Potomac Estuary.
The eutrophication process is a natural phe-
nomenon. However, man has greatly accelerated the
rate of eutrophication mainly by the increasing dis-
charge of nutrients from municipal and industrial
effluents and the increasing use of agricultural ferti-
lizers.
To comprehensively analyze the environmental
aspects of the eutrophication problems in the U.S.A.,
the following questions are being addressed:
1. What is the current trophic status of the lakes,
rivers, and estuaries? How significant is the eutrophi-
cation problem?
2. Which nutrient(s) is causing this rate of eutro-
phication?
3. Can this accelerated rate of eutrophication be
decreased? If so, where should the emphasis of the
control program be placed?
In this paper is presented a summary of the com-
prehensive environmental analysis of the eutrophica-
tion problem in the upper Potomac Estuary. This
case study illustrates the type of analysis that is re-
quired to develop a control program for eutrophica-
tion. Applicability of this study to other water bodies
is also addressed.
H. BRIEF DESCRIPTION OF THE
STUDY AREA
The Potomac River Basin, with a drainage area
of approximately 38,000 square kilometers (km2),
is the second largest watershed in the Middle Atlantic
States. From its headwaters on the eastern slope of
the Appalachian Mountains, the Potomac flows first
northeasterly and then generally southeasterly some
644 kilometers (km), flowing past the Nation's
Capital. The Potomac is tidal from Washington,
D.C., to its confluence with the Chesapeake Bay, a
distance of 183 km (Figure 1).
The study area includes the tidal portion, which is
about 60 meters (m) wide at its uppermost reach
near Washington and broadens to nearly 10 km at
its mouth. Except for a 7.5 m shipping channel and
a few reaches where depths up to 30 m can be found,
the tidal portion is relatively shallow with an average
depth of approximately 5.5 m.
Of the 3.3 million people living in the entire
basin, approximately 2.8 million reside in the upper
portion of the Potomac Estuary within the 7,300
km2 which comprises the Washington Metropolitan
Area. The lower area of the tidal portion, which
drains 8,300 km2, is sparsely populated.
The upper reach above Indian Head, although
tidal, is essentially fresh water. The middle reach is
normally the transition zone from fresh to brackish
water. The lower reach is mesohaline with chloride
concentrations near the Chesapeake Bay ranging
from approximately 7,000 to 11,000 mg/1.
The average freshwater flow of the Potomac River
near Washington, before diversions for municipal
water supply, is 305 cubic meters per second (cms)
with a median flow of 185 cms. The flow of the Po-
tomac is virtually unregulated and is thus character-
ized by extremely high and flashy flows often ap-
proaching 2,500 cms during flood conditions and 30
cms during droughts.
III. WATER QUALITY PROBLEMS
Early historical observations of the water quality
conditions include reports that in the later 1790's
President Adams swam in the Potomac Estuary near
Washington, D.C. By the 1860's when Abraham
Lincoln was president, the canals leading into the
Potomac Estuary, as well as the Potomac Estuary
itself, often emitted objectionable sewage odors forc-
ing Mr.' Lincoln to leave the White House at night.
From the year 1870, when the first sewers and cul-
verts were constructed, to the year 1938, when the
•Director, Pacific Northwest Environmental Research Laboratory, U.S. Environmental Protection Agency
15
-------
CHAIN BRIDGE
ZONE I
WOODROW WILSON BRIDGE
PISCATAWAY
^vy,v
KILOMETERS BELOW
CHAIN BRIDGE
MARYLAND POINT
PINEY POINT
LEGEND
MAJOR WASTE TREATMENT PLANTS
24.1
48.3
72.4
10
10 20
KILOMETERS
FIGURE 1. STUDY AREA
16
-------
Table 1
Water Quality Problems
Upper Potomac Estuary
Kilometers Major Type
of River of
Reach Affected Pollution
Chain Bridge to Hains Point
Mains Point to Piscataway Creek
Piscataway Creek to Maryland
Point
Anacostia Tidal River
11
16
50
13
Frequently high bacterial counts
Low-dissolved oxygen concentrations
Nuisance algal growths
Frequently high bacterial counts and
low-dissolved oxygen concentrations
Major Source
of
Pollution
Overloaded sanitary sewers and
combined sewer overflows
Effluents from wastewater treatment
facilities
Nutrients in wastewater discharges
Combined and sanitary sewer over-
flows
first primary treatment plant was built, almost all of
the sewage from the Washington Metropolitan Area
was discharged untreated into the Potomac Estuary.
The burgeoning population growth in the Wash-
ington Metropolitan Area has compounded the water
quality management problem. The accelerated pop-
ulation growth has completely outstripped attempts
to provide adequate facilities for wastewater treat-
ment. In addition, much of the growth has been un-
controlled in nature and location, and it is now diffi-
cut to provide adequate wastewater collection and
treatment within the limited space available for such
facilities in the area. Changes in composition of the
wastewater, mainly in the phosphorous content, have
also had a pronounced effect on water quality.
Since the first sanitary survey was made by the
U.S. Public Health Service in 1913 (1), the water
quality with respect to bacterial densities and dis-
solved oxygen levels in the Washington Metropolitan
Area has been degraded as a result of the discharge
of either untreated or inadequately treated municipal
sewage.
The upper estuary has been divided into the four
reaches according to type and source of pollution
as itemized in Table 1 and shown in Figure 2.
There are about 90 kilometers of the upper estuary
degraded with the effects of eutrophication being
pronounced in approximately 50 kilometers. In ad-
dition, the Upper Potomac Estuary, including the
Anacostia Tidal River, is subjected to periods of
high concentrations of sediment. During initial
studies of the estuary, major emphasis was placed on
the high bacterial and low-dissolved oxygen problems
(2) (3). More recently, the nuisance algal problem
has also been included.
The time frame of algal problem development has
been developed from several studies as summarized
by Jaworski et al. (4). As shown in Figure 3, there
have been historical invasions of nuisance growths
in the Upper Potomac Estuary.
From a review of data in Figure 3, it would appear
that nuisance conditions did not develop linearly
with an increase in nutrients. Instead the increase
in nutrients appeared to favor the growth and even-
tually the domination by a given species. As nutrients
increased further, the species in turn was rapidly
replaced by another dominant form. For example,
water chestnut was replaced by water milfoil which
in turn was replaced by blue-green algae, mainly
Anacystis.
The massive blue-green algal blooms, which have
occurred every summer since 1960, appear to be
associated with large increases in phosphorus and
nitrogen loadings in the upper reaches of the Po-
tomac River tidal system (Figure 3). The blooms
have persisted since the early 1960's although during
this period the amount of organic carbon from waste-
water was reduced by almost 50 percent when com-
pared to that discharged prior to 1960.
Under warm temperature and low-flow conditions,
large standing crops of this alga develop green mats
of cells. Chlorophyll a concentrations range from
approximately 50 to over 200 u g/1 in these areas
of dense growth which at times extend over ap-
proximately 80 km of the upper and middle reaches
of the estuary. These high chorophyll levels are 5
to 10 times those reportedly observed in other eutro-
phic waters by Brezonik et al. (5) and by Welch
(6). During a dense bloom, the dry weight of cells
ranges from 10 to 25 mg/1 which is almost twice
those reported from the lakes in Madison, Wisconsin.
In the mesohaline portion of the lower reach of
the Potomac Estuary, the algal populations are not
as dense as in the freshwater portion. Nevertheless,
at times large populations of marine phytoplankton
(primarily the dinoflagellates Gymnodinium sp. and
Amphidinium sp.) occur producing what are known
as "red tides."
IV. NUTRIENT CONCENTRATIONS
AND SOURCES
The concentration of nutrients along the estuary
varies as a function of wastewater loading, tempera-
17
-------
PERIODICALLY
HIGH BACTERIAL
DENSITIES
MARYLAND
PERIODICALLY HIGH
BACTERIAL DENSITIES
AND LOW DISSOLVED
OXYGEN LEVELS
PERIODICALLY MODERATE
BACTERIAL DENSITIES.
LOW DISSOLVED OXYGEN
LEVELS AND BEGINNING
OF ALGAL BLOOMS
PISCATAWAY CREEK
PRONOUNCED
NUISANCE ALGAL
GROWTHS
BRACKISH
WATERS
FIGURE 2. ESTUARY REACHES
18
-------
10,000 -
- 100.000
- 80,000
. 60,000
_ 40.000
- 20,000
o
$
z
O
C
O
1910 1920 1930 1940 1950
FIGURE 3. ESTUARY HISTORY
1970
ture, freshwater inflow from the upper basin, bio-
logical activity, and salinity. The annual distribution
of the various nutrient concentrations has been re-
ported by Jaworski, et al. (4), and the summer levels
are summarized in Table 2 for five key stations
along the estuary.
In the vicinity of the Woodrow Wilson Bridge,
there is an increase in alkalinity, total phosphorus,
NO* and NOa nitrogen, and ammonia nitrogen with
a corresponding decrease in pH, all of which can
be attributed to the 1.23 million m3 per day of waste-
water discharged in the Washington Metropolitan
Area. The rapid disappearance of the ammonia
nitrogen between Woodrow Wilson Bridge and In-
dian Head and Maryland Point is caused by the
oxidation of NHs to NO2 + NO3 by the nitrifying
bacteria. The sharp drop in NO2 + NO3 nitrogen
between Indian Head and Maryland Point is at-
tributable to the large uptake by the pronounced
algal growths in this area.
A complete analysis of the nutrient sources in the
Upper Potomac Estuary has been made by Jaworski
et al. (4). A summary of the major sources is pre-
sented in Table 3 for low, median, and high Potomac
River flows.
When considering only the upper basin runoff and
wastewater discharges to the estuary as summarized
in Table 3, it can be concluded that the order of
percentage of nutrients controllable by wastewater
treatment is (1) phosphorus, (2) nitrogen, and (3)
carbon.
While the controllable phosphorus and nitrogen
percentages decrease at higher flows, these conditions
usually occur during the months of February, March,
and April, when temperatures and algal crops are
lowest. Since nuisance algal conditions occur pri-
marily in the upper or the freshwater portion of the
estuary, the higher flow effects are reduced con-
siderably by the time the blooms are most prolific
during the months of July, August, and September.
Under low- and median-flow conditions, both
nitrogen and phosphorus are largely controllable. If
allowances are made for atmospheric contributions
of nitrogen, only an approximate 200 kg/day of
nitrogen could be added to the upper estuary, which
is less than 10 percent of the nitrogen in the waste-
water discharges. Thus, during summer months, algal
control by management of nitrogen instead of phos-
phorus appears to be a feasible alternative.
Using only 0.1 percent of the transfer rate, the
amount of carbon (CO2) potentially available from
the atmosphere was estimated to be approximately
431,000 kg/day (4). Moreover, with the upper
reach of the estuary well mixed due to tidal action,
19
-------
Table 2
Average Range of Concentration
Summer Conditions
Upper Potomac Estuary
Stations and
Kilometers from
Chain Bridge
Chain Bridge (0.0)
W. Wilson Bridge (19.5)
Indian Head (49.3)
Maryland Point (84.3)
301 Bridge (104.7)
PH
(units)
7.5-8.0
7.0-7.5
7.2-8.0
7.5-8.2
7.5-8.0
Alkalinity
(mg/1)
80-100
90-110
70-90
60-85
65-85
Total
Phosphorus
(mg/1)
0.08-0.20
0.30-1.20
0.20-0.40
0.10-0.25
0.05--0.20
NO,+NO>
Nitrogen
(mg/1)
0.3-1.0
0.8-1.2
0.5-1.5
0.1-0.3
0.1-0.2
NH,
Nitrogen
(mg/1)
0.10-0.50
1.00-3.00
0.10-0.50
0.05-0.30
0.05-0.20
Carbon
Nitrogen
Phosphorus
Carbon
Nitrogen
Phosphorus
Table 3
Summary of Major Nutrient Sources
Upper and Middle Reaches of the Potomac Estuary
Low-flow Conditions
(95% of time exceeded)
(Potomac River Discharge at Washington, D.C.=40 cubic meters/sec)
Carbon
Nitrogen
Phosphorus
Upper
Basin
Runoff*
77,100
3,000
450
Percent
of
Total
52
10
4
Estuarine
Wastewater
Discharges
72,600
27,200
10,900
Percent
of
Total
48
90
96
Total
(kg/day)
143,700
30,200
11,350
Median-flow Conditions
(50% of time exceeded)
(Potomac River Discharge at Washington, D.C.= 185 cubic meters/sec)
159,000 68 72,600 32
18,100 40 27,200 60
2,400 18 10,900 82
High-flow Conditions
(5% of time exceeded)
(Potomac River Discharge at Washington, D.C.= 1150 cubic meters/sec)
680,000 90 92,600 10
185,000 87 27,200 13
10,000 47 10,900 53
231,600
45,300
13,300
752,600
212,200
20,900
'Upper basin runoff: both land runoff and wastewater discharges in upper basin.
recruitment of carbon from benthic decomposition
appears to be a significant source of inorganic car-
bon as well. When all potential sources are con-
sidered, it appears that management of carbon for
algal control is not a feasible alternative at the pres-
ent time.
V. EUTROPHICATION CONTROL
REQUIREMENTS
For water quality management purposes, the
Upper Potomac Estuary may be considered hyper-
eutrophic when nuisance plant organisms become
predominant as it is now occurring with the blue-
green algae. Four major water use interferences have
been offered by Jaworski, et al. (4) including the
required reduction in the algal standing crop for
each of the conditions as shown in Table 4.
The first two are related to the oxygen budget.
Studies have demonstrated that during the summer
months more ultimate oxygen demand is added to
the upper estuary as a result of these algal growths
than from the present wastewater discharges, though
this demand may not be fully exerted.
The aesthetic and recreational potential of the
upper estuary is impaired by the extensive mats of
algae which cause objectionable odors, clog marinas,
and cover beaches and shorelines. The potential use
of the estuary as a water supply source could also
be impaired because of possible toxin problems as-
sociated with the blue-green algae.
Of the four interferences, the highest reduction
percentages are for control of algal growths to pre-
vent nuisance conditions. From the data in Table
4, a 75 to 90 percent reduction in chlorophyll a con-
centrations will be required to limit chlorophyll levels
to approximately 25/ig/l, the concentration selected
as the desired upper limit for eutrophication control
in the Upper Potomac Estuary.
20
-------
Table 4
Subjectve Analysis of Algal Control Requirements
Water Quality or
Water Use Interference
DO Depression Caused by
Algal Respiration
Increase of Total Oxygen
Demanding Load
Recreational & Aesthetic
Nuisance Conditions
Toxins
Indications of
Interference
mg/1 of DO below
Saturation
mg/1 of Increase in
Ultimate BOD
Chlorophyll a
Concentration
Undefined
Magnitude of
Current Interference *
1.5 to 3.0 mg/1
15 to 30 mg/1
100 to > 250 ug/l
Unknown
Desired
Limit
0.5 mg/1
5.0 mg/1
25 ug/l**
Unknown
Required Percentage
Reduction of Current
Standing Crop
65-85
65-80
75-90
Unknown
•Under nuisance-bloom conditions, chlorophyll a concentrations range from 100 to >2SO»g/l
** Average over entire water column.
VI. NUTRIENT CRITERIA
The desired nutrient criteria were developed using
data from: (1) algal composition analysis, (2) an-
nual nutrient cycles and longitudinal profiles, (3)
bioassay studies, (4) review of historical data, (5)
comparison with a noneutrophic estuary, and (6)
algal modeling. Each method was used independently
in the development of a nutrient phytoplankton re-
lationship in the Potomac Estuary.
When investigating the role of nitrogen and phos-
phorous in eutrophication of the Potomac Estuary,
a detailed study of the movement of these nutrients
was made using both a real-time dynamic water
quality estuary model (8) and an average tidal
mathematical model (9). The dynamic modef was
expanded to predict the concentration of chlorophyll
a based on the utilization of inorganic nitrogen. In
Figure 4, predicted NO2 + NO3, NH3 and chlorophyll
a profiles are presented. The predicted maximum
concentrations conform closely to observed data in
both distribution and magnitude.
From field data, bioassay studies, and mathema-
tical model runs, it was concluded that the standing
crop of blue-green algae can be predicted using the
nitrogen cycle. This further supports the premise
that the nitrogen availability appears to control the
standing crop. Similar methods also indicated that if
total phosphorus were in the range of 0.03 to 0.1
mg/1, the desired 25/^g/l level of chlorophyll could
be realized.
Based on the six independent methods of analysis
and the 25 fig/1 level of chlorophyll a, nutrient
criteria were developed for reversing the eutrophica-
tion process occurring in the freshwater portion of
the Potomac Estuary as shown in Table 5. Since
there are over 5.0 mg/1 of inorganic carbon
in the estuary, even under maximum bloom con-
ditions, no criterion for carbon could be established
at the present time.
The lower values of these ranges are to be applied
to the freshwater portion of the middle reach and
Table 5
Required Concentration for
Eutrophication Reversal
Parameter
Concentration Range
Inorganic Nitrogen
Total Phosphorus
0.30-0.5 mg/1
0.03-0.1 mg/1
to the embayment portions of the estuary in which
the environmental conditions are more favorable
toward algal growth. The higher values are more
applicable to the upper reach of the Potomac Estuary
which has a light-limited euphotic zone of usually
less than 0.60 meters.
Since the growth of massive blue-green algal mats
is apparently restricted to the freshwater portions
and dinoflagellates are often encountered in the
mesohaline environment, no specific nutrient criteria
have been established for the mesohaline portion of
the Potomac Estuary. It appears that if the afore-
mentioned nutrient criteria are achieved in the upper
estuary, adequate control of the eutrophication proc-
ess in the lower reach of the estuary should also be
realized.
VII. WASTEWATER MANAGEMENT
ZONES
Initially in 1969, to facilitate the determination
of wastewater management requirements, the upper
and middle reaches of the estuary were divided into
three 15-mile (24 km) zones with similar physical
characteristics, beginning at Chain Bridge (see Fig-
ure 1). This zoning concept allows for greater flex-
ibility ih developing control needs.
More recent studies have suggested that Zone I
be divided into three subzones described as follows:
Subzone Description
I-a Potomac Estuary from Chain Bridge to Hains
Point, a distance of 12.1 kilometers.
I-b Anacostia tidal river from Bladensburg, Mary-
land, to the confluence with the Potomac, a dis-
tance of 14.4 kilometers.
I-c Potomac Estuary from Hains Point to Broad
Creek, a distance of 12 kilometers.
21
-------
125-
100-
fc€ 75-
§S
u
z
i-i,
E E
50-
25-
0
1.4'
1.2 -
1.0-
0.6-
0.4 -
0.2 -
0.0
PREOICTE
TEMP. = 27.5C
FLOW = 79.29 cms
NH3 (PREDICTED)
N02 + NO3 (OBSERVED)
NO2 + NOs (PREDICTED)
SALINITY INTRUSION
fo
20 30 40 50
KILOMETERS BELOW CHAIN BRIDGE
60
70
FIGURE 4. PREDICTED AND OBSERVED PROFILES
Discharges into tidal embayments were investigated
on an individual basis.
Using the zone concept, total maximum loadings
for each pollutant were developed for each zone.
Allocation of pound loadings for each discharge can
be obtained by prorating the zonal poundage using
various bases such as population, drainage areas,
geographical subdivisions, and others.
Vin. WATER QUALITY SIMULATION
MODELS
Water quality simulations and wastewater treat-
ment investigations were made using the FWQA
Dynamic Estuary Model (DEM) and the DECS
III, a general purpose estuarine model. The DEM
(8) is a real-time system utilizing a two-dimensional
network of interconnecting junctions and channels
which permits direct inclusion of tidal embayments
in the flow representation. The model is comprised
of a hydraulic component that describes tidal move-
ment and a quality component. The DEM includes
the basic transport mechanisms of advection and
dispersion as well as the pertinent sources and sinks
for each constituent. This model was used to simulate
water quality conditions on an hourly basis and to
determine zonal loadings under low-flow conditions.
DECS III is based on a time-dependent tidal aver-
age solution of the basic mass balance equations
(9). This model was used to investigate seasonal
variations in the nitrogen and phosphorus distribu-
tions in the Upper Potomac Estuary.
Simulations of phosphorus discharges into the
Potomac Estuary were made using second-order re-
action kinetics with a deposition rate of 0.05 mg/day
at a temperature of 29 °C. The allowable phosphorus
loadings were determined based on maintaining an
average of 0.1 mg/1 of phosphorus (P) within Zone
I, 0.067 mg/1 (P) within Zone II, and 0.03 mg/1
(P) within Zone III.
For investigating the role of nitrogen in water
quality management, a feedback system of the intro-
gen cycle was incorporated into the dynamic estuary
mathematical model similar to that proposed by
Thomann et al. (10). The model consists of six
possible reactions: (1) chemical and biological de-
composition of organic nitrogen to ammonia, (2)
22
-------
bacterial nitrification of ammonia to nitrite and
nitrate, (3) phytoplankton utilization of ammonia,
(4) phytoplankton utilization of nitrite and nitrate,
(5) deposition of organic nitrogen, and (6) decay
of phytoplankton.
IX. MAXIMUM CONSTITUENT
LOADINGS PER ZONE
Using the models and coefficients as described in
the previous sections, zonal loadings were determined
for UOD nitrogen, and phosphorus (see Table 6).
The loadings presented are maximum allowable
loadings for each zone, assuming that adjuacent zones
are loaded to their maximums.
The increases in loadings for the lower zones
mainly reflect the increase in the estuary's volume
and tidal transport. Since nitrogen and phosphorus
criteria for the lower zones are more stringent, the
increase in nutrient loadings in this area is not as
pronounced as for UOD.
For the projected 1980 wastewater loading con-
ditions, the anticipated percent removal rates for
Zone I-c would be approximately 93 percent UOD,
96 percent prosphorus and 93 percent nitrogen.
Since Zones II and III do not currently receive as
much wastewater, the removal percentages will not
be as high.
X. SEASONAL WASTE TREATMENT
REQUIREMENTS
1. Ultimate Oxygen Demand
The maximum allowable UOD loadings, as pre-
sented in Table 5 for the three upper zones of the
Potomac Estuary, were developed for low-flow and
summer temperature conditions. During high tem-
perature periods, the effects of nitrogenous oxygen
demanding substances on the dissolved oxygen bud-
get were determined to be quite significant.
Studies have shown that during very warm periods,
when nitrification rates are high, the nitrogenous
component of UOD exerts 250,000 Ibs/day of oxy-
gen demand as compared to approximately 200,000
Ibs/day from the carbonaceous demand. During low
temperature periods, when the ambient water tem-
perature is less than 15°C, the effects of nitrification
on the dissolved oxygen budget have been shown to
be negligible.
Based on these findings, it was recommended that
(1) UOD loading presented in Table 5 be applied
only under summer conditions, (2) the removal or
oxidation of ammonia in wastewater discharges be
provided whenever the water temperature is above
15°C, and (3) a high degree of removal of suspended
solids (a maximum of 15 mg/1 in the effluent) and
carbonaceous oxygen demanding material (a mini-
mum of 90 percent) be provided on a year-round
basis to prevent the accumulation of sludge deposits
in the vicinity of sewage treatment plant outfalls
during cooler weather and to maintain high DO
levels under ice cover.
Table 6
Maximum UOD, Phosphorus, and
Nitrogen Wastewater Loadings for
Low-Flow Summer Conditions
(kg/day)
Zone
I-a
I-b
I-c
II
III
Allowable UOD
1,800
1,400
33,800
85,500
171,000
Phosphorus
90
40
400
680
900
Nitrogen
450
140
1,580
2,600
4,100
2. Phosphorus and Nitrogen
The loadings, as presented in Table 6, were estab-
lished for low-flow conditions. During these periods,
the nutrient contribution from the upper basin is in-
significant when compared to that contained in the
wastewater discharges.
To determine whether the nitrogen and phos-
phorus criteria could be met under varying Potomac
River inflows and varying nutrient contributions from
the upper basin, an annual simulation was made of
conditions from February 1969 to September 1970.
This period was critical because a drought condition
occurred during June and July of 1969, and August
flows were over four times above the average dis-
charge. Thus, both low and high summer flows were
simulated.
Mathematical model analysis of the annual dis-
tribution of phosphorus in the critical algal growing
area showed close agreement between the observed
and predicted phosphorus profiles (see Figure 5).
Also shown in Figure 5 are the predicted annual
phosphorus profiles resulting from year-round waste-
water phosphorus removal in the upper estuary, as-
suming: (1) no control and (2) 50 percent control
of the phosphorus loading originating in the Upper
Potomac River Basin. From the data presented in
Figure 5, it was concluded that both (1) the adher-
ence to maximum allowable phosphorus loadings
from wastewater effluents being discharged directly
into the estuary (see Table 6) and (2) a 50 percent
reduction of the total incoming phosphorus load
from the upper basin, will be required if the recom-
mended maximum phosphorus criteria are to be real-
ized. In order to achieve a 50 percent reduction in
the present phosphorus load from the Upper Po-
tomac River Basin, the current overall wastewater
contribution of 2700 kgs/day must be reduced to
less than 320 kgs/day.
Because of the more stringent criteria, particularly
in the lower zones including longer transport time,
the possibility of recycling previously deposited phos-
phorus from bottom muds and the unpredictability
23
-------
of phosphorus in various forms being transported
from the upper basin, year-round phosphorus re-
moval at all wastewater treatment facilities in the
Potomac River Basin was recommended.
As presented earlier, the necessity for unoxidized
nitrogen control in wastewater discharges to main-
tain a high dissolved oxygen content in the Potomac
Estuary was restricted to that time of year when
water temperatures exceed 15°C. When evaluating
the need for annual nitrogen control to prevent ex-
cessive algal blooms, controllability, duration of
nuisance blooms and temperature become significant
factors.
While spring blooms of diatom algal cells have
been observed, the major nuisance blue-green algal
blooms of algae usually occur during the months of
July, August, and September. During these months,
the controllability of nitrogen by wastewater treat-
ment is usually the greatest and the water tempera-
ture highest.
Mathematical model predictions of inorganic nitro-
gen concentrations in critical algal growing areas
based on (1) no estuary wastewater nitrogen re-
moval, (2) nitrogen removal during periods with
temperatures above 15°C (April-November), and
(3) year-round nitrogen removal are presented in
Figure 6. For the nitrogen loading as given in Table
6, the inorganic nitrogen concentration of less than
0.3 mg/1 can be achieved for drought conditions
such as in June and July. The abnormally high Au-
gust Potomac River flow condition and resulting high
upper basin loading caused the nitrogen level to in-
crease to approximately 0.5 mg/1.
While it may be desirable to maintain nitrogen
concentrations at or below the selected criteria at all
times, the high flows from the upper basin during the
winter and spring months contribute high nitrogen
loadings which increase the nitrogen concentrations
above acceptable levels regardless of wastewater
treatment practices. In considering (1) that nuisance
algal growths occur mainly during the months of
July, August, and September, (2) that seasonal
nitrogen removal is generally adequate for main-
taining the desired nitrogen concentration during this
time, and (3) that unoxidized nitrogen control is
required only for warm temperature periods, it was
recommended that nitrogen removal for algal con-
trol, as in the case of nitrogenous demand for oxygen
enhancement, be limited to periods when water tem-
peratures in the estuary exceed 15°C.
In developing the seasonal requirements, emphasis
was placed on maintaining a balanced ecological
community structure in the upper or freshwater por-
tion of the estuary. More research efforts in both
transport mechanisms and nutrient algal relation-
ships are needed to determine management require-
ments for the lower or saline portion of the estuary.
XI. ESTIMATED COSTS
The present cost of providing for additional waste-
water flows and treatment requirements from the
year 1970 to 2020, including operation, maintenance,
and amortization cost, has been estimated to be
$1.34 billion, with a total average annual cost of
$64.9 million. The unit treatment processes assumed
include activated sludge, biological nitrincation-de-
nitrification, lime clarification, nitration, effluent aera-
tion, and chlorination.
Reduction of the initial capital and operation and
maintenance costs to a per capita basis are shown in
Table 7. The above summary, which does include
replacement cost, indicates that the cost of waste-
water treatment in the Upper Potomac Estuary is
about $13 to $24 per person/per year. This expendi-
ture, which includes the cost of the activated carbon
process, will renovate the water to the chemical and
bacteriological levels to meet drinking water quality
standards.
XII. APPLICABILITY OF POTOMAC
ECOSYSTEM ANALYSIS TO OTHER
EUTROPfflC SYSTEMS
In the summary of comprehensive analysis pre-
sented for the Potomac Estuary ecosystem, water
quality problems of eutrophication and low dissolved
oxygen were addressed. Limits were developed for
nutrients and ultimate oxygen demanding materials.
For other estuaries along the North Atlantic Coast
the pollution stress analysis would be very similar.
However, for some lakes and rivers, the compre-
hensive analyses can be signicantly different.
To cope with this variation aspect of the eutrophi-
cation analysis, a National Eutrophication Survey
was initiated in 1972. The objectives of the survey
are as follows:
1. To assess the current condition of lakes and
impoundments (over 800) which have, or are being
threatened with, eutrophication problems.
2. To develop scientific rationale which can be
used in setting and revising water quality criteria
for nutrients and developing eutrophication priorities
and policies.
3. To assist federal and state governments in set-
ting nutrient control priorities and policies.
4. To evaluate methods, including analytical tech-
niques, that can be used for lake classification.
5. To assess the contribution of non-point sources
of nutrients in varying geographic areas, and, on a
limited basis, relate the contribution of nutrients to
land use management.
In addition to the above, research efforts in the
U.S.A. are being conducted to develop eutrophica-
tion controls for freshwater and marine environ-
ments. Specific objectives of this research are to:
24
-------
0.9-
0.8-
0.7-
0.6-
S O.i.
I
n 0.4-1
0.2-
0.1-
0.0
. PREDICTED
OBSERVED
CONTINUOUS (YEAR AROUND) PHOSPHOROUS
REMOVAL
0.3-
0.2-
0.1-
0.0
1 II I I I
NO PHOSPHORUS CONTROL IN UPPER POTOMAC BASIN
ASSUMING 50% OF PHOSPHORUS LOAD FROM UPPER
BASIN IS CONTROLLED
FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
1969
FIGURE 5. ANNUAL PHOSPHORUS PROFILES
POTOMAC ESTUARY AT INDIAN HEAD
JAN FEB
1970
3.2-j
28-
2.4-
I 2.0-
It
Z
3 0.8-
- 0.4.
0.0
NO NITROGEN REMOVAL
NITROGEN REMOVAL FROM APRIL TO NOVEMBER
YEAR AROUND NITROGEN REMOVAL
FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
1969
FIGURE 6. SIMULATED ANNUAL NITROGEN PROFILES
POTOMAC ESTUARY AT INDIAN HEAD
JAN FEB
1970
25
-------
Table?
Per Capita Cost Projections
Item
Average Population
Initial Capital
Cost/Time Period
Capital Cost/Person/Year
O&M Cost/Year
O&M Cost/Person/Year
Total Cost/Person/Year
1970-1980
3,350,000
$570,000,000
$17.0
$25,100,000
$7.5
$24.5
1980-2000
5,350,000
$528,000,000
$4.0
$46,200,000
$8.6
$13.5
2000-2020
8,000,000
$1,173,000,000
$7.3
$72,400,000
$9.1
$16.4
(1) develop an understanding of the eutrophication
process, (2) develop methods for monitoring eutro-
phication conditions and for predicting, with the
use of mathematical models, the impact of nutrient
sources on eutrophication, (3) develop technology
to control and reverse eutrophication processes and
(4) to establish the practicability of using this tech-
nology through pilot and demonstration scale appli-
cation.
Our research efforts to date appear to indicate
that the approach to comprehensive environmental
problems of the Potomac Estuary is adaptable to
other ecosystems. However, the uniqueness of many
of the lakes and rivers based on morphological and
limnological characterstics requires specific analysis
and data interpretation. It also appears that control
measures will be developed along the morphological
and limnological basis for a given set of ecosystems.
REFERENCES
1. U.S. Public Health Service, "Investigation of the Pollu-
tion and Sanitary Conditions of the Potomac Water-
shed," Hygienic Laboratory Bulletin No. 104, Treasury
Department, February, 1915.
1. U.S. Army Corps of Engineers, "Potomac River Basin
Report," Vol. I-Vol. VIII, North Atlantic Division,
Baltimore District, February 1963.
3. Davis, Robert K., "The Range of Choice in Water
Management, A Study of Dissolved Oxygen in the
Potomac Estuary," Johns Hopkins Press, Baltimore,
Maryland, 1968.
4. Jaworski, N. A., Donald W. Lear, Jr., Orterio Villa, Jr.,
"Nutrient Management in the Potomac Estuary," Pre-
sented at the American Society of Limnology Sym-
posium on Nutrients and Eutropbication, Michigan
State University, East Lansing, Michigan, February
1971.
5. Brezonik, W. H., W. H. Morgan, E. E. Shannon, and
H. D. Putnam, "Eutrophication Factors in North Cen-
tral Florida Lakes," Florida Engineering and Industrial
Experiment Station, Bulletin Series No. 134, Gaines-
ville, Florida, August 1969.
6. Welch, E. B., "Phytoplankton and Related Water Qual-
ity Conditions in an Enriched Estuary," Journal Water
Pollution Control Federation, Vol. 40, pp. 1711-1727,
October, 1968.
7. Lawton, O. W., "The Madison Lakes Before and After
Diversion," Trans. 1960 Seminar on Algae and Metro-
politan Wastes, pp. 108—117, Robert A. Taft Sanitary
Engineering Center, Technical Report W61-3, 1961.
8. Feigner, Kenneth and Howard S. Harris, "Documenta-
tion Report, FWQA Dynamic Estuary Model," FWQA,
U.S. Department of the Interior, July 1970.
9. Thomann, Robert V., "Mathematical Model for Dis-
solved Oxygen," Journal of the Sanitary Engineering
Division, ASCE, Vol. 89, No. SA5, October 1963.
10. Thomann, R. V., Donald J. O'Connor, and Dominic M.
DiTorro, "Modeling of the Nitrogen and Algal Cycles
in Estuaries," presented at the Fifth International Water
Pollution Research Conference, San Francisco, Califor-
nia, July 1970.
11. Environmental Protection Agency, "National Capital
Region Water and Waste Management Report," Wash-
ington, D.C., April 1971.
26
-------
The Ecological System Stability Concept
V. D. Fedorov*
Complexity must be considered to be the main
feature of biological systems. This is so because of
the variety of elements comprising the system and
the special features of their interrelationships (stress
and coordination). The impossibility of studying all
the variables of very complex systems inevitably in-
troduces a subjective aspect into the selection of
some of them, those the specialists deem most im-
portant to an understanding of the nature of the
phenomenon under investigation. The a priori rank-
ing of system variables according to importance
(made consciously or unconsciously) therefore is
necessary before the investigation of biological sys-
tems of any complexity can be undertaken.
The imposed necessity of operating with a limited
number of variables when studying the integral prop-
erties of very complex systems precludes the estab-
lishment of simple cause-effect types of relationships
between the different properties of biological systems.
This is what has given rise in biology to the point of
view that the whole is greater than its component
parts, so that the properties of the whole cannot be
studied effectively after the whole has been broken
down into parts to study the properties of the latter.
This approach has led to the concept of the exist-
ence of special biological laws that cannot be ex-
plained by resort to the positions of the strict prop-
ositions of such sciences as physics and chemistry.
This point of view is quite productive today, because
research in the integral properties of biological sys-
tems that reflects the face of the systems, their ap-
pearance, fate, and coordination, has been moved
up on the agenda.
1. Characteristics and types of biological systems
Let us define a biological system as an object
capable of forming negative entropy, the integral
properties of which are the result of the interaction
of the parts of which it is formed. Let us now dis-
tinguish the following integral properties as primary
in biological systems:
(a) complexity, determined by the diversity of the
elements involved, and their interrelationships, as
well as the stress on the interrelationships in the
system;
(b) autonomy, which determines the special fea-
•Motcow Sute Unlveriity, Moicow
tures of the position of the system at subsystem levels
within the limits of a larger system;
(c) reliability, which determines the probability
of processes taking place in the system, and of the
temporal maintenance of the constancy of its prop-
erties;
(d) stability, which determines the capacity of
biosystems to withstand external disturbing effects,
while at the same time preserving the system's
"face," which is determined by its past history.
There are objective reasons for the subdivision
of biological systems into three groups according to
the differences in the system of regulations governing
their behavior (Fedorov, 1972):
(1) a-type biological systems with central control;
that is a subsystem whose regulatory link with any
element is stronger than the link between such ele-
ment and any other, taken arbitrarily. This category
includes all animal organisms with central and vege-
tative nervous systems;
(2) p-type biological systems with "passive" con-
trol (Milsum, 1964), which can include plant or-
ganisms with their humoral regulation;
(3) r-type biological systems with "random" con-
trol, which can include all ecological systems not
carriers of a special behavior "program" in the DNA
molecules, as is the case in the first two groups.
It is only natural that the importance of the integ-
ral properties listed will be dixerent for difierent
types of biological systems.
Thus, such integral properties as reliability and
autonomy, on which the carrying out of the "pro-
gram" is primarily dependent, will be of primary
importance for systems with a "program" (a-type
and p-type). At the same time, such integral charac-
teristics as complexity and stability, on which their
fate is primarily dependent, that is, the capacity to
withstand random "unprogrammed" environmental
disturbances, will be of primary importance for eco-
logical systems (r-type).
It is useful, for purposes of convenience in justi-
fying methods used to measure the integral proper-
ties of ecological systems, to separate the following
from among the "unprogrammed" disturbances of the
abiotic component.
27
-------
Environmental factors that interact with the living
component of the system, that is, that have an effect
on the system and experience a response from the
organisms, should be included among the intrasys-
tem factors (biogenic elements, for example).
The second group of environmental factors is
made up of so-called extra-system factors. These es-
tablish the conditions for the existence of concrete
ecosystems within the limits of larger system forma-
tions, that experience no noticeable response from
the biotic component (solar radiation, temperature,
humidity, and salinity, for example).
Finally, the third group of factors is made up of
the so-called disturbing factors, those for which the
ecosystems are poorly prepared (or not prepared at
all) by the preceding course of the evolution of the
species, and by the past history of the formation of
concrete communities under concrete biotope con-
ditions. They may include physically heterogeneous
factors, but their distinguishing feature is the sudden-
ness (on the geological time scale) of the effect cre-
ated by change in their intensity which is unusually
abrupt for the system. This action can include an
abrupt change in the reserve of any "normal" chemi-
cal element in the locus (for example, N and P,
which give rise to the problem of eutrophication), or
of a normally rare element (for example, salts of
heavy metals Hg, Pb, Cd) or, finally, the appearance
of a substance that is "new" to life, that has been
synthesized by man (for example, DDT, PCS and
other pesticides).
In fact, it is indeed the intensification of the dis-
turbing factors associated with the variety of forms,
and with the reinforcement of the scales of human
activity in the biosphere, that does much to explain
just how important the problem of studying the sta-
bility of natural systems is today. Therefore, our sub-
sequent elucidation of the material will be limited
to a discussion of certain of the general aspects of
ecological system stability, and to justifying methods
used to measure that stability.
2. Measuring the stability of variables
A biological system has numerous characteristics
that reflect the special features of its organization.
These characteristics can be considered as inde-
pendent variables of the system, and the need arises
in any study of a system to continually estimate
their behavior in time and/or space.
The measurement of the stability of an independ-
ent variable (Si) can be assumed to be the estimate
of the variation in the unit values of a function
around a mean in order to estimate the temporal
change in individual independent variables. Stability
thus is a measure of the temporal inconstancy of any
function. Such estimates of Si can be (Fedorov,
Bokolova, 1972) the mean values of the deviation in
amplitude, Ai [1], or the square of this deviation
A 2 [2], of the variable with respect to its mean value
(1)
(2)
where
Xj is the value of the variable being measured
at the time of measurement;
n is the number of measurements when AXJ
=xH-x,and
x"=ij x,.
1=1
The index of heterogeneity of criterion distribution
that we have proposed (Fedorov, 1973), and which
is an estimate of the ratio of the mean geometric to
the mean arithmetic value of the variable, A>, [3],
A3 =
(3)
is yet another method for estimating Si. But regard-
less of the method used to estimate the stability of
the individual variables, it can be shown that the set
of estimates obtained forms a bimodal distribution
(Fedorov, Sokolova, 1972) when the values found
for Si are ranked. One of the sets of estimates here
includes the indices that characterize the rate at
which the process occurs (that is, the intensity of
the function), so these estimates can be expressed as
a time derivative. Let us call them functional esti-
mates, assuming that
c = .L
"f
f n
(4)
The second set groups the estimates of indices
that can be taken to be some result of the processes
(functions) at the time of observation, so these esti-
mates can be expressed as a time integral. Let us
call them structural estimates, assuming that
(5)
28
-------
The most important functional characteristic of an
ecosystem is its productivity. The latter reflects the
rate of formation of the organic matter (biomass)
by the individual species populations making up the
community. The most important structural character-
istic of ecosystems is the biomass of the populations.
This raises a question. How can the integral prop-
erties of a system be evaluated quantitatively from
numerous individual estimates of independent vari-
ables? Separation of the indices of the ecosystem's
biotic component into functional and structural is
quite productive in finding an answer to the question
just posed because one can compare the mean esti-
mates (or modal class positions) for the two sets of
estimates obtained, that is, for S« and S,.
It is readily apparent that comparison of the S*
and S. estimates leads to the introduction of a certain
measure that characterizes the internal state of the
system, and this measure, as will be shown below,
determines the homeostasis of the ecosystem,
3. The homeostasis of biological systems
The term homeostasis was introduced initially in
1932, by Walter Cannon (Cannon, 1932) to desig-
nate the mechanisms that maintain the equilibrium
of the internal medium (blood sugar constancy, and
constant body temperature in warm-blooded animals,
for example), which latter is constantly being upset
by changing conditions in the outside world and is
constantly being restored by special internal mecha-
nisms with feedback action. These mechanisms are
under the control of the autonomic nervous system,
so are inherent in all living systems with a central
control subsystem; that is, a-type systems, or simply
a-systems.
Much later (Fedorov, 1970a), while analyzing
the organizational features of a- and r-systems, we
turned our attention to the existence of a principle
difference between them. Thus, functional reconstruc-
tions of a-systems are designed to maintain, in the
final analysis, constancy of the structural character-
istics (preservation of the specific external appear-
ance of the systems attributable to the functioning
of numerous mechanisms for regulating metabolism,
for example). Structural reconstructions of r-systems
were designed to maintain the constancy of the func-
tional characteristics. Change in the ratio of the
photophilic to thermophilic forms, and of the
thermophilic to psychophilic species makes it pos-
sible to maintain the constancy of the production
characteristics during the vegetative period when
there is a variation in temperature conditions and
illumination. Thus, the presence of ecological mech-
anisms that periodically maintain the constancy of
the functional characteristics of r-systems within pre-
dtermined boundaries makes it possible to apply the
concept of homeostasis to them after the concept has
been generalized and formalized.
We will take it that the concept of homeostasis can
be extended to include the mechanisms that maintain
the constancy of certain characteristics because of
the inconstancy of others. This means that the oc-
currence of relatively stable properties in one part of
the system can be provided because of the occur-
rence of additional instability in other of its proper-
ties.
Thus homeostasis is a regulation mechanism, one
that provides for temporal ordering of change in
system properties toward stability of groups of char-
acteristics with respect to processes (r-systems), or
to their results (a-systems). In ecological systems
in particular, the occurrence of an additional zone of
inconstancy is associated with change in the relative
abundance of individual populations; that is, with
change in the specific percentage of biomasses of
species while retaining the constancy of the overall
formation rate by organisms of a homotypic associa-
tion.
The relationship
was used as the measure of the homeostasis (G) of
the living component of an ecosystem in the paper
written in collaboration with S. A. Sokolova (Fed-
orov, Sokolova, 1972). However, for reasons of
purely practical convenience associated with intro-
ducing the normalization O^G^=1, the measure of
the homeostasis is best calculated from Eq. (7) if
the value of Si is found by using Eq. (3) when using
the stability index
G= 1--
when S
and when calculating the stability estimates using
Eqs. (1) and (2)
G = l~~c~ when Sg
8
Thus, in a living system a different absolute meas-
ure of variability can be attributed to a different
variable, depending on the method selected for esti-
mating Sj.
This gives rise to a routine question. What is the
cause of the variability in individual characteristics?
There are two such causes.
1. First, there are the internal causes attributable
to the special features of the interaction of the ele-
ments comprising the system. These reflect the laws
of coordination of elements and of the interrelation-
ships between them. Included, for example, are the
variations in the numbers of two populations in a
29
-------
system whose mutual relations are of the "prey-
predator" type. The most important feature of the
changes attributable to internal causes is their fluctu-
ating nature; that is, their change with variable ampli-
tude and frequency within the limits of some region
that is the norm under the system's usual conditions
of existence.
2. Second, there are the external causes, attrib-
utable to the special features of the action of dis-
turbing environmental factors. These are capable of
bringing about deviations in system properties speci-
fied by the variables that will be beyond the limits
of the norm.
Thus, it is precisely the disturbing causes that raise
the question of the permissible measure of deviation
from the norm, within the limits of which the ecosys-
tem will be able to withstand external actions because
of its property of eliminating their consequences as a
result of the ordering state of affairs of the homeo-
stasis mechanisms. This capacity is what brings us to
the concept of stability as an integral characteristic of
systems, expressed by the capacity to withstand dis-
turbances. Consideration of stability as the mainte-
nance of some norm for the properties, attributable
in turn to the norm for the individual variables, per-
mits posing the question on the plane of establishing
a permissible measure of deviations from the norm of
certain of its properties under the effects of concrete
environmental disturbances.
Then let us define system stability as the permis-
sible (without risk of destroying the system) measure
of deviations in specified properties of a system from
the norm, caused in some measure by external dis-
turbing actions. In this definition, system stability is
established first with respect to a limited number of
selected properties, properties which we will call the
specified properties, and second, with respect to a
series of concrete disturbing environmental actions.
System stability in this case can be judged in propor-
tion to the deviations from the norm of the specified
properties caused by the disturbing measure of the
external action. This approach permits us to use the
quantitative estimates of stability to compare estimates
characterizing, on the one hand, the measure of the
disturbing actions and, on the other hand, the evoked
measure of deviations of the specified properties
from the norm. This then gives rise to the need to
establish not simply a measure of deviation of speci-
fied system properties but the permissible measure
of the deviations which, in time, can be eliminated
by the system itself. This measure of deviations elim-
inated by the system is in fact the permissible meas-
ure, and this measure must be considered the meas-
ure of stability of the entire system as a whole.
In concluding this section it seemed important to
us to turn our attention to the idea of the link be-
tween the permissible measure of deviations in the
specified properties of a system and the measure of
the homeostasis. The very idea of such a link is ex-
tremely constructive, because the measure of the
homeostasis was quite uniquely established above,
whereas the question of the permissible measure of
deviations (as distinct from the idea of a link with
the measure of the homeostasis) immediately ac-
quires acuity and the vagueness of the question con-
cerning "norm" and "pathology."
4. Invariants as indices of the stability of biosystems
A conclusion as to the instability of systems does
not follow, of course, from inconstancy of individual
variables (estimates of Si). The already generalized
interpretation of the homeostasis logically leads to
the invariant possibility. It is true, let us stipulate,
that an estimate in the form of a mean [Equations
(4) and (5)] cannot be invariant, strictly speaking,
because here we are estimating some mean instability
of individual indices which are incapable of reflecting
the more integral characteristics of system state.
However, as a very rough approximation, when the
Condition O^G=constant is satisfied, the simplest
primitive form of what we have decided to consider
in the form of an invariant turns out to be the same
mean estimate of the stability of the functional in-
dices of an ecosystem.
A more typical case of an invariant, when the
properties under consideration in the norm are gen-
erally temporally unchanged (or invariant), can be
attributed to a complex combination of consistently
varying individual variables. The definition "con-
sistently varying" implies a strict observance of the
rules for the coordinating of functions, which has
been achieved using the autonomic nervous system in
systems with central control, as a result of which it is
easier to find invariants for a-systems than for r-
systems.
We thereby arrive at a more integral interpreta-
tion of stability (although less constructive) as the
temporal maintenance of invariant properties, thanks
to the combined, ordered cooperation of continually
changing individual system variables.
Let us define the conditions needed for the mani-
festation of invariant properties in biosystems. There
are two such conditions:
(1) the closed nature of the cycle of the elements
of a stagnant environment in ecosystems; (2) the
constancy of the instantaneous rate of change in
system entropy.
Then, in accordance with these conditions, the
invariants of ecological systems can be:
(a) the constancy of the ratios between the syn-
thesis and decay reactions of organic matter in the
system (in units of energy);
(b) the constancy of the ratios between the rates
of movement of individual elements through the links
of the trophic chain, or even of the rate of individual
elements;
30
-------
(c) the constancy of the ratios of negative en-
tropy, forming at different trophic levels in the
system.
However, the ecosystem invariants cited and pre-
dicted on the basis of general considerations are
difficult to use in practice because the determination
of their values requires a great deal of preliminary
experimental work, and both necessary conditions
are almost never satisfied, or are not strictly satisfied.
The concrete generalized indices that can be meas-
ured can be (a) within the limits of a single trophic
link, the BR* index (Fedorov, 1972), where B is
the biomass, and R is the rate of its formation, or
(b) within the limits of neighboring trophic links,
the BpBv index, which reflects the interaction between
predator (BP) and prey Bv).
Equations based on BR* and BPBV estimates can
be derived and the following biologically reasonable
assumptions can be useful as the initial premises:
JV
= L
system variants, postulating that the complication of
the variants, justified by logical premises, will bring
them asymptotically close to the theoretically possible
invariants of an ecosystem. A logically justified ex-
ample of just such a complication of variants can be
the indices BR*D and BR*D/lg W, where D is the
information index of diversity (MacArthur, 1955)
and W is the number of species.
5. Measuring ecosystem stability
Once the basic concepts included in the definition
of ecosystem stability have been denned and con-
cretized, we can move on to discussion of methods
of estimating its total stability. Diverse, and numer-
ous, methods for this kind of estimate can be de-
vised, of course. Several estimating methods that use
the so-called coefficient of stability, which we shall
designate x, are suggested below.
a. The canonical method. If we also attribute the
possibility of deviation from the norm to the dis-
turbing source, then the measure of the disturbing
action turns out to be precisely the measure of the
deviation of an independent variable from the norm,
that is,
B. = const
(10)
1*1 n -1
(14)
RVB1-RVB2
R*2 • B2 R*3 • B3
= const
(11)
k-H
= const (12)
BP-B;
B* • B?
Bk-l.
_ p
c-1
Bj'B^
= const
(13)
What must be emphasized is that the BR* and
BpBv indices are not system invariants in the full
sense of the word. We can only state that they reflect,
in implicit form, the movement of energy in the
system, and therefore can be used. It therefore is
more correct to call them (as well as the biomass and
productivity indices which form them, incidentally)
Then the result of the action, or the system's re-
sponse according to the action in terms of the speci-
fied properties, Ax, may be expressed by the relative
deviation of a dependent variable from the norm,
that is,
'I.-W.
Ay
(15)
The ratio of these magnitudes can be called the
coefficient of system stability
X =
Ax/Ay
yn
(16)
In the ecosystem "norm," x will equal O when the
"norm" of the action of the external factors is pre-
served, as well as when the system's invariant prop-
erties are maintained by the mechanisms of homeo-
stasis.
b. The selective method of estimating the coeffi-
cient of stability is based on the taking into account
of the inhomogeneities in the distribution of the esti-
mates of the actuating source and the response regis-
tered by the system [see Eq. (3)].
Here it is assumed that the index of stability S
KAS)
of the variables being measured and the system
31
-------
properties under consideration (variant) is a suitable
measure for estimating inhomogeneity in the sets of
values for the individual variables being compared.
In other words, an implicit identity between the
method of estimate Ax and the expression
(1 ~S ) is assumed. This method avoids the need
3
for tedious work in estimating norms and measures
for permissible deviations in an ecosystem. Then
x =
(17)
' S
(A3)] act
where S,
, is the index of inhomogeneity in the
distribution of the estimates of variables, as calcu-
lated using Eq. (3). It is important to note that the
estimate for one of the variants to which we attribute
decisive importance when determining system state
(for example, the estimate in terms of BR*D), serves
as the system response in this case. At the same time,
the estimate of the stability of only one of the inde-
pendent variables, to whose action we attribute
system response, also can be used as the disturbing
source.
c. A variant of the selective method is the estimate
of the coefficient of stability in terms of the mean
values of the set of estimates, Si, equated to the dis-
turbing sources and to "system responses"
1 -
'i (res)
(18)
Sj(act)
Let us call this variant the exploratory method, one
used in studying the combined action of disturbing
sources on an ecosystem.
d. The three means method uses the mean values
of the stabilities, determined by any of the methods
relating to the estimates of the disturbing source
(Sd) and to system state, Si and S., to estimate the
coefficient of stability,
X= — (19)
i • o/. \
(A 3)
If the equivalance of 1—S and Sd is assumed,
a variant of the three means method is the estimate
(20)
6. Conclusion
In discussing the representation of stability that
has been developed, let us turn our attention to what
are, in our opinion, several pertinent features that
give our constructions a concrete nature.
I. The first of them concerns the division of the
characteristics of an ecosystem into functional and
structural, so that a quantitative measure of the
homeostasis (G), can be introduced.
II. The same variable, but used in different re-
cording methods, can simultaneously be the basis for
obtaining the characteristics of the disturbing source
(for example, an increase in the total reserve of
nitrogen in the system, AN, leading to eutrophica-
tion), a structural characteristic (the content of dif-
ferent forms of nitrogen at the current moment in
time), a functional characteristic (the rate of change
in the nitrogen content in the environment), and the
invariant characteristic (the rate of nitrogen circula-
tion in the system). This has made it possible to
clearly classify the data obtained according to com-
pleteness, how obtained, how processed, and research
goals, which is a cardinal point in determining the
mean values of the sets of estimates of variables ob-
tained (extra-system, structural, functional, com-
bined-variants, invariants).
III. Of principle importance is the assumption
that when G = constant^O, all estimates of system
variables (AyO are within the limits of the norm,
so that the concept of system "norm" in terms of the
norm for the specified properties was concretized,
and the region of the norm was determined for the
external and internal factors (y± AyO, as well as for
the disturbing environmental actions (x±Axi).
IV. It can be postulated that when G-»O, when
the condition O
-------
BIBLIOGRAPHY
Fedorov, V. D., 1970a, "Special features of the organiza-
tion of biological systems and the hypothesis of an 'out-
break' by a species in a community," Vestnik MGU,
Biological Series, No. 2, pp. 71-81.
Fedorov, V. D., 1970b, "The biotic variety of a phytoplanlc-
ton community and its production characteristics," Bio-
logicheskiye nauki, No. 2, pp. 7-15.
Fedorov, V. D., 1972, 'The problem of the complex in
biology and special features of its resolution," Vestnik
MGU. Biological Series, No. 6, pp. 30-42.
Fedorov, V. D., Sokplova, S. A., 1972, 'The stability of a
plankton community and some characteristics of the en-
vironment," Okeanologiya, No. 6, pp. 1057-1065.
Fedorov, V.D., 1973, "A new index of inbomogeneity in
the structure of a community," Vestnik MGU, Biological
Series, No. 2, pp. 94-96.
Cannon, W. B., The Wisdom of the Body, 1932, London.
MacArthur, R. H., "Fluctuation of animal population, and
a measure of community stability," Ecology, No. 36,
1955, pp. 533-636.
Milsum, H, H., "The nature of living control ssytenw,"
Engng. /., Vol. 47, No. 5, 1964, pp. 34-38.
33
-------
Development of Maximum Permissible Environmental
Loading—Air
B. J. Steigerwald *
I. INTRODUCTION
Most of the discussion thus far in the Symposium
has focused on the development and selection of air
quality goals or standards. Certainly this is a critical
decision. In selecting standards we determine the
severity of the pollution effect as well as those whom
we will protect. Equally important and often more
practical are the decisions relative to how we will
meet the air quality goal or standard. The conse-
quences of these decisions are often immediate and
are very visible and understandable to the public. Es-
sentially, the regulatory strategy or air pollution con-
trol scheme that is selected determines the distribu-
tion of the burdens of control, and their magnitude.
(Tlearly, the regulatory strategy that is selected or
the maximum permissible environmental loading that
is permitted should be the most cost/effective pro-
gram for the attainment of the air quality goal. In
addition we should consider pollution control as ad-
ditional costs of doing business. Control costs should
be paid by the consumer who will benefit from the
cleaner air. For example, a community whose control
regulations force location of power plants into rural
or distant regions should pay a high enough rate for
electricity to offset any environmental effects in the
area where the plant is located. These may include
air pollution, water pollution or effects of strip
mining.
Although the challenge is simple and clear—mini-
mize the ratio of costs to effectiveness and distribute
burdens fairly—there are many practical constraints
in the development of the optimum regulatory
strategy. These limitations usually dominate the situa-
tion and often preclude the full use of the available
analytical tools. Compromises must always be made
to account for lack of data on air quality, meteor-
ology, location of receptors and emissions. Political
factors also must be considered and these are not
necessarily compatible with technical/economic
optimization. Despite these limitations the art or
science of selecting a regulatory strategy is probably
on a par with our ability to set optimum environ-
mental goals. Also, it must always be kept in mind
that air pollution is only one small part of a much
broader system which should be optimized. Other
considerations should include energy supply, trans-
portation, public health care, general economic well-
being, recreation, and environmental factors such as
garbage disposal and water pollution. Most often
these factors cannot be included quantitatively in
decision models of air pollution regulatory strategy.
Often major compromises must be made with limited
quantitative analysis.
H. FACTORS TO BE CONSIDERED
IN ANALYSIS OF PREFERRED
REGULATORY PATHWAY
For purposes of the current discussion, we will
limit ourselves to the rather narrow problem of de-
veloping the best possible air pollution control
scheme or set of regulations to meet a well stated air
quality goal or standard. Even within this narrow
perspective, there are a wide variety of factors that
must be considered and analyzed.
Institutional Factors
Organization of government and the availability of
resources often direct the general form of the control
strategy or limit the scope of the analyses. These
include: form of government-socialistic or capital-
include: form of government—socialistic or capital-
strong local control programs; statutory authority—
freedom provided in enabling legislation, mandatory
regulations or exemptions in basic statute; technical
sophistication of the nation or local governments de-
veloping regulations; resources available for surveil-
lance and enforcement; and extent of public support
for air pollution control programs.
Nature of Pollutants
The effect of the pollutant being considered and
the severity of the effect must influence the nature of
the control program. The effects spectrum that must
•Director, Office of Air Quality, U.S. Environmental Protection Agency
35
-------
be considered is continuous. The spectrum starts at
public health and ranges through material damages
to esthetics and on to nuisance effects such as visi-
bility and odors.
Where the potential effect is on health, affects a
large segment of the population, and is serious, ir-
reversible, and acute, a stringent control strategy
must be applied. This would include both the severity
of the control measures and their time of application.
Examples are beryllium, asbestos, and mercury, three
pollutants recently classified as "hazardous" in the
United States. Source control for hazardous pol-
lutants is under the very restrictive Section 112 of
the Clean Air Act. Section 112 dictates nationwide
emission standards, compliance from 90 days to two
years, and extensive monitoring and surveillance.
Pollutants such as fluoride whose effect is serious,
but not acute and which may not affect large seg-
ments of the exposed population, may allow for
slightly less priority on the time of application of con-
trol measures. Therefore, more reasonable, longer
term control options may be possible.
Some effects may call for no control. Recently in
the United States, we determined that odor control
would not be a nationally manated effort. Odor is
deemed a local nuisance and control should proceed
at a rate determined by local needs and desires.
Selecting an Air Quality Strategy
The scientific character of the pollutant is possibly
the most critical factor in determining the nature of
the control strategy. If there are multiple exposure
routes, air pollution control strategy must be inte-
grated into the broader effort and put into perspec-
tive with timing and intensity.
Many pollutants such as CO, SO2 and total par-
ticulates are emitted directly. Their source/receptor
relationship tends to be uncomplicated; air concen-
trations can usually be related to sources in the same
urban area. Several pollutants that are becoming
very important, however, are formed in the atmo-
sphere often after long transport distances. Atmo-
speric chemistry complicates greatly the develop-
ment of an effective, equitable control strategy. Oxi-
dants, for example, are formed by the reaction in the
atmosphere of certain organic compounds, oxides of
nitrogen and sunlight. The reaction is complex and
can occur within minutes or over several days. Re-
cent data show that this is not strictly an urban
problem. Concentrations of oxidant greatly exceed-
ing the air quality standard of 0.08 ppm one-hour
average are occurring in many rural areas of the
United States 100 or more miles from major cities.
Reports include 0.15 ppm in Ft. McHenry, Mary-
land, which is located 120 miles from Washington,
D.C.; 0.13 ppm in Kane, Pennsylvania, which is
over 150 miles from Cleveland, Ohio; and 0.5 ppm in
Indio, California, located over 120 miles from Los
Angeles. Sulfate particulate matter, including sulfuric
acid, also seems to be formed in the atmosphere as
the result of the oxidation of SO2 gas; only 2% is
particulate sulfate. The situation is very complex.
Current information in this case is not yet adequate
to allow the selection of a workable national control
scheme for meeting a sulfate air quality standard.
The sources of a pollutant, their distribution and
future growth patterns are also key determinants of
the control strategy. To effect significant control,
large industrial sources generally install control
equipment. Area sources such as home heating or
automobiles often require basic changes in fuel
supplies, land use patterns or life style of the popula-
tion. In dry, windy climates such as the South-
western United States, wind-blown dust, including
that from farming and unpaved roads, can be the
major source of particulate matter and may preclude
meeting the air quality goals with any air pollution
strategy.
A formal analysis of alternative control strategies
that incorporates many of these considerations re-
cently has been done in the United States for
NO2.( 1) The analysis resulted in a significant change
in the national NO2 control program. A short sum-
mary of the study might be illustrative of the value
of a preferred control strategy analysis.
After some initial confusion related to the selec-
tion of measurement methods, we determined that
existing levels of NO2 were not excessive in most
U.S. cities. The primary objective of the NO2 analy-
sis was to determine an equitable and efficient air
pollution control strategy that would maintain the
annual NO2 standard of 100 /*g/m3 despite the rapid
growth of NOx sources. A proportional model was
used to simulate air quality data at five-year intervals
out to 1990. The model used current air quality
data, six categories of NOx sources (light, medium,
and heavy duty vehicles, industrial processes, area
sources, and electric power plants), five sets of
growth rates for each source category, two levels of
stationary source control, and five NOx automotive
emission standards. A proportional relationship was
assumed between NOx emissions and NO2 concen-
trations. Annual air quality estimates and control
costs were calculated for each strategy. Fuel penalties
were also calculated for each level of mobile source
control.*
The NOx analysis summary report recommended
that the 1977 automotive standard of 0.4 grams/
mile be revised to 2.0 grams/mile. In addition the
report recommended that more stringent emission
control be placed on new and existing stationary
•See Attachments: Attachment 1 describes NO. Analysis; Attachment 2 illustrates a range of estimated air quality and control costs.
36
-------
sources in regions with high NO2 concentrations.
The recommendation was made after evaluating
ranges of air quality, and costs that included fuel
penalties. As a result of the analysis, adjustments
are now being made to the national control program
for NO2 in the United States. Table 1 illustrates an
example comparison of two pollution control strat-
egies for New York City. Several other strategies not
shown were also compared.
m. GENERAL TYPES OF
REGULATORY STRATEGIES
There are two general approaches to developing
an air pollution control strategy. The first involves
simply the widespread application of reasonably
available control technology. The second involves
an air quality management program, usually over a
relatively small area such as a city, to balance emis-
sion control with air quality needs.
Control Technology
The control technology approach is the more tra-
ditional strategy. A general problem is identified and
available emission control technology is applied to
control the problem. There is little attempt to vary
regulations to match air quality levels or to develop
long-range plans to maintain the air quality standard.
Formal analysis of cost-effectiveness and cost-benefit
of alternative strategies is not emphasized in this
approach. However, control is technically simple and
enforceable. Sources generally know unambiguously
their control requirements and can have confidence
that they will not change continuously. The initial
successes achieved in St. Louis, Pittsburgh and Lon-
don were based on this form of control strategy.
In the Clean Air Act of 1970, the U.S. Congress
generally limited the control technology approach to
new sources. Section 111 of the Act, New Source
Performance Standards, requires all new major
sources of pollution to meet national emission stan-
dards set by the U.S. Environmental Protection
Agency. These standards are based on application
of best demonstrated control technology and take
into account the cost of such control systems. Na-
tional emission standards are set for the lead content
of gasoline. Standards are also set for new auto-
mobiles, trucks, and aircraft. Standards are also
established for selected sources of hazardous pol-
lutants. These are based either on best technology
or on the degree of control needed in the most critical
situation in the country.**
The control technology approach also is being
used increasingly as a compromise solution when air
quality management programs cannot be upheld be-
cause of lack of data or inadequate analytical tools.
Sulfur oxides and particulates are examples, espe-
cially in the vicinity of large isolated sources such
as power plants in very mountainous terrain where
diffusion calculations are unreliable. The control of
oxidant in rural locations is another example.
Air Quality Management
Theoretically, the benefits of the air quality man-
agement (AQM) approach are great. Control is re-
quired only where needed and only to the extent
needed. Usually these benefits cannot be fully real-
ized because of practical limitations, but even if the
AQM approach is eventually discarded as the basis
for regulations, the AQM analysis will provide val-
uable experience in setting enforcement priorities
and in directing research and data collection pro-
grams.
Air Quality Management is required by the U.S.
Clean Air Act as the principal approach for the at-
tainment and maintenance of the national air quality
standards. The entire country has been divided into
247 Air Quality Control Regions (AQCR). Control
region designations are based on considerations such
as meteorology, population density and political
boundaries. Each state must submit a State Imple-
mentation Plan (SIP) that outlines and justifies a
regulatory scheme to attain and maintain all of the
air quality standards in every region or part of
Region within the state. The SIP's are very complex,
Table 1
Comparing NOx Control Strategies (New York)
Strategy
Automotive standard of
0.4 grams/mile
&
Moderate Stationary
Source Control
Automotive standard of
2.0 grams/mile and
Maximum Stationary
Source Control
Annual
Air Quality (ug/m*)
1977 1980 1985 1990
90 90 86 96
91 78 77 87
Annual
Control Cost (10° dollars)
1977 1980 1985 1990
144.8 276.0 457.0 510.1
128.7 194.7 320.6 377.4
Annual
Fuel Penalty (10° dollars)
1977 1980 1985 1990
67.6 134.9 212.2 227.9
60.5 92.8 132.1 136.7
••See Attachments: Attachment 3 lists New Source Performance Standard categories and pollutants; Attachment 4 lists dates for controlling
Mobile Source Emissions; Attachment 5 illustrates Pollutant and Sources of Hazardous Emission Control.
37
-------
but each must include information on current air
quality, an emission inventory, regulations that will
reduce current emissions enough to achieve the air
quality standard and evidence of enforcement capa-
bility.
The development of a complete AQM program
demands great technical sophistication, good under-
standing of the pollutant and much data. A full dis-
cussion of the AQM is beyond the scope of this
paper. It seems appropriate, however, to delineate
some of the data needed for a full analysis and to
describe the status of AQM techniques in the United
States.
Information on air quality is vital to the AQM
approach. The data must be temporally and spa-
tially compatible with the air quality standard, i.e.,
hourly, daily or annual measurements and available
for key receptor sites and in locations of expected
maximum concentrations. Data must be available for
enough years to ensure that severe conditions of
meteorology and emissions are included. Natural
background (wind-blown dust, stratospheric ozone,
organics from vegetation) contributions to urban
levels must also be determined and included in the
analysis. Finally, since the AQM is so quantitative
and so dependent on air quality, standard methods
of collection and analysis must be developed with
known precision and accuracy and the size and sit-
ing of monitoring networks must be carefully de-
fined. In the United States there are approximately
4000 air sampling stations, most using standard or
equivalent methods and operated by the State or
local governments. The data are sent four times each
year to a national data system called SAROAD
(Storage and Retrieval of Air Data) for organiza-
tion and evaluation.
Source emission data are the next vital part of the
air quality management analysis. The emission in-
ventory should include the magnitude and location
of all important sources of the pollutant being eval-
uated. Large sources should be specified individually
along with information on stack height, existing con-
trols, plant age and economic viability. Area sources
such as home heating or automobile emissions should
be disaggregated into geographical areas compatible
with the diffusion model being used. In the United
States, this may range from several square kilometers
when sophisticated diffusion models are being used
to full counties (from a thousand to 30,000 square
kilometers) where the simple proportional relation-
ship between emissions and air quality is used. In the
United States, approximately seventy percent of the
100,000 large point sources (over 100 tons of emis-
sions per year) and all of the 3300 counties (as area
sources) are included in the National Emission Data
System (NEDS).
Data on the type and distribution of receptors that
may be affected by air pollution may also be neces-
sary in some types of air quality management analy-
sis. The need to include receptors depends upon the
type of air quality standard that has been set. Cur-
rently in the United States the health related standard
must be met in all geographical areas by a date set in
the basic legislation (1975 or 1977). Legally, the air
quality standard basis for the control regulations and
the latest time of application are uniform throughout
the country. In practice, there is a priority set on
analysis, monitoring, surveillance, and enforcement
that favors urban areas or other areas where sensitive
receptors are concentrated. Even if allowed by the
Act, there are no usable damage functions that re-
late the magnitude of health effects to changes in
concentrations. Damage functions, therefore, cannot
be included in any air quality management models.
An optimal control strategy cannot be selected
without detailed information on control systems. For
each major source category, information for making
air pollution control decisions must include effective-
ness, availability (including timing) cost and social
impact. Data is fairly simple to collect for control
systems that involve stack gas cleaning equipment
such as precipitators or filters. When pollution con-
trol is achieved through a change in the manufac-
turing process or fuel substitution, we find more
difficulty in assigning a cost to pollution control.
Some of the more recent control programs involve
changes to the transportation system in a city or
land use planning which have a social impact that
cannot always be stated in simple economic terms.
Cost comparisons with more traditional pollution
control equipment may not be possible.
The heart of an AQM analysis is the source/
receptor model that relates emissions to air quality.
In its simplest form this may be a proportional roll-
back model that assumes, for example, that a 40%
reduction in total SO2 emissions for the city will re-
sult in a 40% decrease in air concentrations. Unlike
more sophisticated diffusion models, the proportional
model does not account for source distribution,
height of emission or meterological conditions. (2)
In general, diffusion models take into considera-
tion the pollution diluting capabilities of the environ-
ment (see Table 2). The more dilution that takes
place because of wind, stack heights, or distance
between sources, the lower the ambient air concen-
trations become. The "box" model, "gaussian plum"
models, and the CO model are examples of more
advanced diffusion models. They allow one to pre-
dict the spatial distribution of various concentrations
of pollutants for a given day, or over some long time
period, which is not possible by proportional model-
ing.
A current limitation in meterological diffusion
models is an inability to include the complex atmo-
spheric reactions that both form and destroy pol-
lutants. This is most serious for photochemical sys-
tems involving hydrocarbon, oxides of nitrogen and
oxidants. Extensive research is underway and an
38
-------
Table 2
Diffusion Models
Box Model
Q
C,=-
uhd
C,=-
Gaussian Model
Q
CO Model
Cw=Crexp[(0.55)(l-z/zr)]
C,=Cr exp [(0.87+0.98UPK1-Z/Z,)]
C,=ground level concentration, Cr=roof top concentration, Ci=leeward side of building concentration
Cv=wind side of building concentration (g m-1)
Q =source emissions (g sec'1)
u =average wind speed, ur= roof top wind speed (m sec-1)
h =mixing height (m)
d =\b [urban area] ** (m)
y =cross wind distance from plume axis (m)
-------
150
100
=
si
50
PROJECTED AIR QUALITY
MEASURED AIR QUALITY
1970 1971 1972 1973 1974 1965 1976 1977
CALENDAR YEAR
FIGURE 1. PLAN REVISION MANAGEMENT SYSTEM
IV. MAINTENANCE OF AIR
QUALITY STANDARDS
Programs to maintain air quality standards gen-
erally allow for much more freedom of action and
imaginative solutions than control schemes to roll-
back pollution to attain the standards. Since the
emphasis is on new sources, evaluations can be made
of the effects of city planning, new mass transit sys-
tems, effective solid waste disposal, industrial siting
practices, economic incentives and emerging control
technology. Theoretically, programs to maintain
standards should accommodate themselves fully to
balancing air quality with economic and population
growth. However, there is little experience with such
programs and no working analysis techniques that
have widespread applicability. Since these are future
programs, there is still time for research to make
significant contributions.
Efforts to maintain air quality in the United States
are a complex program of activities, many of them
required by the Clean Air Act. Current state emis-
sion regulations must be restrictive enough to achieve
the standards in the most polluted areas; therefore,
where regulations apply state-wide, there will be
excess control or "overkill" in all other areas. This
"overkill" will allow much growth of new sources to
be absorbed without violation of standards. For ex-
ample, SO2 controls needed in Philadelphia will re-
duce SO2 levels in surrounding counties by about
40% (15 to 25 /ig/m3) even though they are now
well below the standard. Special emission regula-
tions for new sources are a major part of programs
to maintain the standard. The Clean Air Act and
EPA regulations require nationwide emission stan-
dards for many new sources including automobiles
and trucks, gasoline additives, aircraft and all major
industrial processes. These help ensure minimum air
quality impact of future economic and industrial
growth. In addition all new industrial and combus-
tion sources must obtain a pre-construction approval
which can be given only if it is determined that the
air quality standard will not be violated in the vicinity
of the plant. Recently these regulations have been
expanded to include a pre-construction review of
airports, highways, shopping centers, amusement cen-
ters, and other "indirect" sources that attract enough
automobile traffic to threaten the air quality stan-
dard. Finally, a comprehensive analysis is being made
of needed maintenance programs for the more pop-
ulous Air Quality Control Regions in the country.
These include projections to 1985 for emissions and
air quality and development of appropriate regula-
tions to ensure maintenance of the standards, includ-
ing consideration of land use and transportation.
40
-------
Table 3
Implementation Planning Progn
PURPOSES: (a) Relate emissions of sulfur dioxide and
paniculate matter to ambient annual con-
centrations of the two pollutants
(b) Test regional control strategies of cost-
effectiveness in reducing ambient pol-
lutant concentrations
Source Data
Management Segment
Purpose: Create and
maintain emissions in-
ventory data for use
with other modules
Air Pollution
Diffusion Segment
Purpose: Calculate
pollutant contribu-
tions from individual
sources to preselected
receptors
PROGRAM MODULES
3
Control Cost Segment
Purpose: Calculate
emission reductions &
costs associated with
applications of vari-
ous control techniques
Emission Standards
Segment
Purpose: Determine
for each source the
most cost-effective
control technique
needed to achieve in-
put emission stds.
Regional Strategies
Segment
Purpose: Estimate
cost of control and
ambient air quality
associated with imple-
mentation of a re-
gional control strategy
FUTURE IMPROVEMENTS
1) Update pre-set control technique information to reflect
current control cost and technology
2) Include, in the Air Pollution Diffusion Segment, the
Samples Chronological Input Model which has the
capability of calculating short term concentrations
-------
V. SUMMARY
This paper has attempted to present some of the
conceptual considerations that we feel are important
in developing optimum control strategies to achieve
and maintain air quality standards. Strategy develop-
ment is a complex, very practical, and vital part of
the total field of air pollution control. The control
strategy and resulting regulations determine not only
who will control but also the severity of the neces-
sary control measures. No attempt has been made to
discuss the detailed techniques of air quality manage-
ment analysis. Entire books have been written on
many of them.
Currently in the United States, the concepts of air
quality management are used primarily to guide the
development of control strategies. A full quantitative
analysis is seldom the determining factor in setting
regulations because of institutional and political con-
straints and data limitations. Despite many of the
unquantifiable factors associated with air quality
management, improvements in models such as IPP
and development of new techniques similar to the
Plan Review Management System will possibly pro-
vide more objectivity in the selection of future con-
trol strategies.
BIBLIOGRAPHY
1. NOZ Analysis Report, Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency,
September 21, 1973.
2. Proceedings of Symposium on Multiple-Source Urban
Diffusion Models, U.S. Environmental Protection
Agency, 1970.
3. Photochemical Smog, Proceedings of International Sym-
posium on Air Pollution, October 17, 1972, Tokyo.
4. Air Quality Implementation Planning Program, Vol-
ume 1, National Technical Information System—PB.
198299.
5. A Users Manual for the Sampled Chronological Input
Model, GEOMET, Inc., Contract No. 68-02-0281,
1973.
6, Validation and Sensitivity Analysis of the Gaussian
Plume Multiple-Source Urban Diffusion Model,
GEOMET, Inc., Contract No. CPA 70-94, November
1971.
42
-------
Attachment 1
Description of NO, Analysis
• COMPUTER MODEL
EMISSIONS OF NO, AND CONCENTRATIONS
OF NO.
• SOURCES OF NO, EMISSIONS
-Light Duty Vehicles
-Medium Duty Vehicles
-Heavy Duty Vehicles
—Electric Power Plants
-Industrial Plants
-Area Sources
• GROWTH FACTORS
• AUTOMOBILE EMISSION STANDARDS
• STATIONARY SOURCE CONTROL
• COSTS
Attachment 2
Ranges of NO* Air Quality Projections and Costs
Air Quality (ug/m§)
Costs (Millions $)
City
Phoenix
Los Angeles
San Francisco
Denver
New York
Philadelphia
Washington, D.C.
Chicago
Baltimore
Salt Lake City
1975
80
9.4-9.6
181
64.8-66.1
86
28.6-29.2
78
8.4-8.6
101
75.3-76.8
88
30.2-30.8
94
15.5-15.8
120
38.6-39.4
107
12,3-12.6
64
5.4-5.5
1980
58-71
16.5-39.3
97-132
103.4-244.3
67-75
45.6-108.8
72-86
13.4-45.0
74-97
119.0-274.2
74-97
47.2-113.0
69-103
24.2-60.1
97-117
58.7-148.0
84-102
22.7-57.4
54-60
9.6-22.4
1985
56-78
26.0-69.9
94-153
136.9-396.0
68-87
62.4-177.5
81-101
18.0-68.2
68-99
168.1-478.6
82-109
61.8-183.2
64-102
34.6-103.4
93-127
83.4-245.2
78-107
39.6-118.5
45-60
14.5-40.1
1990
61-80
30.3-85.0
104-174
142.3-447.7
80-101
65.2-209.9
95-120
18.8-79.0
74-112
176.4-562.3
95-127
63.1-208.7
72-120
37.0-121.5
103-146
87.1-279.5
89-132
50.9-164.9
48-67
16.8-49.5
43
-------
Attachment 3
New Source Performance Standard
Categories and Pollutants
Steam Generators SO* NOi, TSP
Municipal Incinerators TSP
Portland Cement Plants TSP
Nitric Acid Plants NOi
Sulfuric Acid Plants SOi
Asphalt Concrete Plants TSP
Petroleum Refineries SO* HC
Secondary Lead Smelters & Refineries TSP
Petroleum Storage HC
Secondary Brass & Bronze Refining TSP
Iron and Steel Mills TSP
Sewage Treatment Plants TSP
Attachment 4
Dates for Controlling Mobile Source Emissions
Automobiles
Light Trucks
Heavy Trucks
Aircraft
Motorcycles
NO,
HC
CO
Particulates*
1976
1975
1975
NA
1975
1975
1975
NA
1974
1974
1974
NA
1981
1981
1981
1979
1976
1976
1976
NA
*NA=Not Available
Attachment 5
Hazardous Emission Control—1973
Beryllium Plants
Asbestos
Mercury
Ex traction'Plants
Foundries
Ceramic Manufacturing Plants
Machine Shops
Mining & Milling
Manufacturing
Fabrication
Demolition
Spraying
Mercury Ore Processing Facilities
Mercury Cell Chlor-Alkali Plants
44
-------
Estimating the Maximum Permissible Intensity of the Complex
Effect of Chemical Factors in the Production,Municipal, and
Domestic Environment on Man
N. F. Izmerov *
Investigations of the harmful effect of chemical
compounds on living organisms when the toxins
enter primarily from some one environment (air or
water, food, and so forth) have now taken shape.
Approaches to the hygienic limitation of pollution
of individual objects in the habitat have been de-
veloped in this plan. At the same time, we know
that the same harmful substances can enter the body
simultaneously from the air in production areas and
in living quarters, from the atmosphere of populated
areas, from water and food, from clothing and shoes,
and so forth.
And it has been established that emissions from
the production of aluminum, cryolite, and super-
phosphates settling on soil and water surfaces create
artificial geochemical provinces with increased trans-
fer of fluorine to plants and animals ecologically as-
sociated with man. There are individual microregions
1.5 to 2 km from a cryolite plant in which the fluorine
content in the air was found to measure 0.46-0.85
mg/m3, exceeding the maximum permissible concen-
tration (MPC) for air by a factor of 40-80. The
fluorine content in the soil 4 km from the plant ex-
ceeded the natural content by a factor of 1.5-27.5.
The harmful effect of plant emissions on trees, as
well as on the health of animals and man, has been
noted (Sadilova, M.S., 1964). The results of an in-
vestigation of individual groups in the population liv-
ing around aluminum and cement-slate plants where
elevated fluorine content is present in various environ-
mental objects revealed fluorosis in varying degrees
of seriousness in children and in the adult population
(Audera, A. K., 1966), as well as change in the
thyroid gland resembling enthyroid goiter (Kharito-
nova, V. A., Perfil'yeva, Z. V., 1966). Natural in-
vestigations of animals (Lindberg, Z. Ya., 1965)
confirmed the prevailing role of production emissions
in the formation of similar geochemical provinces
with increased fluorine content in environmental
objects.
The population in agricultural areas absorbs a big
part of the pesticides in the course of everyday living
and not from the air in the work area, while working
in agriculture. The specific content of DDT in food
products varies from 92 to 95%, in water 5%, and
in air 3%, of the total dose absorbed. The effect of
pesticides in a work area can be characterized by
its transitory nature, and by its periodicity (L. I.
Medved', et al., 1971). The complex nature of the
effect on man is characteristic of a great many chem-
ical compounds. One of the most widely propagated
production toxins, for example, is carbon monoxide,
yet is equally a domestic toxin. The products of the
incomplete combustion of natural and artificial gas,
and of other types of fuel in buildings, vehicle ex-
haust gases, and emissions from municipal and do-
mestic enterprises, are the primary sources of air
pollution, along with emissions from metallurgical
industry plants, and the like. Signs of the effect of
carbon monoxide (increased blood level of carboxy-
hemoglobin) have been detected in 78% of those
examined and living in apartments supplied with gas
in individual microregions (Martynyuk, V. Z., Devy-
atko, D. G., 1956).
Hydrocarbons affect the inhabitants of oil drilling
areas during production activities, as well as when
production is not taking place. Changes in the cardio-
vascular system (lability of arterial pressure, changes
in the vegetative regulation of vascular tonus) have
been found in workers in refineries refining high-
sulfur crude oil, as well as among the population
living in the surrounding area (Zagidullin, Z. Sh.,
1966).
However, it should be pointed out that some sub-
stances (carbon disulfide, benzene, carbon tetrachlo-
ride, and others, in particular) are primarily indus-
trial toxins, whereas others, natural food and fungous
toxins, for example, are primarily domestic toxins.
A toxicological estimate of the complex effect of
toxins on the body (upon simultaneous intake from
•Director, Order of the Red Banner of Labor Scientific-Research Institute of Industrial Hygiene and Occupational Dis«as«s of the Academy
at Medical Sciences of the USSR
45
-------
different environmental objects) cannot be identified
with the combined effect of several toxins, and pre-
sents very definite difficulties.
The thrust of the attention given to resolving ques-
tions of setting complex hygienic norms for chemical
compounds naturally is concerned with threshold and
harmless levels of toxin effect on the organism upon
isolated admission. Experimental research in par-
ticular has shown that when there is simultaneous
admission of chemical substances (12 nonelectro-
lytes) through the gastrointestinal tract and the res-
piratory organs at the MFC level or at the threshold
of chronic effect, there is summation of the effects
(S. M. Pavlenko, 1973). Ye. I. Spynu (1967, 1968)
has calculated that the summed quantity of pesticides
entering the body from different environments at the
MFC level exceeds the harmless dose. He therefore
proposes establishing a harmless maximum permis-
sible dose of a drug for man, obtained by totaling
the maximum permissible doses and concentrations
in all environments, with consideration given to their
specific importance:
4
= 2 D
where
Dm is the maximum permissible harmless dose;
Di is the dose of pesticide in a food area ;
D2 is the dose of pesticide entering with the inspired air in
the work area;
D3 is the dose of pesticide entering with the water;
Di is the dose of pesticide entering with the inspired at-
mospheric air.
Inherent in the realization of his proposals is the
difficulty that arises because of the need to determine
the specific importance of the MFC in the different
environmental objects in order to calculate the maxi-
mum harmless dose.
Also proposed for calculating the maximum harm-
less dose for simultaneous admission of a single sub-
stance into the body from different environments,
is the use of the principle of setting hygienic norms
as applicable to the combined effect of several chem-
ical compounds, the effect of which is unidirectional
in nature and based on the additive effect (A. I.
Korbakova et al., 1971). The use of a known formula
for the summed effect has been suggested for these
purposes:
From prod.
MFC for prod
From atmosphere
MFC for atmosphere
From water
MFC for water
Dose from food
POC for food
The availability of norms for many pesticides, and
for certain other substances, in the air in work areas,
in the atmosphere, in reservoirs, and in food prod-
ucts, as well as the use of formulas for calculating
approximate levels, today enables us to obtain some
idea as to the summed content of substances in the
environments, and to recommend permissible quanti-
ties in each of them. At the same time, the differences
in the approaches taken to justify the MFC in dif-
ferent environments make it difficult to use the prin-
ciple described. We know, for example, that only
the MFC in the air in work areas and the mean daily
MFC in the air in populated areas can be established
from the resorptive effect (the maximum single con-
centration can be established from the reflex effect),
and sometimes in terms of the toxicological criterion
for reservoir water for sanitary and domestic pur-
poses, and for food (with a very small safety factor
= 100). This behavior pattern can be, and is, the
most frequent dependency, but it fails to take into
consideration the possibility of a different effect for
the different paths over which admission can occur.
Toxicity, as well as the danger of poisoning, depends
on how certain chemical compounds enter the body.
Fluorodinitriles, for example, are only slightly toxic
when injected into the stomach (LD60 = 997-2917
mg/kg) (Korshunov, Yu. N., 1969), but are highly
toxic when inhaled and the danger of developing
acute and chronic poisoning is very high (these sub-
stances, when inhaled, are included among the highly
toxic and dangerous compounds) (LD50 = 62-67
mg/m3, Linich = 1 mg/m3).
It would appear that aliphatic nitriles injected into
the stomach undergo enzymatic conversion in the
liver and become less toxic by a factor of 100. The
neutralization of fluorodinitriles in the acirc environ-
ment of the stomach also is of definite importance.
The different nature of the effect stimulating toxins
have on the body when they enter over different paths
is important. Thus, bromine is a specific stimulating
toxin when inhaled, and the changes in the respira-
tory system are of decisive importance at all levels
of its effects, and for any duration (Ivanov, N. G.
et al., 1973). But when bromine enters the body
over other pathways, changes in the central nervous
system, in the reticular formation, in the cerebral
cortex (Krylov, O. A., 1960; Z. Ya. Dolgova, Ye. G.
Dolgov, 1964), and in the thyroid gland (Verk-
hovskaya, N. N., 1962) come to the fore. Here the
problem of fixing a complex norm is much more
complicated.
We also must pause to consider the important dif-
ferences that exist between the actions of chemical
compounds that take place under natural conditions
upon inhalation, and when they enter via the gastro-
intestinal tract. In the case of inhalation we most
often must deal with the intermitten nature of the
effect, whereas when admission is over an enteral
pathway the mode usually is monotonic. It now has
46
-------
been established that the intermittent mode of action
in some cases is more harmful to the body. The inter-
mittent action of such nonelectrolytes as toluene
(Babanov, G. P. et al., 1972), carbon tetrachloride,
methylene chloride (I. P. Ulanova, et al., 1972),
trichloroethylene (Lomonova, T. V., 1973), and a
number of pesticides (Burkatskaya, S. N., Mityu-
shina, V. I., 1968) resulted in the development of
chronic intoxication and failure of the adaptive
mechanisms more rapidly than their monotonic ac-
tion. The hygienic significance of the variations in
concentrations of harmful substances is different for
different toxins, and is determined by their proper-
ties. In particular, the solubility of substances in
water and in fats, the rate of saturation of the blood
by a substance, and of its conversion in the body,
the pathways and rate of excretion from the body,
and the ability to accumulate, all play a significant
role.
Moreover, long interruptions between the effects
of most chemical substances can result in a reduc-
tion in the toxic effect (Sidorenko, G. I., Pinigin,
M. A., 1971). At the same time, the biological effect
can be more pronounced when interruptions are
brief (N. A. Tolokontsev, 1960; Lyublina, 1973).
These facts also serve to definitely complicate the
problem of establishing complex norms for chemical
compounds.
Data that show the practical necessity for an inte-
grated approach to establishing hygienic norms for
compounds that are all present at the same time in
several of man's habitats are now at hand. Experi-
ments proving, in particular, the inadequacy of the
MPC for the fluorine in water, and ion in the atmo-
spheric air when the effect is complex (Petina, A. A.,
1967), as well as in the water, and in the air in a
work area when the effect is complex (Avilova,
G. G., et al., 1973), have been devised.
A comparatively short biological experiment (1
month) showed that in the case of complex entry into
the body of the fluorine ion with water and air at
concentration at the MPC level for individual en-
vironments, the summed dose of fluorine in the body
causes changes in many of the biochemical indices
of the carbohydrate and mineral metabolisms, as well
as in the integral indices of body condition (Avilova,
G.G., etal., 1973).
We therefore can conclude that the problem of the
complex effect of chemical compounds as a whole is
not yet solved. The need for hygienic norms for
chemical compounds entering the human body simul-
taneously from different environmental objects is an
urgent one, and requires the development of theoreti-
cal and methodological approaches that will provide
for detailed resolution of this problem. Today it ap-
parently is still too early to talk about a single norm
that will express the integral threshold of the pro-
longed effect and the integral MPC.
The biological magnitude of an integral MPC,
established using highly sensitive enzymological tests
to establish exposure, can be an important index
of the summed effect of a chemical factor for a num-
ber of substances, and this is in addition to the en-
vironmental estimate.
The presence of these substances in the environ-
ment can be limited by making a determination of
the permissible content of the products proper, or
of their metabolites, as well as by the state of the
enzyme systems in the biological media, or of the
most sensitive organs.
It should be pointed out that the establishment of
two magnitudes characterizing the maximum permis-
sible quantities of a substance entering through the
lungs and through the gastrointestinal tract must to-
day be recognized as the most realistic approach
when setting complex norms. At the same time, the
MPC for a substance entering the body via the
respiratory organs must take into consideration the
possibility of its entering the body with the air
breathed in a production area and from the atmo-
sphere.
The MPC for a substance entering through the
gastrointestinal tract should take into consideration
entrance with water and food.
Efforts such as these are particularly important
in the case of substances that have far-reaching ef-
fects (mutagenic, embryotropic, carcinogenic).
Approval, in the form of legislation, of hygienic
norms for 3,4-benzpyrene in individual environ-
ments (the air in a work area, atmospheric air, and
water) is a considerable step forward in this regard.
At the same time, it has been shown that there is
little difference between the minimum effective and
maximum nonblastomogenic doses of 3,4-benzpyrene
for any of the pathways over which entry is had into
the body. The same blastomogenic effect can be ob-
served, approximately, in the case of close summed
doses of 3,4-benzpyrene injected into animals in dif-
ferent ways (intratracheally, cutaneously, and into
the stomach), as well as when the data are con-
verted for man (Yanysheva, 1972; Shabad, L. M.,
Sanotskiy, I, V., et al., 1973).
Thus, despite the natural difficulties, an integral
estimate of the maximum permissible intensity of the
complex effect of chemical compounds on man from
different environmental objects not only is necessary,
it is possible.
BIBLIOGRAPHY
1. Audera, A. K., Vliyaniye promyshlennykh vrednostey
na sostoyaniye rotovoy polosti [The influence of harm-
ful industrial conditions on the condition of the oral
cavity], Riga, 1966, pp. 34-37.
2. Babanov, G. P., Burov, Yu. A., Skolbcy, N. A., Ver-
khovskiy, L. G., Abramyan, G. G., Tritskaya, I. L.,
Isakhanov, T. M., Toksikologiya i gigiyena produktov
neftekhimii i neftikhimicheskikh proizvodstv [The toxi-
cology and hygiene of petrochemical products and petro-
chemical production], 1972, pp. 32-45.
47
-------
3. Burkatskaya, Ye. N., Mityushina, V. I. Gigiyena pri-
meneniya, toksikologiya pestitsidov i klinika otraleniy
[Hygiene of the use and toxicology of pesticides and
the clinical aspects of poisoning], Kiev, 1968, p. 698.
4. Verkhovskaya, I. N., Brom v zhivotnom organisme i
mekhanism yego deystviya [Bromine in the animal
organism and the mechanism of its effect], Moscow,
1962.
5. Zagidullin, Z. Sh., Trudy Ufimskogo institute gigiyeny
truda i profzabolevaniy, Vol. 16, No. 2, 1965, pp.
22-26.
6. Ivanov, N. G,, Germanova, A. L., Pozdnyakov, V. S.,
Klyachkina, A. M. et al., In Book: Adaptatsiya i kom-
pensatsiya pri khimicheskikh vozdeystviyakh [Adapta-
tion and compensation in the case of chemical effects],
Moscow, 1973, pp. 75-92.
7, Kprbakova, A. I., Shumskaya, N. I., Zayeva, G. N.,
Nikitenko, T. K., Nauchnyye osnovy sovrem. metodov
gigiyenicheskogo normirovaniya khimicheskikh vesh-
chestv v okruzhayushchey srede [Scientific bases of
modern methods for setting hygienic norms for chem-
ical substances in the environment], Moscow, 1971,
pp. 35-40.
8. Korshunov, Yu. N., Toksikologiya novykh prom. khim.
veshchestv, Moscow, No. II, 1969, pp. 79-85.
9. Krylov, O.A., Fiziohgiya zhurn. im. Sechenova, Vol.
52, No. 7, 1966, pp. 906-910.
10. Lindberg, Z. Ya., Materialy 13 nauch. sessii Rizhskogo
med. institute [Materials from the 13th session of the
Riga Medical Institute], Riga, 1965, pp. 53-55.
11. Lomonova, G. V., Frolova, I. N., Gritsevskiy, N. A.,
Voprosy psikhofiziologii gigiyen. analiza trudovykh
protsessov [Questions of the psychophysiology of a
hygienic analysis of labor processes], Moscow, 1967,
pp. 57-67.
12. Lyublina, Ye. I., Kolichestvennaya toksikologiya [Quan-
titative toxicology], 1973.
13. Medved', L. I., Spynu, Ye. I.. Burvy, V. S.. Belonozhko,
G. A., Antonovich, Ye. A., Vrochinskiy, K. K., "Scien-
tific bases of modern methods for setting hygienic
norms for chemical substances in the environment,"
Materialy vsesoyuznoy konf. 21-22 oktyabrya 1970 g.
[Materials from the All-Union Conference of October
21-22, 1970], Moscow, 1971, pp. 13-17.
14. Pavlenko, S. M., Gigiyena i sanitariya, No. 1, 1972,
pp. 40-44.
15. Avilova, G. G., Golubovich, Ye. Ya., Mal'tseva, N. M.,
Pankratova, G. N., et al., In book: Adaptatsiya i kom-
pensatsiya pri khimicheskikh vozdeystviyakh [Adapta-
tion and compensation in the case of chemical effects],
Moscow, 1973, p. 92, 103.
16. Petina, A. A., Yernichnykh, L. N., In book: Flyuoroz i
yego profilaktika [Fluorosis and its prevention], Sverd-
lovsk, 1967, pp. 50-55.
17. Sadilova, M. S., Voprosy gigiyeny i profpatologii i
toksikologiya [Questions of hygiene and industrial
pathology and toxicology], Sverdlovsk, 1967, pp. 25-26.
18. Sidorenko, G. N., Pinigin, M. A., "Scientific bases of
modern methods for setting hygienic norms for chem-
ical substances in the environment," Materialy vseso-
yuzn. konferentsii 21-22 okt, 1970 [Materials from
the All-Union Conference of October 21-22, 1970],
Moscow, 1971, pp. 13-17.
19. Spynu, Ye. I., Gigiyena primeneniya, toksikologiya
pestitsidov i klinika otravleniy [Hygiene of the use and
toxicology of pesticides and the clinical aspects of
poisoning], Kiev, 1968, pp. 103-108.
20. Tolokontsev, N. A., Gigiyena i sanitariya, No. 3, 1960,
pp. 29-35.
21. Ulanova, I. P., Avilova, G. G., Bazarova, L. A.,
Mal'tseva, N. M., Migukina, N. V., Khalepo, A. I.,
Eytingon, A. I., In book: Adaptatsiya i kompensatsiya
organisma pri vozdeystvii khim. zagryuazneniy okruzh.
sredy [Adaptation and compensation by the organism
to the effect of chemical pollution of the environment],
Moscow, 1972, pp. 17-21.
22. Kharitonova, V. A., PerfU'yeva, Z. V., Voprosy pro-
filaktiki i lecheniya zoba na Urale [Questions of the
prevention and treatment of goiter in the Urals], a
collection of papers from the Sverdlovsk Medical Insti-
tute, Sverdlovsk, 1961, No. 34, pp. 20-25.
23. Shabad, L. M., Sanotskiy, I. V., Zayeva, G. N., Bruye-
vich, T. S., Katsnel'son, B. A., Yanysheva, N. Ya.,
Shugayev, B. B., Gigiyena i i sanitariya, Moscow, No. 4,
1973, pp. 78-81.
24. Yanvsheva, N. Ya., Gigiyena i sanitariya, No. 7, 1972,
p. 87.
48
-------
The Concept of the Threshold Nature of the Reaction of Living
Systems to External Actions and its Consequences in the
Problem of Protecting the Biosphere Against Chemicals
I. V. Sanotskiy *
Legislative and other types of regulations of the
state of man's habitat, and that of the plant and ani-
mal species associated with him, the toxicological
limitation (hygienic normalization) of pollution of
the production sphere and of daily life, as well as the
derived practical measures of a technological or other
nature, depend in the final analysis on the resolution
of a fundamental question. Is there a threshold of
intensity of harmful effect of external factors, or is
any intensity different from zero harmful? In other
words, can objective scientific facts be used as the
basis for protecting the biosphere, or should sub-
jective decisions (the "acceptable risk" concept on
the one hand, and so-called "zero" pollution on the
other, and so forth) be used to some extent?
What must be emphasized is the fact that one
generally is not talking about the threshold of any
changes in living systems as a result of external ex-
citation, but of the incidence of reactions beyond the
limits of ordinary physiological variations observed
in the process of homeostasis, something that will be
taken up in more detail below.
A mathematical description of the task is not the
subject of this report. The purpose of this report is
to set forth corresponding points of view, and some
of the material used as the basis for for them.
As is known, the occurrence of excitation, that is,
of a complex biological reaction to external action
in the form of functional changes, and the physico-
chemical processes that underlie them, is a property
of living matter at any stage of organization.
The magnitude of the effect depends on the
strength of the stimulation. Bowditch (1871) dis-
covered the "all" or "nothing" law, that is, there is
no excitation ("nothing") when stimulations are
subthreshold, but maximum excitation ("all") is de-
veloped when the stimulus is at threshold strength,
and does not increase when the stimulus is intensi-
fied. However, these propositions now are regarded
only as a rule, because local excitation, local re-
sponse, occurs in the stimulated sector when stimuli
•Inttltute of Labour Hygiene and Professional Diseases, MOKOW
are subthreshold. Nor can the "all" characterize the
maximum that the action potential of the nerve or
muscle fiber can attain. According to another law,
the "law of intensity," the effect of the reaction is
greater the stronger the stimulus causing it, over a
predetermined range of intensity. Even more com-
plex interrelationships, three and more phases, have
been described (Vvedenskiy, N. Ye.; I. V. Sanotskiy,
1950, 1970; P. V. Simonov, 1962). The magnitude
of the reaction, as is known, depends significantly on
the functional state of the reacting substrate.
Content of the Problem
The question of the feasibility of establishing a
threshold of harmful effect for most types of effect
of chemical compounds is not in doubt among most
authors. However, this question is debatable when
it comes to mutagens, blastomogens, and radiation
injuries. No decision was reached at the symposium
organized especially for the discussion of this prob-
lem in February 1970, by the American Association
of Hygienists.
At the same time, the threshold nature of the
effect of all harmful factors (including radiation, and
particularly chemical) can be proven theoretically by
starting from the basic differences between living and
nonliving things (the constant exchange by the
organism of matter and energy with the environment,
the continual restoration by the organism of its
structure, and the constant adaptation to the en-
vironment, including those similar to itself during
the process of reproducing).
The simplified expression
Dy = Do — D«x — Dm«t
where Dy is the dose of toxin reaching the receptor;
Do is the dose of toxin introduced into the organism;
D« and Dmet are the doses of toxin excreted from the
organism, and rendered harmless as the toxin moves
toward the receptor, respectively, shows the threshold
nature of the effect [provided that (Dex + Dmet) is
not decreased in proportion to decrease in D0]. Data
49
-------
from an experiment with nonmetabolized and meta-
bolized substances show that this condition is ob-
served. Indeed, only hypothetical substances abso-
lutely not excreted and absolutely not rendered
harmless can reach the receptors without decreasing
in quantity. There are no such substances.
However, even if the receptors are considered to
be isolated, so to speak, from the integral organism,
it will be found that simple fixation of the toxin by
the receptor still does not mean a biologically signifi-
cant reaction with it. For example, it has been shown
in isolated human crythrocytes that fixation (includ-
ing strong) up to a certain limit by a cell membrane
of lead ions does not lead to an increase in the yield
to an incubation medium of potassium and hemo-
globin (Figure 1), according to data provided by
Grigarzik and Passow, 1958.
Thus, the reconstruction of the molecules can take
place without any obvious functional changes in the
biosubstrate. The occurrence of a "biologically im-
portant" effect depends on the "significance" of the
substrate in the vital activities of the cell, or on the
existence of a reserve metabolic pathway (Dinman,
1972).
The processes of adaptation, and of constant
restoration (activation of detoxication, intensified
production of biological structures bound by the
toxin, such as enzyme molecules, the inclusion of
bypasses for normal metabolism, the activation of
the organism's macrosystems aimed at adaptation to
new living conditions, and others, for example) are
added to this in the integral organism. Thus, damage
develops only when the rate at which the damaging
processes (inactivation, degeneration, and so forth)
occur exceeds the rate at which the processes of
adaptation and restoration occur. These phenomena
were well studied during the fractional use of ionizing
radiation, and these are the facts that became the
basis for the ideas as to the threshold nature of its
effect (Strelin, G. S., et al.).
Similar phenomena have been described as taking
place during the action of chemical factors. Thus,
cessation of intensified metabolization of bromo-
benzene in p-bromophenylmercapturic acid (durinv
extended action) was accompanied by decompen-
sated pathology (T. Shamilov). An increase in the
expiration of methylene chloride, and a correspond-
ing reduction in its content in the blood, led to
physiological adaptation (with a conditionally mono-
tonic 1.5-month effect) (G. G. Avilova and N. M.
Mal'tseva).
Adaptation at the subcellular, cellular, organ,
organism, and population level while under the ex-
tended effect of a damaging factor has repeatedly
been demonstrated when the external effect has been
most varied in its nature. Included here is the repair
of chromosome rearrangements in the case of frac-
tional irradiation (N. P. Dubinin, S. P. Frmonenko,
UJ
cc
UJ
UJ
OJS
£
7.0
6.0
5.0
4.0
3.0
2.0
1.0
YIELD OF K FROM ERYTHROCYTES (PERCENT)
10 20 30 40 50 60 70 80 90 100
2468 TO 12
Pb CONTENT IN SOLUTION (x10-« Mol/ml)
SOURCE: GRIGARZIK H., PASSOW H., PFLUGERS ARCHIV.
VOL. 276, p. 80 (1958)
FIGURE 1. THE Pb CONTENT IN ERYTHROCYTES AND THE
YIELD OF K FROM THE ERYTHROCYTES AS A
FUNCTION OF THE Pb CONTENT IN SOLUTION
50
-------
et al.)» the normal life of human populations under
increased (up to a predetermined limit) radiation
background conditions (high mountains, areas in
which radioactive ores are present, and the like),
repair of chromosome rearrangements in the case of
the extended effect of mutagens: ethylenimine, L. D.
Katosova, 1972; ethylene chlorohydrine, G. K. Isak-
ova, et al., 1971; 2-methylthio-4,6-bis-(isopropyla-
mino)-sym-triazine (prometrin), DDT, 2,4D, L. P.
Yefimenko, 1973, v. N. Fomenko, et al., (Figure 2),
rapid inherited adaptation of bacterial to antibiotics,
of insect populations to insecticides, and so forth.
The law of natural selection is not applicable to a
human population, so the precise facts as to the true
adaptation of man to life under chemical pollution
conditions have not yet been forthcoming.
The theory of harmfulness of the effect is of great
importance in the problem of the threshold nature
of an organism's reactions to external effects. What
is important to the protection of the environment is
not any threshold of reaction of the biosubstrate, but
the threshold of reactions that have the harmfulness
criterion. Similar ideas were formulated by N. S.
Pravdin in 1934, and the theory was developed by
his successors. More and more scientists have re-
cently subscribed to the propositions expressed. Weil
(1972), for example, believes there are dose levels
that have no harmful effects on any animal species
or on humans, although the effect may be statistically
meaningful. I. G. Akoyev, et al. (1972), while con-
firming the nonthreshold nature of primary reactions
to radiation effect, advanced the hypothesis of a
threshold number of damaged systems; that is, the
hypothesis of a threshold of harmful effect.
Thus, the theory of the threshold nature of the
effect does not imply reaction of the biobustrate to
the effect of external factors in general, but rather
reactions that are biologically (including medically)
important.
Analysis of Dose-Effect Curves
The threshold nature of the harmful effect of ex-
ternal factors can be demonstrated by an analysis
of the dose-effect curves. Many of the ideas that have
evolved as to the nonthreshold nature of the effect
have been the result of assumptions and extrapola-
tions. The dose-effect relationship is not linear, so
the establishment of the end of the curve corres-
ponding to the lower level is extremely important
(Weil, 1972; Stokinger, 1972; Weisburger, 1968;
Bock, 1968; Mantee, et al., 1961, 1963; Brues,
1958). The conditions under which the experiment
is performed (species, sex, age of animals, state of
nutrition, and so forth) have a significant effect on
the shape of the curve; so the results of determining
thresholds should be reproduced many times over.
55
»
Z
O
g
cc
CC
LU
CO
UJ
o
•
cc
X
o
2
wks
1.5 2.5
mon mon
4
mon
1/50 LDso prometrin
1/50 LD5Q lindane
months
.1/50 LD5Q butyl ether, 2, 4D
.1/50 LD50 DDT
FIGURE 2. THE CYTOGENETIC EFFECT OF PROMETRIN; LINDANE;
BUTYL ETHER, 2, 4D;AND DDT IN THE BONE MARROW
OF RATS
51
-------
The intensity of the organism's reaction usually de-
creases with reduction in the dose of the substance
administered, with the reaction dropping to zero be-
fore the dose reaches zero (Hatch, 1973).
As in the past, most arguments today stem from
analysis of dose-effect curves obtained as a result
of the effect of mutagens and blastomogens. The
curve of the blastomogenic effect of 3,4-benzpyrene
obtained by N. Ya. Yanysheva (Figure 3) enabled
the USSR to be the first in the world to establish the
maximum permissible concentration for this toxin
(L. M. Shabad, I. V. Sanotskiy, et al., 1973). More
and more researchers now tend to favor the feasi-
bility of establishing safe levels for such substances.
Figure 4 is the curve obtained in our laboratory by
L. D. Katosova (1973) for the cytogenetic effects
in bone marrow as a function of the chloroprene
concentration. The curve has the shape of an ex-
pirant over the concentration segment studied. How-
ever, in this, as well as in other, similar investigations,
only tht yield of a relative number of aberrant cells
in experimental animals beyond the limits of the
variations in the spontaneous level for the control, as
well as for the generally accepted "norm," should be
considered as the harmful effect. As is known, Gad-
dum (1956) considered the safe zone to be the
magnitude LD00 ± 6er, so that the probability of
the onset of a fatal result in this case is 1CH.
The threshold of the harmful effect of a substance
is that minimum concentration of it in an environ-
mental object, the effect of which will cause changes
to occur in the organism (under concrete conditions
for the admission of the substance) that exceed the
limit of physiological adaptive reactions, or latent
(temporarily compensated) pathology.
The commission designated to fix the maximum
permissible concentrations of harmful substances in
the production sphere developed a tentative definition
V)
oc
X
u.
O
100
90
80
70
60
50
40
30
20
10
0
Experimental curve-
Calculated curve*
~ -. . .
. Q •>- tv o
POP
DOSE OF 3, 4-BENZPYRENE (in mg)
p p
u>
* - CALCULATED CURVE OF LOGARITHMIC RELATIONSHIP
WHERE Xn IS THE DOSE (in nig) (REFERENCE POINT);
X IS THE MAXIMUM INEFFECTIVE DOSE IN THE
EXPERIMENT, EQUAL TO 0.02 mg.
SOURCE: YANYEHAVE, N. YA.. GIGIYENA I SANITARIYA, No. 7, 10., 90 (1972)
52
FIGURE 3. CURVE OF THE FREQUENCY OF LUNG TUMORS AS
A FUNCTION OF THE DOSE OF 3, 4-BENZPYRENE
INJECTED INTO THE RESPIRATORY TRACT OF
MONGREL RATS.
-------
.054 .064 0.3 .32
1.45 3.5
log concentr.
in mg/m3
FIGURE 4. THE CYTOGENETIC EFFECT IN BONE MARROW
CELLS OF ANIMALS AS A FUNCTION OF THE
CHLOROPRENE CONCENTRATION
of the criteria for the harmfulness of changes result-
ing from the effect of a particular substance that was
based on the departure of an index beyond the limits
of annual or seasonal variations in the norm (± 2ir),
as well as the stability of the changes, even though
not exceeding the norm.
However, a formal approach such as this is not
always satisfactory because in the first place, the
norm is not known in all cases.
In our opinion, the question of criteria of harm-
fulness may be resolved by the delimitation of a
stage of the true physiological adaptation to changed
environmental conditions and of a stage of latent,
temporarily compensated pathology under the effect
of minimally effective quantities of the toxin.
We have introduced into the system proposals for
methodological procedures for use in resolving the
problem described (I. V. Sanotskiy, N. G. Ivanov,
N. M. Karamzina, V. N. Fomenko, 1971). The pro-
posal was to use the following known biological
criteria of the correspondence of environmental con-
ditions as the basis:
• the normal condition for the population (con-
stancy of numbers of the species);
• the capacity of the organism to adapt to the
environment (the law of the unity of organism and
habitat);
• preservation of the organism as a single bio-
logical system.
This approach served to establish, in particular,
that the use of functional and extremal loads is a deli-
cate matter, and that their intensity must be varied
widely. A comprehensive investigation is necessary
to differentiate physiology and pathology in the inte-
gral organism. Thus, adaptation to stimulating toxins
in terms of a respiration frequency test was ac-
companied by pronounced pathology in terms of
morphological criteria and olfactometry (V. S. Poz-
dnyakov). Repairs to chromosome damage in the
bone marrow as a result of the effect of prometrin
were accompanied by the occurrence of chromosome
rearrangements in the liver tissue (L. P. Yefimenko,
A. Ye. Kulakov, 1969), and so forth.
Constancy in species numbers within known limits
can be predicted under laboratory experiment con-
ditions. The system of tests that was devised resulted
in establishing thresholds of harmful effect for many
toxins on the processes of reproducing future gen-
erations.
Results
Proof of the threshold nature of the intensity of
harmful effect is of enormous importance for intel-
ligent, practical activity to protect and to improve
the sanitary conditions of the environment surround-
ing man at work and at home, and to determine the
permissible load on human populations.
If a harmful substance cannot be removed com-
pletely from production, from community and home
53
-------
environments, and from the biosphere in general, its
content in environmental objects must be subject to
toxicological limitation based on a determination of
the thresholds of harmful effect. If the rate of decay
and removal of the substance from the environment
does not exceed the admission rate, the balance must
be restored, with the thresholds of harmful effect of
the substance on living systems taken into considera-
tion. If the substance is concentrated in the environ-
ment, and this includes the food chains, its emission
must be prohibited, particularly by means of closed
air- and water-supply cycles.
The permissible load of a chemical factor cor-
responds to its maximum permissible concentration
in environmental objects. This we defined (I. V.
Sanotskiy, 1970) as the concentration, the effect of
which on the human body periodically, or through-
out a lifetime, directly or indirectly through ecologi-
cal systems, as well as through possible economic
losses, causes neither somatic nor psychic diseases
(including latent and temporarily compensated) nor
changes in the state of health exceeding the limits
of adaptable physiological reactions detectable by
modern research methods immediately or later in the
life of present and subsequent generations.
Known is the fact that a multifactor analysis with
broad formalization of the initial data is needed to
make a synthetic estimate of the correspondence, or
noncorrespondence, between a chemical component
in the environment as a whole and hereditary, as well
as acquired, properties of living systems. The most
difficult stage of formalization in the case of socio-
logical estimates is that on questionnaire methods.
Nevertheless, certain publications along these lines
already have been obtained.
The problem of the distribution of individual
sensitivities, and of a concrete population, causes
great difficulties.
Collaboration between scientists with different
specialists, collaboration between scientists from
different countries, is the most effective way to en-
sure progress in the theory and practice of protect-
ing the biosphere against chemical pollution.
BIBLIOGRAPHY
Akoyev, I. G.; Maksimov, G. K.; Malyshev, V M., Luche-
voye porazheniye mlekopitayushchikh i statisticheskoye
modelirovaniye [Radiation damage to mammals and
statistical modeling], Moscow, 1972.
Vvedenskiy, N. Ye., Vozbuzhdeniye, tormozheniye, narkoz
[Excitation, inhibition, narcosis], St. Petersburg, 1901.
Dubinin, N. P., Problemy radiatsionnoy genetiki [Problems
of radiation genetics], Moscow, 1961.
Dubinin, N. P.; Tarasov, V. A., In coll: Sovremennyye
problemy radiatsionnoy genetiki [Contemporary prob-
lems of radiation genetics], Moscow, 1969, 3.
Yefimenko, L. P.; Kulakov, A. Ye., In coll: Mater, konfer.
molodykh nauchnykh rabotnikov. Instltut gigiyeny truda
i profzabolevaniy [Materials of the conference of young
scientific workers. Institute of Industrial Hygiene and
Occupational Diseases], 1969, Moscow, 139-140.
Isakova, G. K.; Ekshtat, B. Ya.; Kerkis, Yu. Ya., Gigiyena
i sanitariya, 1971, No. 11, pp. 9-13.
Katospva, L. D., "An evaluation of the danger of the muta-
genic effect of ethylenimine and chloroprene," Author's
summary of candidate's dissertation, Moscow, 1973.
Kurlyandskaya, E. B.; Sanotskiy, I. V., Gigiyena truda i
profzabolevaniya, 1965, 3, pp. 3-9.
Ivanov, N. G.; Kudinova, O. V.; Pozdnyakov, V. S.;
Klyachkina, A. M., "Reduction of the sense of smell in
animals as an index of the harmful effect of industrial
toxins," In coll: Toksikologinovykh prom. khim. v-v.
[Toxigenic industrial chemicals], No. 13, Moscow, 1973,
p. 33.
Pravdin, N. S., Rukovodstvo promyshlennoy toksikologii
[Handbook of industrial toxicology], Moscow, 1934.
Pravdin, N. S., Metodika maloy toksikologii promyshlennykh
yadov [The methodology of minor toxicology of indus-
trial toxins], Moscow, 1947.
Sanotskiy, I. V., "The Influence of Different Doses of Radio-
thorium on Reproduction," Candidate's Dissertation, Mos-
cow, 1950.
Sanotskiy, I. V.; Avkhimenko, M. M.; Ivanov, N. G., In
coll: Toksikologiya novykh promyshlennykh khimiche-
skikh veshchestv [The toxicology of new industrial chem-
icals], No. 9, Moscow, 1967, p. 71.
Sanotskiy, I. V., In book: Melody opredeleniya toksichnosti
i opasnosti khimicheskikh veshchestv [Methods for de-
termining the toxicity and danger of chemicals], Moscow,
1970.
Sanotskiy, I. V.; Ivanov, N. G.; Karamzina, N. M.; Fo-
menko, V. N., In coll: Nauchnyye osnovy sovremennykh
metodov gigiyenicheskogo normirovaniya khimicheskikh
veshchestv v okruzhayushchey srede [The scientific bases
of modern methods for fixing hygienic norms for chem-
icals in the environment], Moscow, 1971, pp. 63-68.
Simonov, P. y., Tri fazy v reaktsii organizma na vorzrasta-
yushchiy stimul [Three phases in the reaction of the or-
ganism to increasing stimulus], Moscow, 1962.
Strelin, G S.; Med. radiologiya, No. 2, 1960, p. 77.
Fomenko, V. N.; Strekalova, E. Ye.; Katosova. L. D.;
Chirkova, Ye. M.; Sal'nikova, L. S.; Silant'yeva, I. V.;
Yefimenko, L. G.; Kulakov, A. Ye., "Pharmacology.
Chemptherapeutic agents. Toxicology," Itogl nauki i
tekhniki [Summaries of science and technology], Vol. 5,
1973, pp. 128-145.
Hatch, T., WHO Bulletin, 1973, Vol. 47, No. 2, pp. 153-
161.
Shabad, L. M.; Sanotskiy, I. V.; Zayeva, G. N.; Bruyevich,
T. S.; Katsnel'son, B. A.; Yanysheva, N. Ya.; Shugayev,
B. B., Gigiyena i sanitariya, No. 4, 1973, pp. 78-81.
Yanysheva, N. Ya., Gigiyena i sanitariya, No. 7, 1972, pp.
87-91.
Yarmonenko, S. P.; Palyga, G. F., Med. radiologiya, No. 3,
1963, p. 66.
Bock F.C. Nat. Cancer Inst. Monogr., 1968, 28, 57-63.
Brues A.M. Science, 1958, 128,693-699.
Dinman B. Science, 1972, 175, N. 4021.
Gaddum J. H. Brit. J. of Pharmac and Chemother, 1956,
11(2) 156-160.
Grigarzik H., Passow H. Pflugers Archiv. Bd. 267, S. 73-92
(1958).
Hatch T.E. Arch. Environ. Health, 1971, 22, 687-689.
Frohbere H., Baner A. Arzneimittel-Forsch, 1973, 23, N. 2,
230-236.
Mantel N., Heston W. E., Guriam J. J. Nat. Cancer Inst.
1961, 27, 203-215.
Mantel N. Clin Pharm. Ther. 1963, 4, 104-109.
Stokinger H. E. Arch. Environ. Health, 1972, 25, N. 3,
153-157.
Weil C .S. Toxicol and Appl. Phormocol., 1972, 21, N. 4,
454-463.
Weil C. S. Inf. Congr. Pharmacol. San. Francissa Calif.
1972, Abstrs. Invit. Presentat. (1972) 258-259.
Weisbirger J H., Weisburger E. K. Food Cosmet. Toxicol.
1968, 6, 235-242.
54
-------
Questions of Extrapolating Data in Evaluating the Mutagenic
Action of Environmental Factors on Man
N. P. Bochkov *
Experiments using animals are unavoidable in
evaluating the mutagenic action of environmental
factors on man. No one can dispute this rule. Our
success in gaining knowledge of the genetic danger
to man from ionizing radiation, alkylizing com-
pounds, and other mutagens is associated with large-
scale experimental research on animals.
In the 1930's, G. Meller, N. V. Timofeyev-Res-
sovskiy, and other geneticists, for the first time posed
the question of the danger of ionizing radiation to
man's heredity based on extrapolating results ob-
tained using animals.
Experimental genetics presently has a vast arsenal
of methods for checking the mutagenic action of
various factors on animals. They include cytogenetic
methods applied to somatic and embryonic cells,
consideration of dominant lethals, recessive lethal
mutations and recessive and dominant apparent mu-
tations. Experiments have been performed using
drosophilae, mice, rats, guinea pigs, and rabbits. The
most valid conclusions have been obtained from ex-
periments using drosophilae and mice. The other
animals have been used, as a rule, in cytogenetic
experiments, or in experiments that took dominant
lethals into consideration. Gene mutation considera-
tions in mammals are very time consuming and ex-
pensive. Nor does consideration of the dominant
lethals even begin to settle the question, because this
is a collective group of genetic damages, usually
vague in nature.
It is desirable, when considering the present state
of the problem of extrapolating data obtained on
animals to man, to discuss the following questions.
1. What are the circumstances that force us to
turn to extrapolation?
2. What permits us to make extrapolations?
3. What hinders making extrapolations?
4. How should extrapolations be made?
1. What are the circumstances that force us to turn
to extrapolation?
The enormous importance of direct data in eval-
uating the effect of mutagenic factors on man's he-
redity is well known. However, research in the mu-
tagenic effect of environmental factors must be con-
ducted quickly as possible. Consequently, even vast
quantities of data and the most improved methods
for considering gene mutations in human embryonic
cells cannot accomplish the primary task, that of
verifying the mutagenicity of environmental factors,
since they require more time, from several years to
several decades. Therefore, models must be used.
These can be human cells in a culturel, or animals.
Chromosome aberrations and aneuploidy in hu-
man somatic cells can be studied by the effect of
mutagenic factors on the organism, and on animals
to consider chromosome and genome mutations in
somatic cells. Procedures that use human cells are
sufficiently well-developed for this purpose. How-
ever, similar experiments are impossible when we
turn to embryonic cells. This raises the question of
whether it is right to judge embryonic cells by so-
matic cells? Obviously not, because the formation of
embryonic cells includes a number of mitotic di-
visions and meiosis. The nature of chromosome be-
havior in meiosis differs sharply from mitosis, which
leads to different quantitative and qualitative be-
havior patterns in the mutation process in embryonic
and somatic cells. This raises for the geneticist the
question of the desirability of testing mutagenic fac-
tors on animals (so as to take chromosome aber-
rations into consideration) and extrapolating the
data to man.
But it is nevertheless sometimes possible, although
difficult, to obtain human embryonic cells for cyto-
genetic research, and thus estimate chromosome mu-
tation frequency. It would seem that in this case we
can manage without extrapolations. However, as an
analysis of the data in the literature shows, there is
a great difference in the frequencies of gene and
chromosome mutations for the effect of the same
mutagen. It is not simply a matter of the different
nature of the qualitative responses, but rather that
chemical mutagens can induce one type of mutation,
yet have no effect on another.
•Institute of Medical Genetics, Academy of Medical Sciences of the USSR, Moscow
55
-------
Attempting to delineate the question of gene mu-
tations, we again are faced with the choice of one of
two approaches; either induce gene mutations in a
culture of human cells (the procedures have now
been developed), or experiment on animals. Nu-
merous data show the great differences in mutation
frequencies in vivo and in vitro, and this is what
makes experiments on animals, and extrapolation,
necessary. Whether some of them are adequate, or
not, will be gone into in what follows.
Thus, three circumstances force us to turn to ex-
trapolations; the difference in the mutagenic response
of somatic and embryonic cells, the difference in the
frequency of induced, gene, and chromosome mu-
tations, and the difference in the frequencies of mu-
tations in vivo and in vitro.
2. What permits us to make extrapolations?
Many of the conditions that force us to turn to
extrapolations in evaluating mutagenic factors in the
environment were enumerated above. But at the
same time, all are, as it were, contraindications. Can
we make extrapolations anyway, or is doing so a
useless task? What is it that permits us to answer
the question of the correctness of making extrapola-
tions in the affirmative?
The structure of the individual chromosomes, and
of the karyotype, in higher animals conforms to a
single principle. The chromosome is a strictly ordered
nucleoproteinic structure, which undergoes identical
changes in the process of reproduction and function-
ing. With the single principle of the chromosome
structure as a basis, it can be assumed that there are
common behavior patterns in chromosome reactions
to mutagenic effects when the mutagens reach the
chromosomes. Experimental data confirm these as-
sumptions. Based on a great deal of research in the
mutation process, we can speak of its universality.
We can see by the example of radiation and of alky-
lizing agents that the similarities in the mutagenic
responses of the higher organisms are greater than
the differences, especially as regards the qualitative
response. All substances that have been found to be
mutagenic for mammals cause chromosome abera-
tions in human somatic cells in vitro and, where
tested, in vivo as well.
3. What hinders making extrapolations?
The problem of extrapolation would not arise if
there were no differences in the mutation response of
the various organisms. The data obtained for any
organism simply would be used to evaluate the hu-
man mutagenic danger. However, this is not so, and
the more the mutation process is studied, the more
the restrictions placed on extrapolation.
The first major restriction on extrapolation is as-
sociated with the difference in the metabolism of
chemical substances in animals and man. Let us, by
way of an example, cite thalidomide, the embryotoxic
effect of which is strongly manifested in man, but
not at all in rats. Many chemical compounds are
toxic for one species, yet decompose rapidly in an-
other. This most often is associated with hereditary
features of metabolism.
What follows from the extrapolation restrictions
considered above is that the metabolism of the mu-
tagen in man and in animals must be known if cor-
rect extrapolations are to be made. Obviously, ex-
trapolations will be more accurate for radiation
mutagenesis than for chemical.
The second restriction placed on extrapolation is
associated with the dissimilar qualitative behavior
patterns in the mutation process in different animals.
Moreover, it is apparent that induced mutagenesis
can be stronger or weaker in man. There is no basis
for assuming that it always is more intensive in man
than in animals.
What causes the quantitative differences in the in-
duced mutation process in different animals? Induced
mutagenesis, as has been established recently by
different authors, is an unusually complex process,
one that can be influenced by many factors. The pri-
mary damage from mutagens, apparently, is of a po-
tential nature. The cells contain reparative systems.
All of this leads to quantitative variations in the
mutation process. Reliable procedures for evaluating
the role of the primary and secondary processes in
the differences in the mutation response remain to
be developed. The importance of the phase of the
cess cycle for mutation frequency, the importance of
type of cells, of the species and age of the animals,
has been discovered for radiation and chemical
mutagenesis. The relationship between mutation fre-
quency and mutagen dose can be different in different
animals, and this fact makes extrapolation partic-
ularly difficult.
We can cite several papers to illustrate these prop-
ositions. N. I. Shapiro and Yu. Ya. Kerkis were the
first to establish the fact of dissimilar radiosensitivity
of the chromosomes of different species of animals.
Dominant lethals were taken into consideration in
the experiments conducted by N. I. Shapiro. It was
found that the numerical incidence of lethals in mice,
rats, and rabbits, was dissimilar, with the rabbit
showing a higher death rate of zygotes before
implantation.
More comprehensive work was conducted by
M. Lyon and B. Smith, who evaluated the frequen-
cies of dominant lethals at different times after ir-
radiation. They established the fact that it is not
just a matter of the differences in the general genetic
radiosensitivity, but also that the radiosensitivity will
vary differently in different animals, and at different
times.
56
-------
The dissimilar frequency of dominant lethals in
different animals and their dissimilar shift during
spermatogenesis have been noted in the effect of
ionizing radiation, as well as of chemical mutagens.
This can be seen in the data provided by M. Lyon,
B. Smith, and D. Bateman. It is important to note
that the curves are dissimilar, and that the mean
frequencies differ.
The differences in the frequency of induced aber-
rations in different organs and in different types of
cells in different animals can be seen in the work
done by Yu. Ya. Kerkis, et al.
The extrapolation problem is complicated by the
fact that there are age differences in the response to
mutagenic actions. This has been described by a
number of authors for animals, and we have done
so for man.
The following rule has thus far been applied to
extrapolations: gene mutations increase linearly with
the dose, and chromosome mutations as the square,
with the genome mutations question unresolved.
These conclusions have not always been reached
on the basis of strict statistical criteria. For example,
it is now well-established that the dependence on the
dose is not the square and determinate for radiation-
induced aberrations.
The sum of the experimental data permits the con-
clusion to be drawn that the coefficients of the equa-
tions of mutation frequency in different species of
animals (or types of cells) are different, and depend
on the mutagenic factor dose. All this speaks to the
enormous difficulties standing along the road of
extrapolations. Things are even further complicated
by the fact that the yield of muttaions per dose unit
differs for different dose ranges. And as a rule, the
experimenter is not working with the same doses
with which man comes in contact.
4. How should extrapolations be made?
The question of the mutagenicity of any factor
for man is decided primarily on a qualitative basis.
Extrapolations therefore, are of the same nature.
What rules can we propose on the basis of the sum
of contemporary knowledge?
(a) First of all, the researcher is confronted with
a dilemma; should the extrapolations be made from
human cells in a culture, or from animals? The eval-
uation of the mutation process is a very complex
procedure. After all, three types of mutations (gene,
chromosome, and genome) must be taken into con-
sideration, as must at least two types of cells (so-
matic and embryonic), and the metabolism of the
chemical mutagen. There is, therefore, no unambig-
uous answer to this question in all cases. The de-
cision is different in each concrete case. Preference
sometimes may be given to a culture of human cells,
sometimes to animals.
(b) If mutagenicity of any factor is detected for
an animal, it must be considered mutagenic for man,
with the exception perhaps of those cases when the
metabolism of the mutagen in animal and in man
differs sharply at precisely the point of application
of the mutagenic activity. Failure to observe a muta-
genic effect of a compound in one, or even in several
species, does not mean that it is safe for man, but
only that we may hope it is. This approach seems
objective to us, because of the great importance of
underestimating the presence of a mutagenic effect
attributable to any chemical substance (more im-
portant than overestimating). Thus, a mutagenic
effect established for any animal object must be di-
rectly extrapolated to man, although this may not
always be confirmed. Negative results in investigating
a mutagenic effect must be extrapolated with ex-
treme care in the sense of the absence of a mutagenic
effect by a certain chemical substance on man's
heredity, and only for the condition that the metab-
olism of the compound in the research animal and
in man is exactly alike.
(c) If the mutagenic effect of any factor is estab-
lished for animals, the task of quantitative extrapola-
tion arises, because the basic approach to evaluating
environmental factors must be the population ap-
proach. If numerical data are obtained for several
species of animals, it is better to proceed using mean
data, rather than using data on one species of ani-
mals, when evaluating the mutagenic danger to man.
Of course, the task of extrapolation will be difficult
if drosophilae and mice are included in this group.
Preference usually should be given to mammals, par-
ticularly in the case of chemical mutagenesis. Quan-
titative conclusions cannot be precise, even when
all extrapolation requirements are met as strictly as
indicated. The estimates will be approximate in any
case. Therefore one must strive to determine the
order of the danger, not the precise figures. Thus,
quantitative extrapolations should be made more
carefully than qualitative ones.
In the case of quantitative extrapolations, it is
important to evaluate the total mutagenic response,
proceeding from the different sensitivity of the phases
of the cell cycle, according to the weighted mean
principle.
(d) The geneticist is confronted with two complex
questions when evaluating the mutagenicity of a
factor; what animals to use in the experiments, and
which tests to select for the genetic evaluation. Ob-
viously, the two main species of genetically well
studied and convenient objects should be mentioned
first, drosophilae and mice. A great many muta-
tions can be studied in each of these species, but
expanded experiments on mice to take recessive
mutations into consideration are very time-consum-
ing and expensive. The drosophila therefore is most
often used. Experiments with mice as a rule are
limited to dominant lethals. Experiments must use
mammals when a widely distributed substance, or
factor, is being checked, or when there are reasons
57
-------
for an assumption as to its mutagenic effect. When
the mutagen does not penetrate the testicles in ani-
mals, but does so in man, experiments can be set up
so that the mutagen is injected into the scrotum of
the animals.
(e) It is not always necessary to attempt to make
extrapolations. If experiments can be set so that
human cells are used, it becomes mandatory that
extrapolations not be made. There are no adequate
models for evaluating chemical mutagenesis in man.
Therefore, the current problem, and one that applies
in all cases, is that of obtaining direct data on the
influence of mutagenic factors on man's heredity. It
is necessary, for example, to extrapolate the cyto-
genetic effects in somatic cells. Modern procedures
permit doing this directly on human cells as a result
of the effects of the mutagenic factors in vivo and
in vitro. The large-scale development of procedures
for taking gene mutations in human somatic cells
into consideration will narrow the circle of ques-
tions for which it will be necessary to extrapolate
data. At the same time, and in conclusion, it should
be emphasized that as our knowledge of the mu-
tation process increases, the limits of the potentials
of extrapolations will become clearer, and the selec-
tion of the models used to evaluate environmental
mutagenic factors will be more objective.
Extrapolation, in the final analysis, is based on
the results of experimental modeling. Consequently,
the better the experimental object is selected, the
more accurate will be the extrapolation. Well known
from experimental pathology is how attentive and
careful one must be when modeling any of the
human pathological processes.
58
-------
The Question of Genetic Danger from Environmental
Pollutants
L. M. Filippova *
Pollution of the habitat of living organisms by
biologically active agents—chemical (the wastes
from different industries, erosion loss from agricul-
tural areas, and the like), as well as physical (in-
crease in the radiation background, thermal pol-
lution, noise, and others)—has become a side-
effect of the development of the human society.
The progressing pollution of the natural environ-
ment has a number of biological consequences, with
one of the principal ones the danger of the occur-
rence in the majority of cases of negative hereditary
changes in man, as well as in animals and plants,
under the effect of chemical and physical agents in
the environment. The great many chemical com-
pounds entering the natural environment, individ-
ually and in a variety of compounds, are capable
of inducing gene, genomic, and chromosome muta-
tions, which cause an increase in the frequency of
congenital deformities, defects and anomalies, many
hereditary diseases, stillbirths, spantaneous abor-
tions, and the like.
Major chromosome damage (aberrations) causes
a genetic defect the scale of which is such that the
damage is rarely transmitted by heredity because of
the poor viability, or sterility, of the carriers, and
this is precisely why aberrations have little impor-
tance from the standpoint of population.
Recessive gene mutations play a special role from
the standpoint of population. They are harmful to
the organism in the majority of cases. Transmitted
and accumulated from generation to generation in
the latent (hetero2ygous) state, they increase the
"genetic load" of the population. The problem of the
"genetic load" is a particularly serious one for a
population in which the effect of natural selection
has weakened, and this would apply to the human
population.
Systematic study of the genetic danger of agents
that pollute the environment, of products of chemical
synthesis widely used in industry, agriculture, and
daily life [1], has shown that many of them definitely
have mutagenic activity and can damage the genetic
material of living organisms. Mutagenically active
•Main Administration of the Hydrometeorological Service, Moscow
compounds have been found in medicinal com-
pounds, pesticides, dyes, and others.
Therefore, no one today can have any doubt about
the fact that all chemical agents that in some way or
another come in contact, in particular, with man, as
well as with animals and plants, should be subject
to genetic control. The results of such control should
be taken into consideration in setting hygienic stan-
dards, and should further serve as the basis for
drastically limiting the content of mutagenic sub-
stances in the environment. The lexicological stan-
dards accepted at the present time certainly cannot
be recognized as adequate in all cases. While there
are many cases when the mutagenicity of a com-
pound correlates with toxicity, slightly toxic chemical
products often have mutagenicity. And it is this
mutagenicity, and not the toxicity, that should be the
limiting factor when setting standards for a chemical
compound such as this. Paraquinone dioxime, for
example, is widely used in the tire and rubber in-
dustry and while it has no pronounced toxicity (an
organoleptic index, color, is used in its standardiza-
tion) it does induce lethal gene mutations and
morphoses in Drosophila and chromosome rearrange-
ments in a culture of peripheral blood leukocytes
and in a culture of human embryonic fibroblasts [2].
Moreover, the mutagenic concentration of triethylene
melamine is less than the toxic by a factor of 100
[3].
The evaluation of the genetic threat of environ-
mental pollutants must proceed from the fact that
in the majority of cases the organism will be sub-
jected to the effects of a large number of mutagenic
agents in small concentrations. There are, therefore,
two problems of primary importance:
1. the problem of the simultaneous, or sequen-
tial, effect of several mutagenic agents;
2. the problem of defining the threshold of the
mutation process.
1. The living organism is subjected to the action
of a complex set of chemical and physical agents
that penetrate the organism over different paths and
from different media.
59
-------
Recent papers have dealt with the mutagenic effect
of two chemical agents, two physical (irradiation +
temperature, for example) agents, or a chemical
agent in combination with a physical agent (chemical
mutagen + temperature, chemical mutagen + ir-
radiation, and so on), making it possible to estimate
the situation, albeit tentatively, in connection with
the simultaneous presence in the environment of
many mutagenic agents. The possibility of additivity,
as well as of synergism of the genetic effect of two
mutagenic factors, not necessarily highly active indi-
vidually, has been pointed out. This increase in the
level of mutagenesis also has been observed for
chromosome aberrations, as well as for gene muta-
tions [4, 5, 6]. It appears that the reasons for the
intensification of the effect of combined mutagenic
effect can be very different indeed. For example, the
effect of one of the agents can be to change the phy-
siological state of the cell, change the hereditary ma-
terial to the labile state, damage the cell control
mechanism, inhibit reparative processes, and so on,
with the result a reduction in the resistance of the
organism to the effects of the second agent. Nor can
we exclude the possibility of the appearance of a new,
more highly mutagenic substance when two com-
pounds interact with each other, and with cell meta-
bolites.
2. This gives rise to a question. Will the effects
thus far discussed be valid for the case of extremely
low concentrations of mutagenic substances in the
natural environment? In other words, is there a
threshold of mutagenic effect of chemical agents and
radiation? It obviously is impossible to answer the
question categorically. And not only because our
present knowledge of the question is limited. This
also is so because the genetic effect includes a com-
plex set of damages to the hereditary material, each
different from the other, and characterized by speci-
fic mechanisms and kinetics.
Shortwave radiation, and chemical substances to
an even greater degree, are capable of inducing point
mutations and simple, primary chromosome breaks;
that is, those changes in the hereditary material the
basis of which is an elementary chemical act, or a
single ionization. Analysis of the kinetics of these
"single-strike" mutation events attests to the linear
nature of the dose-effect ratio [7, 8]. The dose-effect
line can be extrapolated through the zero point,
pointing to the absence of a dose threshold. Actually,
the formation of a single pair of ions, or the appear-
ance of a single molecule of a chemical mutagen in
a cell, is perhaps sufficient to induce point mutation
as a result of just one DNA base change. The oc-
currence of mutations in living organisms is prob-
abilistic in nature. Even the tiniest doses of mutagens
can induce mutation changes (which can be ex-
tremely serious, right up to being lethal, caused
by a single mutation in one of millions of genes),
simply among fewer persons. So, naturally enough,
the probability of genetic damage will decrease with
decrease in the mutagen dose. The proposition that
there is no dose threshold for inducing "single-strike"
mutation events is now well confirmed by direct ex-
periments [9-13], Those rare cases of induction of
mutations attributable to the inhibition of certain
enzymes needed for the production of DNA pre-
cursors are the exceptions. If such an enzyme in the
norm does not limit the DNA or chromosome repli-
cation rate, the inhibitor will be effective if some
minimum magnitude is exceeded [14].
But things are different in such damage to chro-
mosome material as complex rearrangements, for
which two, or more, independent chromosome breaks
are necessary. Here the dose-effect curve differs from
that described above. These structural changes, upon
irradiation, can be caused by accumulations of ions
from different tracks that occur independently. Their
frequency increase is not directly proportional to the
dose, but rather is more rapid, and there is a thresh-
old dose [7]. The same thing is seen for the dose
ratio in the case of the "double-strike" ("multi-
strike") effect of chemical agents, the result of which
is the appearance of two (of several) independent
breaks with subsequent regeneration and recombina-
tion of the chromosome material and the appearance
of complex rearrangements of the chromosomes.
Accordingly, the kinetics of "single-strike" muta-
tion events (the occurrence of gene mutations and of
simple chromosome breaks) can be characterized by
the linear dose-effect ratio and absence of a thresh-
old. The frequency of the "multistrike" mutation
processes (the occurrence of complex chromosome
rearrangements) increases more rapidly than the
dose, and has a threshold.
The experimental curve for the induction of muta-
tions attributable to physical, as well as chemical,
agents usually has the following, typical, shape.
of
ll
o
LU
oc
u.
O
a
MUTAGEN DOSE
FIGURE 1. DOSAGE-MUTATION RELATION
60
-------
The linear segment, ab, corresponds to the preferred
induction of gene mutations under the influence of
low mutagen doses. The be segment rebects the in-
crease in the percentage of large, complex, chromo-
some rearrangements and the decrease of induction of
gene mutations when the mutagen doses are in-
creased [15]. The linear segment, ab, usually is very
much longer for a chemically induced mutation pro-
cess than is the case for shortwave radiation. Many
highly specific, active mutagens (nitroso- and diazo
compounds, for example) induce gene mutations
with much greater frequency than chromosome aber-
rations. Virtually almost all the chemical mutagens
induce mainly gene mutations in the low dose range.
It is obvious that the mutagenesis curve will, in
practice, undergo substantial modification from com-
pound to compound because of the high degree of
specificity of the chemical mutagenesis (discussion of
the nature and causes of which are pointless in this
report). Nevertheless, the tendencies described will
be at the bottom of the mutation process.
Some researchers assume they have every reason
to talk about a threshold for toxic or teratogenic ef-
fects, as well as for mutagenic (including the induc-
tion of gene mutations). Well known is the fact that
mutation frequency is largely dependent on the phy-
siological state of the cell at the time of the action,
the special features of the cell's metabolism, on a
great many intra- and extra-cellular factors. Small
quantities of a potentially mutagenic agent that enter
the organism can be completely inactivated, and
rendered harmless, by the cell's defense mechanisms,
while induced pre-mutation changes can be healed
by the reparative systems. These small concentra-
tions of the mutagen thus will, for all practical pur-
poses, be inactive. It would seem that in cases such
as these there is an obvious mutagen effect threshold.
But can we, at this point, say that mutagenesis has a
threshold, in the strict sense of the word?
First of all, the magnitude of this "threshold" is
not absolute for the same compound, and will vary
greatly not only from object to object, but even from
gene to gene in the same genotype. Consequently,
this so-called "inactive" concentration will, to a very
large extent, characterize the degree of resistance of
the organism to the particular mutagenic action, and
not the threshold of mutagen action. This resistance
on the part of the organism is a factor that is subject
to changes as it is affected by definite conditions.
Important, and convincing experimental data [16]
attest to the fact that agents that are cell metabolism
modifiers cause significant changes to the induced
mutagenesis picture. Under real, natural conditions
the actions of a great many different chemical and
physical agents "subthreshold" under other condi-
tions and concentration, and can have a definite mu-
tagenic effect. Hydrogen peroxide, for example, will
not induce mutations in many organisms, even in
high concentrations, because it is decomposed by
their catalase. But if a toxin that blocks catalase
(potassium cyanide, or sodium azide, for example)
is added the mutation frequency will increase, even
when the smallest quantities of peroxide are involved.
Well known is the fact that an increase in the
quantity of oxygen, carbon monoxide, hydration,
visible light, temperature, and other environmental
factors, can sharply increase the frequency of muta-
tions induced by chemical mutagens and by irradia-
tion.
There can be a sharp increase in mutagenicity as
a substance metabilizes in the environment. Trans-
formation of a substance upon entering the environ-
ment can, for example, lead to a greater capacity to
be adsorbed, or to penetrate, the cell, such as is the
case with inorganic mercury, which is transformed by
microorganisms into methylated mercury [17].
Finally, the transformation of a "prethreshold"
into a "postthreshold" dose can take place as a result
of the additive, and even more so of the synergid,
effect of two, and more, mutagenic agents, which we
discussed above.
So, two things evidently can happen as a result of
the action by small doses of mutagens.
1. The organism responds to the effect of even
the smallest doses of a mutagen by straining the pro-
tective systems. This leads to inactivation, to the
rapid elimination of the mutagen molecules, or to
repairation of the primary damage to the genetic
material. Mutation will not occur in this case because
the processes involved in the mutagen action and in
those opposing such action are in equilibrium. The
action of the mutagen and, correspondingly, the or-
ganism's reaction are nonthreshold.
2. The strain of the protective systems reaches a
maximum with increase in the genetic load. Beyond
this point the action of even the smallest dose of
mutagen (and not necessarily the same one) will
lead to the formation of a mutation. Thus, the oc-
currence of a pronounced (pathological) change in
the hereditary structures has a threshold in the major-
ity of cases.
What follows from the above, then, is that it is
of primary importance when standardizing the total
genetic load of harmful substances in the environ-
ment to proceed from the nonthreshold nature of the
effect of mutagenic agents, and not from the thresh-
old nature of the occurrence of a pronounced (path-
ological) character. Even small quantities of a great
number of chemical and physical agents in the en-
vironment are constituents in the total genetic load
on the living organism. There can be little doubt that
if the systems for intracell regulation and the cell's
reparative systems protect the organism from the
harmful effect of a small dose of a mutagen, the cell
will be in no condition to withstand the simultaneous
action of tens, or of hundreds, of small quantities of
61
-------
mutagens that would not be fatal to the organism in-
dividually. And this is understandable, because ab-
sence of effect from low doses of a mutagen is not
because the low doses have no effect in the organism,
it is the result of the active counteraction launched
by the organism's protective systems, the capabilities
of which, naturally, are quantitatively limited.
mutagenic agents. The only real way to control the
content of these compounds is to standardize these
products at the emission source, or even sooner,
while they are being manufactured.
BIBLIOGRAPHY
CONCLUSION
Today, we are faced with the urgent need to ex-
pose the level of genetic damage attributable to en-
vironmental pollution. Hence, the task of genetic
monitoring of the human population is an extremely
important one. All cases of spontaneous abortion,
anomalies of genetic etiology in the newborn, as well
as in older groups, and among those exposed con-
stantly, or periodically, to the effects of some par-
ticular agent (at work, or while undergoing treat-
ment, for example), must be systematically taken
into consideration and studied, in order to assess
changes in the frequency of gene and chromosome
mutations in germ and somatic cells.
We must introduce genetic methods in standard
toxicological research in standardizing chemical prod-
ucts, and this in the immediate future. This task is
well within the realm of possibility. A series of
genetic studies can be made using the same animal
that takes part in an acute, or a chronic, toxilogical
experiment (dominant lethal mutations, chromosome
aberrations in sexual, or somatic, cells, and others).
Forms of genetic activity not accessible for direct
analysis in mammals should be tested along parallel
lines on adequate models. The determination of per-
missible quantities of chemical compounds in the en-
vironment should proceed from the nonthreshold
kinetics of the mutation process, and, as a minimum
from the summation of the genetic effect of even the
smallest doses of mutagens. The permissible concen-
trations should be infinitely small in the case of active
1. Rapoport, I. A., Filippova, L. M. ZhVKhO im. Mende-
leyeva, 1970, 6, 681.
2. Revazova, Yu. A. "The Hygienic Importance of Re-
search in the Genetic Activity of Industrial Com-
pounds." Author's Dissertation, Moscow, 1967.
3. Matter, B. E., Generoso, W. M. Abstracts. 2d Annual
Meeting of the European Environmental Mutagen So-
ciety, Czechoslovakia, 1972, p. 40.
4. Oster, I. I. In: Advances in Radiobiology, Oliver and
Boyd (Eds.). Edinburgh, 1957, p. 475.
5. Kaufmann, B., Gay, H., Rotnberg, H. Exptl. ZooL,
1949, 111, 415.
6. Nikiforov, V. G. V kn. Obshchaya genetika [General
Genetics]. "Nauka" Publishing House, Moscow, 1965,
p. 166.
7. Gershkovich, I. Genetika [Genetics]. "Nauka" Publish-
ing House, Moscow, 1968, p. 201.
8. Rapoport, I. A. V sb. Supermutageny [In collection:
Supermutagens]. "Nauka" Publishing House, Moscow,
1966, p. 9.
9. Sparrow, A. H., Underbrink, A. G., Rossi, H. H. Sci-
ence, 1972, 176, 916.
10. Smith, G. V sb. Genetika raka [In collection: Cancer
Genetics], Foreign Literature Publishing House, Mos-
cow, 1961, p. 442.
11. Freese, E., Freese, B. Radiat. Res., SuppL, 1966, 6,
97.
12. Glembotskiy, Ya. L., et al. V sb. Radiatsionnaya gene-
tika [In collection: Radiation Genetics]. Academy of'
Sciences of the USSR Publishing House, Moscow, 1962,
p. 300.
13. Marcovich, H. These des Sci., 3846, 1957.
14. Freese, E. In: The Evaluation of Chemical Muta-
genicity Data in Relation to Population Risk. Proceed-
ings of the Workshop held at Research Triangle Park,
N.C. 1973, 171.
15. Rapoport, I. A. Khimicheskiv mutagenei [Chemical
Mutagenesis]. "Znaniye" Publishing House, Moscow.
1966, p. 24.
16. "Evaluation of Genetic Risks of Environmental Chem-
icals." Report of a Symposium held at Skokloster,
Sweden, 1972, p. 13.
17. Miller, M. W., Berg, G. G., Eds. Chemical Fallout.
C. C. Thomas, Springfield, Illinois, 1969. 531 pp.
62
-------
Hygienic Bases for Protecting the Environment
C. I. Sidorenko,* Ye. I. Korenevskaya,* M. A. Pinigin,** G. N. Krasovskiy ***
The problem of protecting the environment has
taken on special urgency in recent years because of
the scientific and technical revolution taking place in
the world.
Among the many ecological, technological, and
biological aspects of this problem in need of study
and urgent decisions, the most important task is that
of protecting human life, activities, and health
(taking the interests of future generations into con-
sideration). The hygienic aspects of the problem of
protecting the environment therefore now are of para-
mount importance. Here at home the principle of
priority in protecting health, and the preservation of
favorable hygienic living conditions for the popula-
tion, was clearly expressed in a number of resolutions
by the Central Committee of the Communist Party
of the Soviet Union, the Supreme Soviet, and the
Council of Ministers of the USSR on improving pro-
tection of the environment and on the rational use
of natural resources.
Wide-ranging governmental measures aimed at
preventing pollution and sanitation of the environ-
ment in the interests of the population are based on
hygienic norms that set a maximum permissible con-
tent of harmful substances in water, atmospheric air,
and food products. A hygienic norm, or the maxi-
mum permissible concentration (MFC), is that con-
centration of a chemical substance the observance
of which results in no unfavorable effect on the
health, well-being, or working capacity of the popula-
tion, and of future generations, as detectable by
modern research methods, while at the same time
causing no deterioration in the population's hygienic
living conditions.
The development of the principles, and of the
methodology, for establishing the MFC is a major
achievement of Soviet hygiene, the more so because
Soviet scientists (N. V. Lazarev, A. N. Sysin, S. N.
Cherkinskiy, and V. A. Ryazanov) made the basic
contribution to the development of this scientific
trend. The State Sanitation Code of the USSR, which
is designed to protect man's environment, now in-
cludes over 420 norms for atmospheric pollution that
involve the isolated and combined effects of these
substances.
•Professor
••Docent
•*'Senior Scientist
Setting a hygienic norm for chemical factors in the
environment is based on two main principles. One is
the concept of the presence of biological thresholds
in the effect of these factors. The other is the
principle of a limiting harmfulness criterion. Thus,
when the MFC of chemical substances in the waters
of reservoirs is scientifically substantiated, complex
experimental research is conducted, resulting in the
establishment of (a) the threshold concentration in
terms of influence on the organoleptic characteristics
of the water (odor, taste, color, turbidity), (b) the
threshold concentration in terms of the influence on
the overall sanitary regime in the reservoir (for which
only initial changes in the biochemical processes of
the mineralization of organic substances in reservoirs
are detected), (c) the ineffective (subthreshold) con-
centration, which has no unfavorable toxic effect on
the human body.
The concrete value of the MFC is established in
terms of the most sensitive index (the harmfulness
criterion), characterized by the least magnitude.
The USSR has established two norms for atmos-
pheric pollution; the maximum single, and the mean
daily, MFC. The maximum single MFC has been
established to prevent reflex reactions in man during
short-term effects of atmospheric pollution, while the
mean daily MFC has been established to prevent its
resorptive effect.
Experimental research is the basis for setting hy-
gienic norms in the field of general and municipal
hygiene. The importance of experiments stems:
first, from the interests of preventive sanitary in-
spection, a procedure that can implement measures
against chemical pollution of the environment during
the planning stage for new enterprises and plants,
and thereby prevent the unfavorable effect of such
pollution on the population;
second, from the possibility under experimental
conditions of revealing relatively accurately the re-
lationship between each biological effect and the level
of concentration of the chemical substances, some-
thing that is difficult to do under natural conditions
because of the additional effects of all the different
factors present in man's habitat.
63
-------
Experimental research in establishing safe levels
of chemical pollution for man is proceeding in two
versions:
(a) an experiment, using volunteers, to reveal
threshold and subthreshold concentrations by reflex
reactions during brief periods of inhalation of a sub-
stance in terms of its influence on the organoleptic
characteristics of water that is completely safe for
the subjects;
(b) a lengthy, continuous experiment on animals
to establish threshold and subthreshold concentra-
tions in terms of a toxicological criterion of harm-
fulness of the substance.
The first stage in the development of hygienic
normalization in the USSR (the 50s through the 70s)
has been characterized by the accumulation of a
great deal of factual data to justify the MFC for
those chemical factors which are most often the cause
of environmental pollution. Theoretical generaliza-
tions of the accumulated data have begun to be made
in recent years because of the need to carry out the
important task of increasing the reliability of the
MFC that have been established by perfecting the
methodological and theoretical bases for hygienic
normalization.
The problem of the combined and complex effect
of substances is deserving of special attention. We
know that mixtures (combinations) of substances
can cause different effects, synergism, antagonism, an
independent effect, depending on the nature of their
toxic effect and on content level. The experimental
research that has been conducted to estimate the
nature of the combined effect of substances polluting
the air, as well as the reservoirs, has shown that an
additive effect occurred in many cases. However, this
question still is far from resolved because of the ex-
treme complexity of the composition of pollutants,
and the relatively low level of their content in the
environment.
This means that we must make further improve-
ment in the methods used to study the combined ef-
fect when concentrations of substances are at low
levels, as well as to study the principles for estimating
the nature of the summed biological effect of com-
plex mixtures. Experimental research is now in pro-
gress along just such lines in institutes in the USSR.
Even more complicated is estimating the danger of
the complex effect of a substance entering the body
simultaneously over different pathways (from dif-
ferent media). The current practice of establishing
hygienic norms for the same chemical substances
separately for atmospheric air, the air in production
areas, for water in reservoirs, and for food products
fails to consider the possibility of their entering the
body simultaneously over different pathways. This
gives rise to the need to develop methodological ap-
proaches to complex (unified) hygienic normaliza-
tion for chemical compounds in the environment.
In view of the difference in the principles and
criteria for establishing maximum permissible con-
centrations of harmful substances for different media,
what we are talking about at this stage is the develop-
ment of approaches to unified hygienic normalization
for substances, the permissible content of which can
be limited by their resorptive, that is, their toxic,
effect.
The integration of hygienic normalization, or uni-
fied hygienic normalization, in different environ-
mental objects, while an urgent requirement of the
times, is an extremely complex problem, the solution
of which certainly will not be a unique one in all
cases.
We know that a different content often finds its
way into the very concept of "unified hygienic nor-
malization."
It is not our purpose to provide a precise and uni-
versal formulation of this concept. We will simply
point out that at this time we are not including hi
this concept the task of establishing a unified MFC
magnitude for all media. The problem of integration
of hygienic normalization is associated specifically
with answers to the following questions:
1. The development of principles and methods for
estimating the danger and safety of harmful sub-
stances upon their complex admission to the body, as
well as unified approaches to estimating the simul-
taneous pollution of different environmental objects.
2. The development of unified criteria for the
harmfulness and harmlessness of a chemical factor
for the body, and scientific justification for unified
principles and conditions for conducting toxicological
experiments (length of experiments, selection of
doses of substances, research methods, interpretation
of results obtained, selection of safety factors, and so
forth).
3. The carrying out of organizational measures
providing for unified planning of scientific research
in the study of toxic substances as applicable to all
media and the simultaneous review of data obtained
by representatives of the different branches of
science.
The first step could, for example, be the establish-
ment of unified norms protecting the interests of all
branches of the national economy in the rational use
of reservoirs (ensuring sanitary conditions of water
use by the population, the interests of fish protection,
the preservation of hydrobionts, the use of reservoirs
for watering agricultural crops, and so forth). Ac-
cumulated practical experience in the use of sanitary
and fishery MFC, with the orientation being toward
a more rigid norm from these two standards, has
shown the desirability and soundness of this ap-
proach. Unfortunately, future realization of this task
has been made difficult by the lack of methodological
bases for the normalization of MFC magnitudes in
hydrobiology and agriculture.
64
-------
There probably are different ways to estimate the
danger from chemical substances upon their com-
plex admission to the body. However, the establish-
ment of the biological equivalency (the isoeffective-
ness) of doses, and of concentrations, of substances
entering the body over different pathways is, as we
see it, of decisive importance in developing ap-
proaches to unified hygienic normalization.
Various estimates have it that under contemporary
conditions, human health is threatened by several
thousand chemical substances, very few of which
have been studied, relatively speaking. At the same
time, knowledge of the degree of danger to be an-
ticipated from each environmental pollutant is ex-
tremely important.
Despite the success achieved in the field of hy-
gienic normalization for environmental pollution, the
rates at which norms are validated lag sharply behind
practical requirements. This is what increases the
urgency for scientific development of methods for the
accelerated establishment of MFC for chemical pol-
lutions.
Accelerated normalization for harmful substances
can be accomplished by developing calculating and
rapid methods. Formulas for predicating MFC in
atmospheric air, and in the waters in reservoirs, in
terms of the physico-chemical characteristics of the
substances, of the parameters of their toxicity, and of
the norms established for other media, that are based
on correlation and regression analyses, have been
proposed.
Another approach to accelerated validation of hy-
gienic norms is based on a study of the "concentra-
tion-time" and "dose-effect" relationships in a short
experiment with subsequent extrapolation of the data
obtained to an extended period of time.
Research to validate the methodological conditions
for the practical use of such an approach is presently
in progress along these lines at the A. N. Sysin In-
stitute of General and Municipal Hygiene of the
Academy of Medical Sciences of the USSR.
Estimating the danger from chemical compounds,
with consideration given to their intermittent effect,
is an extremely urgent problem in hygienic normali-
zation. We know that the concentration of harmful
substances in the environment is constantly under-
going temporal change, and as experimental research
shows, the intermittent effect can, in some cases
diminish, and in others strengthen toxic effects as
compared to the effect of permanent concentrations.
It is extremely important to the estimate of the
intermittent effect to establish the general behavior
patterns in the development of the toxic effect for
both the permanent and interrupted effects of
chemical compounds, because it is a practical im-
possibility, given experimental conditions, to repro-
duce the picture of the variability of the concentra-
tions of the substances polluting the environment.
First among these general questions that should be
included are the quantitative dependence of the bio-
logical effect on the concentration of the substance
and the duration of its effect, comparative estimates
of the permanent and interrupted effect of harmful
substances on the body, quantitative estimates of the
capability of the substances and the cumulation of
the effect, and so-forth.
Wide-ranging special research in validating the
principles and methods used to study the long-term
consequences (blastomogenic, genetic, gonadotropic,
teratogenic, allergenic) of the effect of chemical pol-
lution is needed to resolve the question of the feasi-
bility of using these data for hygienic normalization.
Questions of the threshold nature of the types of bio-
logical effects of harmful chemical substances in-
dicated still have not been developed sufficiently in
the theory of hygienic normalization. The question of
the feasibility of normalizing carcinogens in the en-
vironment now has been resolved favorably, in prin-
ciple (on the basis of the MFC we can take a dose
that will "cause" an effect beyond the limits of
human life), but the feasibility of normalizing muta-
gens and allergens still is under discussion. Hygienic
science must develop this most important field of
research more effectively.
One of the key questions of hygienic normaliza-
tion involves the problem of extrapolating toxicologi-
cal data from animals to man. A lexicological ex-
periment can be called biological modeling, because
there can be no identity between the bodies of
animals and man, and it is this that makes it neces-
sary to resolve the question of the conditions under
which experimental data can be put into hygienic
practice. The extrapolation methods used in hygienic
research (direct transfer of data, the use of safety
factors, and so forth) are quite empirical in nature.
It has been established that the method that makes
use of direct transfer of toxicological data from
animals, from white rats, for example, to man must
be used with care in order to validate hygienic norms.
The desirability of using "universal" safety factors
also raises doubts, because the value of the coefficient
of extrapolation must be found in terms of the species
of laboratory animals, and of the specific properties
of the toxin being studied. The method used to con-
vert doses per unit of body surface of animals and
man in order to grade species differences with the
toxins also is in need of correlation.
The validation of the conditions under which ex-
perimental data can be extrapolated from animals
to man is attended by great difficulties, but we believe
that study of questions concerned with the compara-
tive sensitivity of man and animals to toxins, and the
recognition of general behavior patterns in species
differences between mammals toward the effect of
chemical pollution, has to be considered a promising
approach to resolving this complex problem.
65
-------
Thus, the development of methodological and
theoretical research is the most important prerequisite
for increasing the scientific validation and reliability
of hygienic norms. It is no accident that in accord-
ance with the program for scientific collaboration
between the USSR and the USA on the problem of
"Environmental Hygiene," the combined efforts of
the scientists of both countries are concentrated on
finding answers to precisely these current questions
of hygienic normalization. The first results of the
joint research show that we now have great potential
possibilities for creating an even more perfect
scientific methodology for perfecting legislation in
the field of protecting the environment, which is the
basis for initiating measures that will improve hy-
gienic conditions.
66
-------
Hygienic Normalization Under Scientific and Technical
Progress Conditions
A. P. Shitskova, M. I. Gusev,* S. M. Pavlenko, I. L. Karagodina
Scientific and technical progress is sharply chang-
ing the nature of the effect of society on nature, and
is bringing about changes in the biosphere. The pri-
mary factors affecting the biosphere are urbanization
and industrial, scientific, and technical progress. They
cause change to take place in the environment, in
atmospheric air, water, soil, nutrition, and in other
things, and the appearance of new factors, such as
vibration, ultrasonics, static electricity, electromag-
netic oscillations, and others.
Control of the biosphere under these conditions is
one of the most important tasks confronting man-
kind.
The classics of Marxism-Leninism considered this
problem to be a social one, right from the very be-
ginning. Control of the biosphere is a conscious
activity on the part of society, aimed at regulating
life processes, including protecting the health and the
biological perfection of existing populations, as well
as of future generations.
The implementation of measures designed to im-
prove the environment requires the development of
quantitative indices for the condition of different
environmental factors and criteria of harmlessness
that ensure optimum conditions for man's (society's)
vital activities under predetermined social conditions.
Hygienic normalization during a period of rapid
technical development is universal for all environ-
mental objects, and is brought about by practical
demands. Scientifically developed quantitative indices
(maximum permissible concentration, maximum per-
missible value, maximum permissible level, and
others) were until recently concerned mainly with
individual substances and environmental factors.
Research in the course of hygienic normalization
for harmful substances includes comprehensive ex-
perimental research methods on animals, observa-
tions of people, and natural investigations of the state
of the environment.
The whole arsenal of modern physiological, bio-
chemical, pathophysical, physicochemical, clinical,
and other research methods is used to study the bio-
logical effect of small concentrations of toxic sub-
• Institute ot Hygiene, Moscow
stances in order to provide hygienic normalization
for these latter. Hygienic normalization, as a rule, is
aimed at establishing threshold and subthreshold con-
centrations of the substances or environmental factors
studied.
The estimate of the interaction between toxic sub-
stances and the organism depends on the dose-time-
effect ratio. This means that all environmental factors
can become pathogenic if their duration and intensity
exceed the threshold level of the effect.
The principles of hygienic normalization have been
perfected with the development of hygienic science
and new tasks, such as recognition of the conse-
quences of multifactor environmental effects on the
organism, have arisen before the researchers in the
modern stage. As a result, in addition to establishing
permissible limits for individual environmental
factors, questions concerned with the study of the
combined effect of toxic substances on the organism
have become currently important. We know that the
simultaneous effect of toxic substance on the organ-
ism can be the same as the effect of each individually
(independent action), can be weaker (antagonism),
or stronger (synergism). Simple summation (addi-
tive action) can take place in the latter case, when
the sum of the concentrations of two or more toxic
substances with a definite effect is equal to unity.
Based on present data, and confirmed by experi-
mental research in the field of municipal hygiene for
atmospheric air and water in reservoirs, there usually
is an additive effect, that is, the effect of summation,
in the case of the combined effect of toxic substances.
It is the opinion of I. V. Sanotskiy that potentia-
tion is extremely rare in the field of industrial hy-
giene when there is a lingering combination of the
effect at the level of the threshold of lingering effect.
Thus, the accumulated data from the experimental
study of the nature of the combined effect of harmful
substances in low concentrations entering the organ-
ism by whatever pathway have made it possible to
formulate principles for the normalization of com-
binations of substances included in the corresponding
legislative resolutions.
67
-------
Hygiene has recently had to study the combined
effect of physical and chemical factors under produc-
tion conditions and in populated areas.
The best studied of the physical factors include
noise, vibration, ultrasonics, and temperature varia-
tions. Their combinations with chemical substances
can be quite varied. The combined effect of physical
and chemical factors differs from their isolated effect
on the organism. Thus, clinical and experimental re-
search has established that local and general vibra-
tion intensifies the toxic effect of heavy metals (N.
A. Gordonova, 1967; L. Ya. Tartakovskaya, 1970).
Experiments on animals have shown that general
vibration combined with manganese, mercury, or
lead has a more pronounced effect on some indices
of the functional state of the central nervous system,
metabolism, and others. New, distinctive, phenom-
ena not noted when the effect is isolated sometimes
will be observed. However, changes pointing to the
attenuation of the mutual effects of different agents
can be detected, along with potentiation. Thus, vibra-
tion definitely moderated the anemia-causing action
of mercury or lead, showing a direct effect on the
hematopoietic processes in the bone marrow (L. Ya.
Tartakovskaya, Ye. N. Fresh, A. G. Mal'kova,
1972).
Manganese exhibits less of an effect on the delicate
structure of the peripheral nerve-muscle endings, and
the regeneration of damaged elements begins much
more quickly as compared to the isolated action of
manganese. The data provided by Ye. M. Neizvest-
naya, L. G. Babushkina, and A. G. Gol'dePman
(1972) show attenuation of the mutual effects of
vibration and MnC2, as well as of silicon in lung
diseases.
The combined effect of ultrasonics and ethanol
intensifies the biological effect obtained from the
isolated effect of these factors. Changes in the central
nervous systems of animals under the repeated effect
of this complex under study were of the same nature
as they were under the effect of ultrasonics alone,
but in a more pronounced form (A. D. Shipachyova,
1969). This leads to the assumption that the physical
factor, that is, the ultrasonics, is the lead component
in this complex. Thus, reactions of the organism to
the combined effect are ambiguous. The reactivity
of organs, and of functional systems, at different
levels can be distinguished by their special features,
and the integral biological effect is not the result of
simple arithmetic addition, but reflects the specific
differences between the reactive structure of the
organism.
Certain common "critical" functional systems
which, in the final analysis, determine the relation-
ship between the processes of pathogenesis in the
course of the response reaction, also can be distin-
guished.
Intensive scientific and technical progress has
greatly complicated the questions confronting hy-
giene in recent years. The need arose to develop new
methodological approaches to making a complex
estimate of the effect of environmental factors. To-
day, research is in progress along these lines to arrive
at complex hygienic normalization of pesticides in
environmental objects (L. I. Metdved', Ye. I. Spynu,
et al., 1969). Some calculations of the maximum
harmless dose for the simultaneous entry of sub-
stances from different media have been proposed (A.
I. Korbakova, N. I. Shumskaya, et al., 1971).
One of the possible methodological variants in de-
veloping approaches to complex hygienic normaliza-
tion can be the use of graphic calculation methods
(N. I. Sidorenko, N. A. Pinigin, 1971, 1972).
Attention also is being given to the experimental
study of the special features of the complex effect of
toxic substances in the environment. The complex
effect of the substances cannot be equated to the
combined effect without the corresponding experi-
mental verification. It can be assumed that the com-
plex effect will have its own special features. This is
so primarily because the same chemical substances
will act on the different receptor fields in a manner
that will depend on the pathway over which they
entered the body (R. V. Anichkov, 1955; A. G.
Bukhtiyarov, 1955), and will differ in the nature and
rate of adsorption, conversion, and excretion from
the body (Ya. A. Teysinger, 1959; I. L. Gadaskina,
et al., 1971).
Special experimental studies were made at the F.
F. Erisman Institute to study the general behavior
patterns of this effect by the most widely used
chemical substances found in organic synthesis pro-
duction, such as ethanol, cyclohexanone, benzene,
methanol, formaldehyde, carbon tetrachloride, and
other nonelectrolytic toxins (S. M. Pavlenko, V. A.
Guseva).
The scheme developed for studying the complex
effect of substances in a continuing experiment pro-
vided for revealing the effect of the toxins at the level
of threshold and inactive doses and concentrations.
Maximum attention was given to the selection of
common, pathogenetically meaningful, and the most
sensitive, specific tests and functional loads in order
to seek out the special features of the effect of the
same substances when body entry was over peroral,
inhalational, and complex pathways.
It has been shown by the example of the study
made of the behavior patterns involved in intoxica-
tion by nonelectrolytic substances that at the basis
of the origin of their chronic intoxication are me-
chanisms that affect the central nervous system, as
well as some of the metabolic processes, regardless
of the entry pathway, although the formation of
chronic intoxication did proceed differently. All
response reactions occurred earlier for most of the
substances when entry was via a complex pathway
than when the pathway was inhalational or peroral.
68
-------
However, the toxicity of the nonelectrolytes
studied changed unidirectionally upon complex intro-
duction in the continuous experiment, despite the
fact that benzene, alcohols, cyclohexanone, and other
of the substances studied are opposite in type of
effect from industrial narcotic-toxins, which belong
to different groups in the system of nonelectrolytes.
We, in our studies, paid special attention to dis-
covering the breakdown in the body's adaptive mech-
anisms (adaptations) to the effect of the toxic sub-
stances studied when they entered via different path-
ways, using a set of functional tests. A comparative
estimate of a series of integral and specific functional
loads (static efficiency, a cold test, an alcohol test,
injection of a decisive dose of the substance itself,
anodization of the brain, and others) was made for
this purpose.
The experiments showed that in the presence of
even negligible symptoms of intoxication, all the
functional loads used yield a good effect, revealing
any change in the functional activity of the organism.
So far as the stage of apparent normalization is con-
cerned, only specific pathogenetically meaningful
functional loads reveal the latent effect of the toxin.
The specific functional test for substances with a
narcotic effect is anodization of the brain.
The studies made showed that the formation of
chronic intoxication proceeded differently when in-
dustrial nonelectrolytic toxins at the threshold level
entered over different pathways.
Basic changes in the indices for the functional
state of the organs and systems were noted only
during the first three to four months of exposure
when entry of substances was complex, as opposed
to separately (inhalation or peroral).
The state of nonspecifically increased resistivity
appears in the first weeks of the experiment. The
stage of compensation for the pathological process
most usually takes three to four months to develop.
Data also were obtained which indicate the possibility
of the development in this case of latent phases of in-
toxication from the toxins studied in shorter periods
of time as compared to the separate effects of these
same substances. This fact is of scientific interest,
because it makes it possible to correctly estimate the
danger of chronic intoxication, with the complex
effect of environmental factors taken into considera-
tion.
Study of the intoxication mechanisms makes it
possible to approach the development of unified
methods for estimating the biological effect of toxins
(nonelectrolytes) when they enter over different
pathways (peroral, inhalational, and complex).
Generalization of the experimental data helps re-
veal the general behavior pattern in the reaction of
the organism to the simultaneous entrance of harm-
ful organic substances (nonelectrolytes) over dif-
ferent pathways, and usually is expressed in the effect
of the summing of their effect.
Along with the study of the complex and combined
effects of substances on the organism, attention re-
cently has been concentrated on the development of
methodological questions with a view to ascertaining
the behavior patterns that could lead to integrating
the MFC for toxic substances simultaneously in food
products, water in reservoirs, atmospheric air, and
other environmental objects. Calculation methods
based on ascertaining the correlation relationship
between different parameters of toxicity and
chemical constants were developed in this connec-
tion.
The development of the multifactor relationship
between the parameters of toxicity of acute and
chronic effects and the physicochemical indices for
substances normed in food products, water in reser-
voirs, atmospheric air, and in the area of working
spaces, is the basis of mathematical modeling re-
search.
The use of parametric and nonparametric analysis
methods has made it possible to expose the correla-
tion relationship between the above-indicated indices,
based on which calculation formulas have been ob-
tained for use in finding the tentative value of the
permissible content of toxins in different media.
On the other hand, the development of accelerated
methods for establishing the MFC for substances in
various environmental objects, based on a study of
the special features of the biological effect of toxins
and the exposure of common pathogenetically mean-
ingful indices of toxicity for different entry pathways,
was begun in a combined experiment. This approach
permits the development of unified methods for es-
timating the biological effect of chemical substances
and makes it possible to conduct studies simultane-
ously, and in parallel, that deal with hygienic nor-
malization of toxic substances in different media.
Our data show that predicating the harmless level
of chemical compounds in different environmental
objects while combining the use of mathematical
modeling and detailed study of the behavior patterns
of the biological effect of toxic substances is a promis-
ing methodological direction to take in the integral
approach to hygienic normalization.
Scientific and technical progress, created by man,
cannot be considered as something that is alien to
the biosphere. It simply is a qualitatively new stage
in its development, one that can be regulated from
the point of view of the interrelations between man
and nature. Therefore, we ought not consider the
presence of changes occurring in the environment
as a fatal inevitability accompanied only by negative
effects on the organism. It would be more correct to
consider environmental pollution (of atmospheric
air, water, soil, and others) as a phenomenon occur-
ring as a result of what still are imperfect techno-
logical processes. All the conditions needed to per-
fect these processes are available in the socialist
-------
society. Mankind today is at the stage of a gradual,
intelligent, transformation of the biosphere into a
zoosphere. Therefore, the activity of the hygienists
during the period of the scientific and technical rev-
olution should be directed at studying the effect of
the newly-forming environmental factors on the or-
ganism on a new scientific methodological base. This
permits developing multiprofile complex indices in
accordance with which it will be possible to correlate
the permissible parameters for forecasting the en-
vironment, working conditions, living conditions, and
the health of the population; to protect man's health
heredity against the defective action of mutations;
to preserve existing genetic information; and, by
social reorganization, to promote the creation of a
healthy, modern, society.
In his report to the 24th Congress of the CPSU,
L. I. Brezhnev pointed out that the consolidation of
the achievements of scientific and technical progress
and of socialist management of the economy in our
country permits the establishment of rational forms
between nature and society.
70
-------
Maximum Permissible Human Stress1
VaunA. Newill **
I. INTRODUCTION
Formulation of sound environmental policies de-
pends upon the availability of adequate information
upon which to base regulatory decisions. Current en-
vironmental health research must provide data for
adequate assessment of health risks that arise from
environmental pollutant exposure. Pollution abate-
ment to reduce or negate these health risks requires
financial expenditures and possibly even lifestyle
changes. Figure 1 is a simplified schematic diagram
that presents several of the principal factors in the
decision-making process concerning air pollution
control. It indicates that the degree of health protec-
tion attained is a function of pollution control costs.
The minimum acceptable level of health protection
is, at the very least, that level necessary to protect
from death; and the immediate regulatory action
program adopted for air pollution control should
certainly protect from illness as well.1
The degree of health protection to be selected
above the minimum acceptable level from Figure 1
is a matter for political decision. The appropriate
authorities must decide on the level of health pro-
tection desirable for their society. Increments of
health protection above the minimum acceptable
level are generally purchased at ever increasing incre-
ments in control costs. The zone in which increased
health protection (benefit) is obtained at increasing
control costs (the cross hatched area in Figure 1)
can also be thought of as the region of social decision
making. The level of health protection desired must,
of course, take into account the existing air pollution
effects, but other considerations are also important,
including general social, cultural and economic
factors, as well as the magnitude of other health
problems. In deciding where to spend limited money
resources, the total set of risks that man must bear
need to be considered and the greatest efforts should
be expended to reduce the largest risks, be they en-
vironmental or other societal problems.
Each regulatory action is taken for society's bene-
fit. However, each action also has a societal dis-
benefit because it consumes a certain quantity of the
POSSIBLE MAXIMUM
PRACTICAL MAXIMUM
kCCEPTABL
"TGION IN WHICH
WIINIMUM^COSTS ARE BALANCED
AGAINST RESULTING
FITS
NIL
COST OF CONTROL OF AIR POLLUTION
SOURCE: BASED ON WLD. HLTH. ORG. TECHN. REP. SER., 1972, No. 506
FIGURE 1. DEGREE OF HEALTH PROTEC-
TION VS. COST OF AIR
POLLUTION CONTROL.
society's resources for its accomplishment. Obviously
there must be some give and take between the
societal benefits and disbenefits. Decisions involving
this process are difficult because the group to bene-
fit may be different from the group that receives the
disbenefits and thus, it is difficult to make this an
equitable process.
Another important concept follows from further
consideration of Figure 1. Infinite health protection
is unattainable, thus society must either decide what
level of health risk is acceptable, a decision that will
be reflected in the societal financial investment in
environmental pollution control; or the decision will
be made indirectly since the societal financial invest-
ment in environmental pollution control will dictate
the level of health protection to be achieved.
II. ENVIRONMENTAL PROBLEM
ASSESSMENT
Environmental problems are identified in two gen-
eral ways. One is by focusing on the environmental
factors (agents) to which the population is exposed.
•Much of the content of this paper is based on internal EPA reports. Only those from which many passages were taken are cited. References
other than EPA reports are cited.
••Special Assistant to the .Administrator, U.S. Environmental Protection Agency
71
-------
Another is by focusing on disease determinants and
mechanisms.
By such considerations many problem areas are
identified. To decide which problem areas are the
most important some criteria must be applied.
Factors considered in placing the problems into a
priority order for action include such considerations
as: (1) characteristics of the pollutant, e.g., ubiquity,
expected levels of exposure to the population, innate
toxicity, bioaccumulation and persistence; and (2)
the pollutant's potential for playing an important role
in public health problems, e.g., affecting the fre-
quency, severity or trend pattern of a specific disease.
Pollutants suspected as being related to diseases that
are common, severe and increasing in prevalance,
should gain the highest priority for regulatory ac-
tion.2
Now let's consider some issues in the detailed
assessment of specific problems.
Integrated information from several disciplines is
useful in the health assessment of these problems.
The major disciplines involved are epidemiology,
clinical research and the whole set of experimental
studies that will be included under the term toxi-
cology, being fully aware that many other disciplines
are involved in these studies.
Each of these discipline approaches is useful. A
well-designed program .integrating the unique capa-
bilities and the crosswalks between these disciplines
can provide a most satisfactory basis for regulatory
action. Epidemiology can provide studies of popula-
tion (communities or disease groups, current or long-
term) exposure in real life settings. The advantages
of epidemiology are: the exposure is natural, there is
no need for extrapolation of data to the human, the
most vulnerable groups in the population can be
studied, and both current and long-term low-level
exposures can be evaluated. The major problems
relate to quantifying the exposure, to dealing with the
many covariates, to obtaining dose-response data and
to deciding association vs. causation.
Clinical research studies can be used to gather
human data on either normal or diseased persons
regarding absorption, metabolism and excretion of
pollutants and can be used for indepth studies of
humans accidently exposed to high levels of pol-
lutants. These accidently-exposed humans provide
an opportunity to develop new study parameters and
response indicators. The advantages of clinical
studies are: the pollutant exposure is controlled so
that improved dose measurements are obtained; since
each person can usually be used as his own control,
covariates are well controlled; vulnerable (highly
susceptible) subjects can be included in the studies;
cause-effect relationships are more easily ascertained;
there is no need for extrapolation to humans; and
thus the derived information gives maximum input
into standards. The problems are that the exposure
is artificial, there can be no long-term exposures and
thus only acute effects are determined, and there
can be real hazard to the exposed persons.
Toxicologic studies can use many response sys-
tems, such as whole animal, isolated organs, cells or
biochemical systems. The advantages of toxicologic
studies are: maximum dose-response data can be
obtained, though this information is incomplete at
the low end of the curve; data acquisition is rapid;
cause-effect relationships are more sure; and mech-
anism of response studies, such as kinetics of pol-
lutant absorption, distribution, metabolism and ex-
cretion, can be performed. While known quantitative
exposure requirements are most easily satisfied under
controlled or experimental exposure settings, appro-
priate human disease models in animals have not
been available and thus have not been evaluated in
this manner. Furthermore, laboratory experimental
studies cannot provide complete assurance of dose
effects because community exposures cannot be dup-
licated in the laboratory. In addition, the difficulties
of extrapolating the data to the human remain,
particularly estimating the threshold of human re^
sponse.
Pollutants, Response and Human Exposure
"The effects of pollutants on human health de-
pend on the physical and chemical properties of the
pollutant, on the duration, concentration and route
of exposure and on the human uptake and metab-
olism of the pollutant. Man's biological response is
likewise a function of occupational, psychosocial and
climatologic factors and is tempered by the phe-
nomvna of tolerance and adaptation. These exposure
factors underlie attempts to understand the impact
of pollutant exposure on human health.
"The physical and chemical properties of pol-
lutants determine their potential as a health hazard.
These properties—including size, density, viscosity,
shape, electrical charge, volatility, solubility and
chemical reactivity—all affect the absorption, reten-
tion and toxicity of the pollutants. Many pollutants
do not retain their original identities after entering
the environment. Thermal, chemical and photo-
chemical reactions occur when pollutants move
through the environment from source to receptor.
These factors affect the final physical and chemical
state at the point of human exposure and help de-
termine the toxic potential of the pollutants." •
The Response Spectrum3
Environmental pollutants can affect the health of
individuals or communities over a broad range of
biological responses. One can conveniently think of
five biological response stages of increasing severity
72
-------
as illustrated in Figure 2: (1) a tissue pollutant bur-
den not associated with other biological changes,
(2) physiologic or metabolic changes of uncertain
significance, (3) physiologic or metabolic changes
that are clear-cut disease sentinels, (4) morbidity or
disease, and (5) mortality. Boundaries between cate-
gories may occasionally overlap. Furthermore, each
category shows a range of responses rather than a
simple all-or-none phenomenon.
At any point in time more severe effects, such as
death or chronic disease, will be manifest in rela-
tively small proportions of the population. In very
few cases can death or disease be attributed directly
and solely to pollutant exposure. Death and disease
are end products of repeated cumulative insults
(cumulative risks) from sources such as diet, ciga-
rette smoking, physical inactivity, infectious chal-
lenges, and accidental injury. In general, the role of
environmental contaminants in the mortality or mor-
bidity experience of a community is difficult to
quantify because so many other determinants of
death and disease cannot be adequately measured.
The lower levels of the response spectrum shown
in Figure 2 are subclinical manifestations of pol-
lutant exposures. Larger portions of the population
are affected to these levels of response. Patho-
physiologic responses such as impaired mucociliary
clearance and bronchoconstriction, and physiologic
changes of uncertain significance, such as neuro-
behavoral responses, are more adaptable to experi-
mental studies on animals or humans than is the
case with acute or chronic disease and can be more
readily associated with specific pollutant exposures.
Pollutant burdens are tissue residues resulting from
pollutant exposure. Pollutant burdens are highly
specific effects of exposure, can be readily quantified
in population studies and may be used as indicators
of environmental quality. If the bridge between the
lower and higher levels of the response spectrum can
be established, the disease risk associated with pol-
lutant burdens or physiologic changes can be shown,
and ultimately the role of pollutant exposure in the
total community morbidity and mortality experience
can be defined.
ADVERSE
HEALTH
EFFECTS
PATHOPHYSIOLOGIC
CHANGES
PHYSIOLOGIC CHANGES OF
UNCERTAIN SIGNIFICANCE
POLLUTANT BURDENS
PROPORTION OF POPULATION AFFECTED
SOURCE: PROCEEDINGS OF THE CONFERENCE ON HEALTH EFFECTS OF
AIR POLLUTANTS, OCT. 3-5,1973; COMMITTEE ON PUBLIC
WORKS, UNITED STATES SENATE 93-15; p. 646.
FIGURE 2. SPECTRUM OF BIOLOGICAL RESPONSE TO POLLUTANT EXPOSURE
73
-------
Some groups within the population may be espe-
cially susceptible to environmental factors. Notably
these include the very young, the very old, and those
affected by a disease. Susceptibility, however, may
be temporary as well as permanent. Inherited ab-
normalities such as alpha antitrypsin deficiency and
abnormal hemoglobins are examples of permanently
altered sensitivity. Temporary increased sensitivity
may be associated with periods of growth, with
weight reduction, with pregnancy and with reversible
illnesses.3
Diseases commonly result from complex causal
webs rather than single factors.4 Environmental pol-
lution may contribute a number of strands to such
webs. Other strands may arise from such diverse
origins as genetic heritage, nutritional status and
personal habits. Moreover, pollutant exposure may
alter the severity of disease without altering its fre-
quency.3
Exposure Response Matrix
The duration and concentration of pollutant ex-
posure are measures of the total dose to the human.
It must be remembered that the rate at which the
total dose is received may influence response.
Health effects of an environmental pollutant may
be either short-lived (acute) or relatively permanent
and irreversible (chronic). Acute or chronic effects
may occur after a single exposure to a hazardous
substance. This can be illustrated by acute radiation
exposure which can cause acute radiation gastro-
enteritis and chronic leukemia. An air pollution epi-
sode may have similar effects though the permanent
sequelae of acute episodes have never been ade-
quately studied. Likewise, acute and chronic effects
can result from long-term exposure. For example,
excess acute respiratory illness and chronic respira-
tory disease have been repeatedly demonstrated in
high air pollution exposure cities. Acute effects and
short-term exposures are less difficult to study than
chronic effects or long-term exposure. Moreover, the
effects of dose rate, i.e., large dose in a short interval
vs. repeated small doses over a long period, have
seldom been investigated systematically. Little effort
has been expended on the monitoring of long-term
exposure and disentangling the causal webs under-
lying chronic effects of pollutant exposure. Such
effort has been hampered, however, by methodologic
difficulties.
Evaluating Exposure-Response Relationships
Given adequate characterization of exposures and
health effects, many additional considerations bear
on the scientific validity of the exposure-response
relationship. Hill5 has given an exposition of criteria
that can be used to judge whether an observed
exposure-disease relationship is causal. Hill's criteria
were developed as guides for occupational health
studies. With minimal modification, however, they
can be applied to general population studies. These
criteria are: Consistency of observed associations
coherence of results, plausibility of the association
and strength of the association. Brief descriptions of
each of these as well as exposure response gradient,
intervention, and control of covariates are included.
1. Consistency of Observed Association:
Consistency of observed association is, perhaps,
the most important criterion. Does the health effect
occur in various age, sex and race groups? Has the
effect been repeatedly observed in different places,
circumstances and times? Even small differences that
are not quite statistically significant bear great weight
in standard setting when the criterion of consistency
is met. When the same effect is observed in a variety
of population groups under varying conditions and at
different times, the likelihood of a constant error or
fallacy becomes progressively less.
2. Coherence of Results:
When animal studies, experimental human ex-
posures, and epidemiologic data are coherent, i.e.,
they all demonstrate the same or similar health ef-
fects of exposure, the bits and pieces of evidence,
when brought together, form a mosaic of health
intelligence. By reinforcing results obtained in one
research approach with studies in another, a co-
herent health intelligence system will provide scien-
tifically strong and readily acceptable guides for
costly air quality controls.
3. Plausibility of the Association:
Initial studies may uncover unexpected relation-
ships between exposures and effects. Mere statistical
associations must be rejected when no reasonable
biological explanation, based on experimental evi-
dence, can justfy the association. On the other hand,
when experimentation points to a disturbed physio-
logic process which may lead to some clinical mani-
festation, subsequent epidemiologic studies designed
to test such hypotheses are well grounded in biologic
plausibility. New exposure-response associations re-
quire support from other disciplines before causal
inferences can be readily defended.
4. Strength of the Association:
When disease frequency is nine to tenfold greater
in exposed than in non-exposed populations, the
exposure-effect relationship is extremely strong. Re-
lationships of this magnitude have been found in
studies of cigarette smoking and respiratory disease
including lung cancer and bronchitis. However, pol-
lutant exposures are generally less intense and less
reactive than cigarette smoke inhaled deeply into the
lungs. Differences in exposure between high and
low pollution neighborhoods seldom exceed factors
74
-------
of two or threefold in concentration. At present
ambient air concentrations, relative differences in
effects between high and low exposure areas are un-
likely to be as striking as ratios observed in smokers
vs. nonsmokers. However, for common and frequent
disease events such as acute respiratory disease,
relatively small differences in disease experience can
have a large and costly impact.
5. Exposure-Response Gradients:
When a stepwise increase in exposure can be
associated wit ha stepwise increase in the frequency
of the adverse health effect, the evidence is strong
for a cause-effect relationship. Linear relationships
over an exposure gradient became increasingly diffi-
cult to explain by third intervening variables. Re-
sponse gradients can be investigated in relation to
exposure gradients across geographic areas, differ-
ences in length of residence in high exposure areas,
and migration gradients constructed from various
combinations of childhood and adult exposures of
the same individuals. Human exposures occur nat-
urally over an exposure gradient, especially over time
and across areas. Exposure-response gradients ob-
viously can be easily created in experimental settings.
6. Intervention:
Protection of health requires society to intervene
in public exposures to air pollution. This interven-
tion is largely based on observed exposure-health
effects associations. As desirable air quality is
achieved, will the frequency of adverse health effects
be affected? If so, the causal nature of the exposure-
response association is strongly supported. The high
cost of air pollution control warrants a national pro-
gram of community health and environmental sur-
veillance in those places where achievement of re-
quired air quality will require vigorous abatement
efforts.
7. Control of Covariates:
The causal nature of an exposure-response as-
sociation is convincingly exposed when, after the
effects of known covariates are first displayed, in-
creased disease risk within covariate classes can be
clearly demonstrated in high exposure populations.
For example, in any studies of chronic bronchitis,
prevalence, smokers and males show more disease
than nonsmokers and females respectively. An air
pollution-chronic bronchitis study should reveal the
above smoking-sex differences in prevalence rates,
thereby assuring readers that the study has internal
consistency. If smoking-sex specific groups in high
exposure neighborhoods have excess chronic bron-
chitis, the hypothesis that air pollution exposure
causes excess chronic bronchitis is considerably more
convincing than if the smoking-sex covariates were
not analyzed. Most epidemiologic studies of air pol-
lution require similar analysis of excess disease risk
within covariate categories, with particular attention
given to age, sex, smoking, socioeconomic level and
duration of residence at current location. When
covariates are systematically analyzed for relation-
ships to the health indicator under study, residual
excesses in disease frequency can be attributed to
area differences in pollution exposure with a reason-
able degree of confidence.6
III. USE OF HEALTH RESEARCH
DATA
The need for regulatory action is predicated on
the demonstration of an effect at the levels of a pol-
lutant to which the popular is or can be predicted
to be exposed. The kinds of control options available
to prevent such effects and to choose among are en-
vironmental standard setting, registration, certifica-
tion, licensing, limiting use, banning and economic
incentives such as effluent charges or taxes. For sim-
plicity, the only option that will be address is en-
vironmental standard setting.
Environmental Standard Setting
An environmental standard in itself is a complex
issue. The standard is the maximum permissible
level of exposure of the public to a pollutant per-
mitted for a specified period of time. For a pollutant
that is only present in the air, the environmental
standard is the primary ambient air quality standard.
However, when the pollutant is one where human
exposure comes not only from the air, but from the
food and water as well, the environmental standards
become a set of standards, each representing an allo-
cation of a portion of the permissible exposure
through each of the media from which exposure
comes. No such environmental standards have been
set in the U.S.A. to date.7
The actual permissible level of exposure permitted
must be decided in some philosophical framework.
At the present time the -framework varies somewhat
by medium of exposure and by effect of exposure, in
some instances by legislative mandate, and in others
by differing beliefs of those responsible for the pro-
gram. It certain is the scientist's role to be in-
terested in these problems and to supply a coherent
and rational basis for the framework within which
regulatory decision-making can proceed.
Certain other concepts are useful to consider when
discussing standard setting, such as threshold dose,
dose-response, extrapolation of data from animal to
man, and combined effects.
Threshold Dose
A biological effect may not be observed until the
dose reaches a certain level. This level is called the
75
-------
threshold dose and it may be defined as, the mini-
mum dose required to produce a detectable effect.
The concept is important from both a practical and
a theoretical point of view, since a true threshold
implies that below a given dose there is no adverse
change whatsover with regard to toxic effect of the
substance studied; this allows a safe limit or standard
to be specified. As one uses more sensitive responses,
thresholds decrease. For example, illness frequently
is a more sensitive measure of response than death,
and both will be preceded by physiologic changes
heralding the onset of disease. For many pollutants,
such as ionizing radiation, there may be no threshold
for the response.3
The existence of a threshold may be due in part to
the build-up in man of tolerance to the particular pol-
lutant. Tolerance represents the ability of man to
endure pollutant exposures without apparent ill ef-
fects. The level of tolerance to environmental agents
may be directly related to a number of characteristics
including age, sex, and nutritional state. The concept
of adaptation signifies an increase in tolerance with
long-term low-level exposure to a given adverse en-
vironment. Adaptation is characteristically related
to the stressful components of the environment. The
ability to adapt varies in a population and is deter-
mined by anatomic, physiologic and biochemical
characteristics of individual organisms.
"Safety"
Operationally the threshold concept is being used
in standard setting activities in the United States
regulatory agencies. However, many persons, finding
that definitions for words such as "safe level," "no
effect level" and "virtual safety" (among others) are
being used to satisfy their intellectual recognition
that such absolutes do not exist, are unhappy be-
cause the use of these terms gives the general public
the implication of safety as an absolute. Thus, the
concept of acceptable risk is being introduced and
hopefully environmental agents soon will be assessed
in terms of relative risk rather than as safe or unsafe.
Dose Response
Dose-response data extracted from either com-
munity or experimental studies is crucial for the de-
cisions that must be made. However, when the infor-
mation is available, as it rarely is, it has been gath-
ered in animals at rather high exposure levels and
needs to be extrapolated to man who will be ex-
posed at lower levels. Man, however, will generally
be exposed in greater numbers and for longer periods
of time so that low-level risk can be manifest. We
urgently need more specific research into the issues
of extrapolation so that better decisions can be made.
Since dose-response information is frequently lack-
ing, other techniques need to be considered.
Extrapolation8
Studies of laboratory animals have been used to
assure that the toxicity risk from an environmental
chemical is acceptably low. Such assurance should
be given before a chemical is introduced into the
environment in a fashion where there can be human
exposure. However, relating animal studies to pos-
sible effects in man poses difficult and practical
problems. As more chemicals for which there is
sufficient toxicity testing are introduced into the
environment where man can have similar exposure,
we can learn more about the validity of the testing.
Until that time we rely on such information as
". . . cancers in laboratory animals are essentially
the same as human cancers and with a single pos-
sible exception, all known human carcinogens are
carcinogens in laboratory animals, laboratory animal
studies seem to predict for man. Thus, for ethical
and practical reasons, data derived from use of
laboratory animals for toxicity testing is the foun-
dation of efforts to protect the public from the pos-
sible harmful effects of new and old chemicals in
the environment."
The normal sequence of animal tests of a new
chemical agent begins with studies to determine the
mechanism through which laboratory animals re-
spond to a compound, the nature of metabolites, the
distribution of the parent compound, the metabolism
in the tissues and organs, and the rates and routes
of elimination of the compound or its derivatives.
Comparison of the similarity of the results of these
tests can be made among various animal species,
thus laying the groundwork for prediction of the
exposure events in man. Attempting to systematize
this process is a prime area for research. Also, it
must be remembered that this approach is most
useful for observing effects that occur soon after
the compound is administered, but is less useful for
observing subtle long-term effects.
When attempting to bridge the gap between non-
human and human toxicity one must recognize two
steps involved in the extrapolation of laboratory
toxicity data to man. First, the extrapolation of what
might be called the average or median mouse to the
average or median man. This consists of extrapola-
tion from the average or even the most sensitive one,
two or three laboratory animal species to an average
man living under conditions with a commonly en-
countered genetic make-up. The second step requires
identifying and allowing for variability in the human
population. The median mouse to median man extra-
polation is the easiest step. Biomedical research re-
sults are beginning to illuminate the differences and
similarities between species and are beginning to
76
-------
provide a rational basis for extrapolation. The sec-
ond step, allowing for variability and diversity—
both genetic and environmental, is much harder.
There is, then, a basis for comparison of the
median mouse to the median rat to the median dog
to the median man. More and more patterns that are
useful for extrapolation to man are being recognized
and can be identified in the course of the study of
the pharmacological disposition of a chemical. Most
of the differences that have been observed suggest
that man is more sensitive than the usual experi-
mental animal, and this must be kept in mind when
developing safety factors.
The following principles were stated in the Wein-
house and Rail8 communication as this subject re-
lated to carcinogenesis testing:
1. The major principle that must be accepted if
we are to deal at all with long-term assessments of
toxicity in man is that effects in animals, properly
qualified, are applicable to man.
2. Methodologies do not now exist to establish a
safe threshold for long-term exposure to toxic agents.
3. The exposure of experimental animals to toxic
agents in high doses is a necessary and valid method
of assessing carcinogenetic hazard to man.
4. Agents should be assessed in terms of relative
risks rather than as safe or unsafe.
Combined Pollutant Effects
Physical-chemical interactions occur among the
pollutants in the environment which in turn alter
the biological activity of the contaminants as well as
the reactions within the tissues of man and animals.
The resulting bioeffects may be synergistic, additive
or antagonistic, resulting in a reaction whose magni-
tude is greater, equal to, or less than the sum of
the individual constituents.
One of the most recent and best demonstrations
of synergism of two pollutants was presented by
Bates and Hazucha in October, 1973.9 Figure 3
shows the effect on four measures of pulmonary
function of 0.37 ppm of SO2 and 0.37 ppm of ozone
exposure independently, and then the enhancement
of effects when both gases are present simultane-
ously. It is interesting to note that the effect of the
two gases administered simultaneously is greater
than the sum of the effect of each individually; thus
a synergistic effect has been demonstrated that might
account for episodes in the U.S.A., Netherlands and
Japan.
Least Case and Worst Case Range Estimates6
Derivation of range estimates based on least case
and worst case assumptions is another way to de-
velop a workable plan upon which action decisions
can be based. In least case assumptions, quantitative
assessment of exposure or response is made on the
assumption that current exposures are not repre-
sentative of long-term trends or that adverse health
responses should first be attributed to all known
covariates, and only residual excesses of illness fre-
quency be attributed to pollutant exposure. For
example, in the frequently-encountered situation
where high exposure and low socioeconomic status
geographically coincide, the effect of low economic
level on illness frequency would be identified first.
Any excess illness which could not be accounted for
by economic level would be quantified as a residual
effect. After all covariates were considered in turn,
the final residual excess would be called an air
pollution-related health effect. Least case estimates
attribute the smallest possible effect to pollutant ex-
posures, do not allow for interaction between co-
variates and exposure, and give a maximum quanti-
tative estimate of human exposures associated with
adverse responses.
Worst case assumptions attribute adverse health
effects first only to covariates which are known to
be strong determinants of disease frequency, such
as cigarette smoking in relation to chronic bron-
chitis. But covariates which are not well founded as
determinants of illness frequency are eliminated from
final analysis, and air pollution exposure is assumed
to have contributed to the relatively larger residual
in excess illness frequency. Likewise, if information
on past exposures is of low quality or unavailable,
current exposures are assumed to represent past ex-
perience, and chronic disease frequencies are calcu-
lated as a function of exposures at current levels.
Worst case assumptions, therefore, give minimum
estimates of exposures associated with adverse health
responses, and tend to maximize the proportion of
disease frequency attributable to pollutant exposure.
The truth may lie at either end of the quantitative
range which can be derived from least case and
worse case assumptions. When health intelligence
cannot give precise quantitative information, the de-
cision maker should be provided with least and worst
case range estimates. At this point, the degree of
control becomes a function of other policy con-
siderations, including control costs, alternate control
strategies (and the health effects of these), the
severity of magnitude of the effect, the population
at risk, etc. Failure to present range estimates leaves
less room for control options and forces decisions
based on one set of numbers derived from arbitrary
interpretations about study results.
Least Cost Protective Standards'
Air quality criteria supply the existing rationale
for standards. The process of standard setting re-
quires policy decisions concerning protection of pub-
lic health at least cost to society. Sound quantitative
exposure response data allows the policy maker to
make clear-cut decisions which can be defended.
77
-------
H
8
100
90
SO
70
90
80
70
60
50
0
SOURCE:
0.37 ppm SO2 (N = 4)
0.37 ppm O3 (N = 3)
0.37 ppm S02+O3 (N = 4)
FVC
EXPOSURE
RECOVERY
MMFR
REV,
EXPOSURE
RECOVERY.
2.0
2.5 0
0.5 1.0 1.5
TIME, hr
Proceedings of the Conference on Health Effects of Air Pollution, Oct. 3-5, 1973; Committee on
Public Works, United States Senate 93-15, p. 534.
This figure shows the effect on four measures of pulmonary function of 0.37 parts per million of
SO2 and 0.37 parts per million of ozone independently, and then the enhancement of effect
when both gases are present simultaneously.
100
90
80
70
60
; 50
100
90
80
70
60
50
FIGURES. EFFECT OF INDEPENDENT AND SIMULTANEOUS EXPOSURES TO SO2 AND O3
When range estimates are broad due to inadequate
health intelligence one of two possible kinds of error
may be made. The resulting standards could be too
lenient and not protective enough or they could be
too strict and excessively costly for the real benefit
derived. Thus, society must pay a price for not gen-
erating adequate environmental health information.
It must also be remembered that generation of health
information itself can be costly and that in some in-
stances it may be less costly to overcontrol than de-
velop the specific information required.
Margins of safety built into standards are judg-
ments required to bridge the gap between the inade-
quacy of health intelligence and the need to stop
continued exposure to hazardous pollutant levels.
Greater degrees of uncertainity generate larger safety
margins. Unfortunately control costs usually increase
exponentially as exposure is limited more severely.
The health benefit from the additional degree of
control may be insignificant or actually negative if
overly stringent controls cause health disbenefits, as
could occur if scarce resources are diverted from
health care or protection to unnecessary pollution
abatement.
The function of a health intelligence system in the
standard setting process can be graphically displayed
in the form of marginal cost curves (Figure 4). The
true social cost of pollution is equal to the sum of
the cost of control and the health and welfare cost of
exposure. Health costs of disease, and more espe-
cially of physiological dysfunction, cannot be readily
estimated with current scientific knowledge. Health
and welfare costs increase with higher exposures,
while control costs increase as exposures are lowered.
True social costs of pollution are minimal where the
two marginal cost curves intersect. If society will
78
-------
tolerate some pollutant-associated excess health
costs, curve AD (Figure 4) may represent this social
preference. On the other hand, if society places a
very high premium on preventable disease or reduc-
ing risk of disease, curve A'D' (Figure 4) may well
represent this attitude. On both curves, the points
B or B' represent the lowest pollutant exposure as-
sociated with adverse health (or welfare) effects.
Under the low health cost assumption (curve AD),
society will accept the health costs (and associated
adverse health effects) represented by the vertical
distance between B and C. Under the high cost as-
sumption, society will not allow any adverse health
effect caused by exposure and will require control to
point B', where high marginal cost of control is
equated with derived health benefit.
Excessively stringent standards are shown at
points A and B under the low cost assumption and
at point A' under the high assumption, while too
lenient standards are reflected in point D under the
low assumption and C' and D* under the high as-
sumption. The least cost standard, in terms of true
social costs, is at point C under the low health
cost assumption and at point B' under the high
assumption. Least cost standards protective of health
require adequate quantitative exposure-response in-
formation, knowledge of control costs and range
estimates of health cost of pollution effects.
Even range estimates of pollutant related health
costs are difficult to derive. Many direct and indirect
health costs must be considered, including: Im-
mediate and delayed health effects of short-term and
long-term exposure; the contribution of pollutant
exposure to the occurrence and severity of major
public health problems such as acute and chronic
respiratory diseases, heart disease, cancer and con-
genital defects; and adverse health effects of control
strategies. While these estimates are formidable
tasks, failure to make the estimates will prevent even
the possibility of achieving least social cost stan-
dards, except entirely fortuitously. Required annual
reports of air pollution control costs clearly warrant
better and more quantitative health effects data and
O
QC
»-
O
u
LL
O
i
COST OF
CONTROL
HEALTH COST
HIGH
ASSUMPTION
LOW
ASSUMPTION
B
UJ
oc
1
X
UJ
X
I-
o
oc
POLLUTION CONCENTRATION
LEAST COST STANDARD: C OR B'
EXCESSIVE LY STRINGENT STANDARD: A, B, A'
EXCESSIVELY LENIENT STANDARD: D, C', D1
FIGURE 4. LEAST COST PROTECTIVE STANDARDS
79
-------
concerted efforts to convert health effects to health
costs.
Recently the National Academy of Sciences was
asked to review the health basis for the ambient air
quality standards. One part of that review was a
Conference on Health Effects of Air Pollutants that
was held on October 3-5, 1973.10 All existing
U.S.A. ambient air quality data were presented as
least case, worst case and best judgment values.
BIBLIOGRAPHY
1. WHO: Air Quality Criteria and Guide for Urban Air
Pollutants. Wld. Hlth. Org. Techn. Rep. Ser. 1972, No.
506.
2. Newill, V. A.: PSAC Briefing on the EPA Air Pollu-
tion Health Effects Research Program, Washington,
D.C. May 22, 1972. This briefing was based on many
in-house reports from the Division of Health Effects
Research, NERC-RTP, EPA, North Carolina.
3. Message from the President of the United States Trans-
mitting the Report of the Department of Health, Educa-
tion and Welfare and the Environmental Protection
Agency on the Health Effects of Environmental Pollu-
tion, Pursuant to Title V of Public Law 91-515. 92nd
Congress, House of Representatives Document No.
92-24 I.February 1, 1972.
4. MacMahon, B.; Pugh, T. F.; and Ipsen, J.: Epidemi-
ologic Methods. Little, Brown and Company, Boston.
Massachusetts, 1960, pp. 18-21.
5. Hill, A. B.: The environment and disease: Association
or causation. Proc. Royal Soc. Med. 58:295-300, 1965.
6. Shy, C. M.: Health intelligence for air quality stand-
ards. Presented at the Meeting of the President's Air
Quality Advisory Board, St. Louis, Missouri, March 27,
1973.
7. Newill, V. A.: Nature and source of pollutants. Pre-
sented at the American Academy of Pediatrics Con-
ference, Evanston, Illinois, June 11, 1973.
8. Weinhouse, S. and Rail, D.: Personal communication.
9. Bates, D. V. and Hazucha, M.: The short-term effects
of ozone on the human lung. Proceedings of the Con-
ference on Health Effects of Air Pollutants, NAS,
NRC, October 3-5, 1973, pp. 513-540. U.S. Congress
Document Serial No. 93-15, November 1973.
10. Finklea, J.: Conceptual basis for establishing stand-
ards. Proceedings of the Conference on Health Effects
of Air Pollutants, NAS, NRC, October 3-5, 1973, pp.
619-668. U.S. Congress Document Serial No. 93-15,
November 1973.
80
-------
Maximum Permissible Human Stress Pollutants with Multiple
Pathways: Ionizing Radiation, Asbestos and Lead
John H. Knelson *
I. INTRODUCTION
All human activity is associated with a degree of
stress; that is with a degree of risk to the health
and welfare of the human organism. Only recently,
however, has man begun to contemplate seriously
the degree of stress and the risks associated with
rapidly evolving technology. Initially, these concerns
were focused directly at the most obvious site of risk,
the workplace itself. Now, concepts from industrial
hygiene and medicine are serving as important bases
for development of the more general field of en-
vironmental hygiene. As technology becomes more
complex and pervasive, the problems of environ-
mental hygiene also become more complex and more
pervasive; obviously then, regulatory decisions based
on considerations of environmental health have in-
creasingly significant social consequences.
Assessment of the influence of environmental fac-
tors on human health is one of the most important
components of comprehensive environmental analy-
sis. The measurement of human health, however, is
in itself a complex and incompletely understood
task, and it becomes increasingly complex when one
considers the multiplicity of environmental stressors
which may act upon the body. It is essential, there-
fore, that some simplifying concepts be employed to
organize and interpret environmental health effects
data. How do we define "Maximum Permissible Hu-
man Stress"? How do we determine environmental
standards for pollutants that permeate the environ-
ment, enter the human body, and exert influence
through multiple mechanisms? These are parts of
the bigger question, "How should we modify or re-
direct our technological development to assure that
environmental hazards are maintained at a level
that represents the best balance between acceptable
risk and acceptable cost?"
II. DOSE-RESPONSE RELATIONSHIPS
The determination of an association between an
environmental agent and a putative health effect re-
quires that the exposure concentration in the environ-
ment be measured accurately and unambiguously.
This rather self-evident prerequisite to health effects
studies is frequently not met. Although conceptually
simple, it is in practice often very difficult to meas-
ure with acceptable accuracy the pollutant dose to
which a population is exposed. This problem is com-
pounded when the agent exists in several forms in
the environment and enters the human organism
through a variety of pathways and from a number
of sources. Even when good chemical analysis is
possible, determination of the actual dose delivered
to the individual is difficult. Environmental monitor-
ing at a particular point gives a relatively poor esti-
mate of the spatial variability of pollutant concen-
trations within an area—and the exposure over time
of an individual residing in that area is even more
indeterminate.
Some materials, such as lead, accumulate in
human tissue. That accumulation, a function of
dosage integrated over some period of time, can be
expressed as "pollutant body burden." This total
body burden, the total quantity of a substance dis-
tributed within body tissues or physiologic compart-
ments, cannot be partitioned into contributions de-
rived from each of the multiple portals of entry, nor
does body burden correlate consistently with re-
sultant health effects.
In the case of asbestos, monitoring of exposure
becomes excruciatingly difficult for not only is par-
ticle composition crucial, but size and aerodynamic
characteristics are also important. Although "fer-
ruginous bodies" are found in the sputum of asbestos
workers, no way has been developed to associate
these bodies quantitatively with the intensity of en-
vironmental exposure. With ionizing radiation, con-
venient personal monitors—film badges—are avail-
able. But they do not measure that fraction of total
body dose inhaled or ingested as radionuclides. In
general, then, measurement of the dose of any en-
vironmental pollutant an individual receives is still
technically difficult.
•Acting Director, Human Studies Laboratory, National Environmental Protection Agency, U.S. Environmental Protection Agency.
81
-------
Measurement of health effects attributable to en-
vironmental pollution is also fraught with many
methodologic difficulties. Definition of health and
description of statistically significant deviation from
the norm is greatly complicated by the multitude of
factors that determine the health status of any pop-
ulation. These difficulties in the measurement of
health effects pervert the meaning of threshold.
Though paying homage to the true definition, in the
real world we set a threshold level at the point where
available health effects data fail to show any sig-
nificant dose-response relationships. Historically, the
more data accumulated the lower the threshold level
is set. In this context, the threshold, when not just
a statistical artifact, is purely an operational concept
which contains more information concerning the
level at which our uncertainty begins than about the
underlying reality.
As the tools for measuring dose as well as those
for measuring human response are improved, it be-
comes possible to describe a continuum of effect
down toward zero dose. Thus, even a very small en-
vironmental insult probably has a very small but
measurable impact on human health. It is then no
longer a question of "What is a safe level?" Rather,
the question becomes "What level of risk is accept-
able?" The concept of threshold limit value is re-
placed with the much more complicated concepts of
Cost-Benefit analysis, which allow us to define Ac-
ceptable Risks. It is no longer sufficient for the scien-
tist to provide a threshold value and suggest a margin
of safety. He must provide the decisionmakers with
the best possible objective description of health ef-
fects as related to dose, and also with an estimate of
his degree of uncertainty concerning that relation-
ship.
HI. ENVIRONMENTAL COSTRESSORS
Implicit in the description of Dose-Response re-
lationships is the ranking by relative priority of the
described relationships. A comprehensive control
strategy requires knowledge not only of the well
described health effects of a particular environmental
agent, but knowledge of the importance of that agent
relative to the importance of other risk factors, not
all environmental, which may lead to the same
health effects. The risk attributable to a particular
environmental agent must be judged in comparison
with the risks attributable to other factors, such as
with eating, drinking, smoking, physical activity, and
mode of life in general. Only in this way may we
efficiently employ society's finite resources; the cost-
benefit ratio should be minimized not just with re-
spect to all activities which come under the broad
umbrella of Public Health. It should be clear, how-
ever, that control activities will not invariably be
directed at the most important risk factors, as risk
factors are not all equally amenable to control.
Risk factors not only sum up, they may also inter-
act. It is well known that smoking increases the risk
of industrial asbestosis. Ionizing radiation in some
forms, smoking, and asbestos fiber inhalation are all
associated with various types of lung cancer. Yet, we
are almost totally ignorant of the relative contribu-
tion of these factors, much less the possible inter-
actions among the three. More recently, we have
begun to suspect that low level lead intoxication and
certain kinds of ionizing radiation may act syner-
gistically to produce some forms of mental retarda-
tion. We are only beginning to explore the relative
importance of environmental costressors.
IV. RANGE OF SUSCEPTIBILITY
Humans are not uniformly susceptible to environ-
mental stress. The historic episodes that first drew
worldwide attention to the lethal consequences of
severe air pollution affected primarily the very young,
the aged, and the infirm. Those early observations
were evident even to the untrained. Since then, more
careful documentation of range of susceptibility has
been possible.
In epidemiologic studies, the strongest correlations
with pollutant exposure have been seen in the inci-
dence of respiratory infection in children, in the in-
creased number of attacks in asthmatics, and in the
increased intensity of symptoms in people with
cardiopulmonary disease.
Ionizing radiation is known to have an especially
deleterious effect on the developing mammal, but we
have no firm evidence that any subsets of the adult
population are particularly susceptible. Excessive
lead intake, by whatever pathway, is associated with
impaired central nervous system development in
young mammals, and may be particularly harmful
in people of all ages with some types of kidney dis-
ease. Particular susceptibility to inhaled or ingested
asbestos in the absence of other environmental co-
stressors has not been described up to now.
V. SPECTRUM OF RESPONSE
Whereas there is a range of susceptibility within
a population, "spectrum of response" refers to the
various ways an individual can manifest the effects of
environmental stress. The spectrum of response is
usually a function of pollutant concentration, dura-
tion of exposure and the presence of costressors such
as exercise, thermal stress, or other coexisting pol-
lutants. We have somewhat arbitrarily classified the
spectrum into categories ranging from the mildest
to the most severe. They are: (1) pollutant body
82
-------
burden, (2) changes of uncertain significance, (3)
pathological changes, (4) morbidity, and (5) mor-
tality. Because the proportion of any population
falling into any category is greatest with (1) and
the least with (5), it is evident that the concept of
"range of susceptibility" is closely related to that of
"spectrum of response." Combining the concepts of
dose, spectrum of response, and range of suscep-
tibility allows us to construct a three-dimensional
dose-response surface from which both risk and the
proportion of the population at risk can be estimated
for each dosage level. This kind of prediction is the
final goal of environmental health effects research.
We are trying to answer the question, "How many
individuals in each category of our population are
subjected to what levels of risk when subjected to
pollutant concentrations within a given range?"
VI. THE EXPOSURE-RESPONSE
MATRIX
Comprehensive environmental analysis as a foun-
dation for comprehensive environmental control re-
quires not only a firm health effects data base, but
interpretation of the data in a manner useful to those
responsible for establishing the control strategy. The
concepts necessary for that interpretation have been
described. Assembling the proper variables into an
exposure-response matrix should provide, for.each
environmental stressor, a point of departure for
eventually assessing the Maximum Permissible Hu-
man Stress attributable to pollutants with mulitple
pathways.
The exposure-response matrix uses dose, estab-
lished by environmental monitoring and modeling or
by pollutant burden analysis, as the independent
variable. The dependent variable is the Spectrum of
Response. To these are added an intervening vari-
able, the Range of Susceptibility. Without this-third
covariate, such a plot is simply the classic dose/
response relationship. With the "z" axis representing
the Range of Susceptibility, we have additional in-
formation of use in total assessment of risk. It allows
us -to relate the degree of response at a particular
dose to specifically susceptible population subsets.
Estimates of Maximum Permissible Human Stress
can be derived from the exposure-response matrix
in combination with a sector of the relevant popula-
tion sizes for each population subset. From the set
of outcomes, .a weighted total must be generated, the
set or sector of weights expressing the importance of
the outcome. These weights are necessarily judg-
mental. However, their use with the exposure-re-
sponse matrix does at least provide a tool for apply-
ing judgmental decisions objectively and uniformly
in developing the cost/benefit information that is
needed for- establishing environmental standards.
VIL ENVIRONMENTAL AGENTS
WITH MULTIPLE PATHWAYS:
IONIZING RADIATION, ASBESTOS
AND LEAD
Three examples of environmental agents with
multiple pathways for entry into the human and
multiple mechanisms of action will be discussed
briefly to elaborate some of the principles for estab-
lishing maximum permissible human stress. These
examples have been chosen primarily because they
illustrate the problems involved in dealing with pol-
lutants of multiple pathways, not because the prob-
lems with these pollutants have been solved. It is
anticipated, however, that analyses of these prob-
lems will demonstrate some helpful first steps in ar-
riving at answers when one is dealing with compre-
hensive environmental analysis and control.
Ionizing Radiation
This example of an environmental stressor, or
category of stressors, is particularly important be-
cause it is probably the one for which we have more
health effects data than any other. Thus, some im-
portant analytic, conceptual, and control philosophy
precedents have been set. Many prototype pro-
cedures for environmental analysis and control have
been developed by a generation of scientists who
devoted their careers to studying the problems of
ionizing radiation. It is an environmental stressor of
multiple pathways par excellence for it is transported
through all media of the biosphere, entering the
human via all imaginable portals.
It is significant to note that the most recent sum-
mary of effects on populations exposed to environ-
mental ionizing radiation begins with a careful com-
prehensive analysis of nature, sources, and transport
mechanisms of the environmental stressor. Without
this analysis, estimation of health effects is inade-
quate. Too frequently, toxicologic evaluation of other
environmental stressors is attempted without suf-
ficient data concerning pathways from the various
sources to the human recipient.
Biologic effects of ionizing irradiation are classi-
fied according to influence on genetic material,
growth and development, and somatic health. Genetic
effects predominated early concern, and have be-
come increasingly important as more has been
learned of the molecular basis for disease. Increased
knowledge of genetic effects has not lead to a greater
precision of dose-response estimates, however, as
new complexities have been discovered. The prin-
ciple of dose-rate independence has lost some cred-
ibility as mechanisms for cellular repair of radiation
damage become better understood. However, dose-
rate delivery during an organism's relatively short
developmental period must still be considered in
establishing life-span standards because of the greater
83
-------
susceptibility of the developing organism to effects
of ionizing radiation. Important analogies with other
environmental stressors emerge as one considers the
problems of establishing other dose-rate standards.
Averaging times for most pollutant standards are
established with crude dose-rate effects data, if any.
Many environmental stresses have more pro-
nounced effects during developmental stages of the
organism, as does ionizing radiation, and this serves
not only to emphasize the range of susceptibility
concept, but also to raise the problem of pollutant
interactions. Other environmental agents, especially
heavy metals, drugs, and food additives are known
to be preferentially distributed to certain tissues dur-
ing the development period and these may exert a
synergistic effect with other environmental factors
which must be recognized in a comprehensive analy-
sis of maximum permissible stress.
The somatic effects of ionizing irradiation, e.g.,
cancer induction, provide, perhaps, the most striking
example that multiple pathways and interactions of
environmental stressors must be considered in estab-
lishing acceptable risk levels for each pollutant. Many
pollutants are carcinogenic. They permeate the en-
vironment, entering the host by all portals. Their
effects may well be additive or even synergistic. They
represent the most important challenge to compre-
hensive environmental analysis.
The comprehensive environmental analysis of
ionizing radiation has been more extensive than for
any other environmental stress. Yet, even with the
extensive data available, establishing the exposure/
response matrix as described is very difficult. Esti-
mates of acceptable risk are possible only within a
range of one order of magnitude. Although genetic,
developmental, and somatic effects have been ex-
plored, the genetic effects of radiation are most im-
portant in estimating acceptable risk.
Four techniques have been used to calculate
genetic risk:
(1) Risk relative to the effects of natural back-
ground radiation;
(2) Risk for specific genetic conditions;
(3) Risk for severe malformation and disease;
(4) Risk in terms of overall ill health.
Natural background radiation averages about 10
mrem per year. However, there is considerable vari-
ation in this figure depending on geographic loca-
tion, type of dwelling, and altitude. Varying concen-
trations of naturally occurring minerals that emit
radiation account for most of the geographic dif-
ferences. Structures built of stone containing these
minerals will subject the inhabitants to more back-
ground radiation than wooden buildings. Screening
of cosmic (mostly solar) radiation by the atmosphere
decreases with altitude. Thus, man throughout his
existence has been exposed to radiation, and the
effect of that natural or background level serves as
a convenient denominator for expressing relative
risk. If man accepts a range of risk, depending on
the factors described, from background radiation
approaching zero to that over 200 mrem per year,
it can be argued that he should accept additional
unspecified risk the same order of magnitude as-
sociated with environmental stressors or man's cre-
ation. "If the genetically equivalent significant ex-
posture is kept well below this [background] amount,
we are assured that the additional consequences will
neither differ in kind from those which we have ex-
perienced throughout human history nor exceed
them in quantity" (BEIR Report, 1972). This esti-
mate, based only on an estimate of dose and none
of health effect is more philosophical than scientific,
but the advantage of avoiding all the assumptions
of human radiation genetics.
Risk estimates for specific genetic abnormalities
can be made by comparing the current incidence of
observed genetic conditions to those expected from
increased radiation. There are no directly applicable
human data, so those from experiments with mice
must be used. Chronic radiation to developing mouse
sperm cells results in about 0.5 x 10~7 recessive
mutations per rem per gene. Developing oocytes are
much more resistant. Therefore, the average for both
sexes is less than 10~7, and must because of the un-
certainties in the data be estimated as somewhere
between 10-8 and 10~7 per rem per gene.
Spontaneous mutation rates in the human are
estimated to be less than 10-". This is 10 to 100
times the rate induced per rem. Thus, the ratio of
radiation-induced to naturally occurring mutations
is 1/10 to 1/100, and the additional dosage neces-
sary to double the spontaneous mutation rate is 10
to 100 rem. This is strictly an "order of magnitude"
estimate. More subtle arithmetic might place the
estimate at 20 to 200 rem for a "doubling dose."
Another statistical approach to the concept of
"doubling dose" utilizes data from Hiroshima and
Nagasaki. The observed death rates in children of
exposed survivors does not differ from matched
population controls who were not exposed. However,
if some assumptions concerning average acute pa-
rental dose (100 rem each) and liveborn death rate
from "spontaneous" genetic abnormalities are ac-
cepted, the chances are less than 0.10 that the true
death rate was as great as one percent higher in
children of irradiated parents. Thus, 100 rem at
most resulted in an effect twice that occurring with-
out additional radiation, and the "doubling dose"
would be 50 rem. This is within the estimate of 20
to 200 rem for "doubling dose" made on the basis
of entirely different assumptions.
To estimate the impact of induced genetic im-
pairment on the overall risk of severe malformation
and disease, one must know the increase in induced
84
-------
genetic impairment and the contribution of such im-
pairment to disease in general. Using the estimate
of 20-200 rem as the "doubling dose," the increase
in relative mutation risk is in the range of 0.005-
0.05 per rem. The continuous exposure rate of 5
rem per 30 years (or 170 mrem per year) would, in
equilibrium, cause an increase of 2.5-25 percent
in the overall incidence of genetically-related disease.
But what fraction of malformations and other rec-
ognizable disease states are mutation-related? Esti-
mates of the total burden of human disease related
to hereditary factors are virtually meaningless.
Therefore, even the best estimates of radiation-
related genetic abnormalities (or spectrum of re-
sponse) mean little until better data concerning the
range of susceptibility within the population are
available.
Risks in terms of overall ill health are included
because the other estimates deal with more specific
disease states that could result in an under-estimate
of radiation risk to health in general. Again, only
order of magnitude relative risk estimates are pos-
sible with available data. The most reasonable esti-
mate is that 20 percent of all ill health may have
some genetic component. If 20 rem is the mutation
rate doubling dose, an exposure of 5 rem per gen-
eration (5 rem per 30 years or 170 mrem per year)
would increase the total ill health burden over a
long period of time by 5 percent. If 200 rem is the
doubling dose, the increased burden would be 0.5
percent.
Thus, with a variety of assumptions and with a
relatively large data base, the standard for the
multiple pathway environmental stressor—ionizing
radiation—is set at 170 mrem per year. This stan-
dard has been established without the refinement
of the three-dimensional exposure-response matrix
because of the many difficulties cited. However, many
attempts have been made to provide the data neces-
sary. The inadequacies are logistic, not conceptual.
As more is learned of ionizing radiation effects, the
matrix can eventually be completed to provide a
more accurate estimate of maximum permissible
human stress.
Asbestos
Although associated historically only with occu-
pational exposure, over the past ten years consider-
able evidence has accumulated which associated as-
bestos disease with nonoccupational exposure as
well. This evidence also suggests a spectrum of re-
sponse ranging from those observed in persons ex-
posed in their occupations to those observed in per-
sons only indirectly exposed by family contact or by
neighborhood location.
Factors which complicate studies of the effects of
asbestos air pollution include:
(1) The long lapse period between exposure
and illness (often 20, 30 or 40 years). The long
periods between exposure to asbestos and develop-
ment of mesotheliomas that have been observed
could be indicative of a breakdown of human
defenses in elderly people. This suggests a range
of susceptibility, i.e., individuals past the age of
asbestos workers who live in the vicinity of emittor
industries may be affected by acute or chronic
exposure to relatively low levels of asbestos.
(2) Lack of knowledge concerning the signifi-
cance of intensity of exposure. Although there is
evidence suggtsting a continuous spectrum of re-
sponse for lung disease, there are no data for
establishing dose-response relationships for the
development of cancer. This dilemma emphasizes
the difficulty inherent in establishing standards for
a substance of multiple pathways and different
pathophysiologic mechanisms.
(3) Size and variety of asbestos fibers. Whereas
the occupational exposure limit is based on counts
of asbestos fibers > 5^i in length, this counting
technique has little usefulness in ambient sit-
uations. Fibers emitted into the ambient air break
up into fibrils the detection of which requires the
utilization of electron microscopy. Consequently,
equivalent quantities of asbestos pollution cannot
be detected in occupational and ambient situations
by determinations of numbers of particles > 5n
in length. When asbestos is implanted into the
pleura of rats, cancer is produced only by fibers
greater than 5/x in length, and usually at least
40/i in length. Fibers less than 5/t in length appear
to be less carcinogenic. However, this is not in-
terpreted as indicating that shorter fibers are com-
pletely harmless. Preliminary studies of asbestos
pollution in water indicate that by far the greater
portion of the asbestos consists of small particles,
although some larger fibers do occur. Hence, the
problem of asbestos pollution in water may be
quite different from the problem of airborne as-
bestos pollution.
(4) The synergistic effects of cigarette smok-
ing or exposure to other concomitant air pol-
lutants. In industry, insulation workers who
smoked had eight times the risk of lung cancer
of smoking workers exposed to asbestos and 90
times the risk of nonsmoking workers not ex-
posed to asbestos. Comparable increased risks
may be experienced by smokers who live in the
vicinity of asbestos industries.
Since timely and accurate measurements of total
quantities of asbestos in the ambient air would be
impractical, or impossible, in most areas, protection
of the general population must rely on control of
emissions. This heavy reliance on emission control
is similar to the strategy applied for ionizing radi-
ation. Considering the complete inability to quanti-
tate the effects related to specific exposure, plus
85
-------
the repeated association of fatal disease with non-
occupational exposure, it would appear desirable
that emissions be restricted to the limits compatible
with the best available control techniques. Such con-
trol recommendations should not be construed to be
recommending elimination of all asbestos from the
ambient air. Nonoccupational illness has been as-
sociated almost exclusively with indirect occupational
exposure or residence in the immediate vicinity of
an asbestos source. There is no evidence that the
small number of fibers to which the general popula-
tion is exposed affects health or longevity; this is
similar to the situation with "background" ionizing
radiation.
With available data, how much of the exposure-
response matrix can be filled in for asbestos? The
first problem is environmental monitoring. Asbestos,
a hydrated mineral silicate, occurs in several forms.
Most widely used in United States' industry is
chrysotile, a magnesium silicate. Other forms include
iron, sodium, and calcium as well as magnesium
silicate. Virtualy nothing is known of the relative
toxicity of these forms of asbestos from data per-
taining to human exposure, and very little com-
parative animal toxicology is available. Because of
this, monitoring virtually never includes specific
analysis of composition. Instead, the airborne as-
bestos filaments are collected on impactor plates or
filters, and the particles sized and counted by micro-
scopy. The relative toxicity of longer fibers (greater
than 5 micrometers) and motes (nonfibrous par-
ticles) is a question of continuing concern. Although
some studies do show a deleterious effect of motes,
their toxicity is apparently much lower than that of
the longer fibers. Whereas pulmonary pneumoconiosis
and silicosis due to paniculate silicate inhalation
have much in common with pulmonary asbestosis,
it seems only the fibrous form of asbestos is as-
sociated with gastrointestinal carcinoma and meso-
thelioma. With respect to airborne pathways, oper-
ational monitoring systems as well as particle or fila-
ment sizing and counting techniques provide less
than adequate data. These may be improved as auto-
mated optical devices are developed and qualitative
determinations using such techniques as microprobe
analysis are employed. With respect to ingestion,
especially from drinking water that has been trans-
ported through asbestos pipes, the monitoring prob-
lem is even more complicated. Again, the source of
asbestos fibers associated with gastrointestinal carci-
noma and mesothelioma is ambiguous. It is difficult
to know how much of the ingested dose is first in-
haled and then swallowed after tracheobronchial
mucocilliary transport, and how much may come
directly from contaminated drinking water. The only
way this question can be answered for exposed hu-
man populations is with much more sophisticated
and comprehensive environmental monitoring. One
can appreciate, therefore, the difficulties in char-
acterizing environmental exposure to asbestos for
the purpose of establishing the exposure-response
matrix.
The spectrum of response for asbestos exposure
has been mentioned. In all cases, however, the health
effect is disabling and eventually fatal for some indi-
viduals. Therefore, all effects lie at the extreme end
of the response spectrum, and large weighting factors
are applicable.
Range of susceptibility estimates are made diffi-
cult by the inadequate assessment of population at
risk and poor understanding of the importance of
environmental costressors. Tobacco smoking seems
the most obviously implicated costressor. But ab-
sorption of carcinogenic hydrocarbons with subse-
quent desorption after inhalation has also been sug-
gested as an additional cofactor. Trace element con-
tamination of the mineral matrix may also play a
role in the overall toxicologic assessment of asbestos.
In general, the assumption has been made that ef-
fects of asbestos exposure are dose-rate independent
and the concept of "fiber years per cubic centimeter
of air" has been employed. From estimates of pop-
ulation risk, the probability of early signs of lung
disease is less than one percent for 100 fibers years
per cc. This would be equivalent to four fibers per
cc for 25 years or two fibers per cc for 50 years ex-
posure. There are virtually no data to estimate size
of population at risk, or range of susceptibility, for
asbestos-related gastrointestinal carcinoma and meso-
thelioma. Thus, as with ionizing radiation, standards
must be set as low as possible because of lack of
health effects data and necessity for maintaining a
reasonable and prudent safety margin. A great deal
more information is necessary before the exposure-
response matrix for all pathways and effects of as-
bestos can be substantially filled out.
Lead
Environmental standards for lead which have been
proposed, take into account more information on
spectrum of response and range of susceptibility
than is available for most other pollutants of multi-
ple pathways. There is little controversy over the
basic toxicology of lead compounds. Toxicity of
lead has been recognized since antiquity. It has most
usually been associated with oral ingestion, how-
ever, and only more recently has airborne lead as-
sumed greater importance. The concern with lead
as a pollutant of multiple pathways has emerged
since the introduction of organic lead compounds to
automotive vehicle fuel because of their antiknock
properties. There has been some controversy over
the relative importance of inhaled versus ingested
lead.
As usual, environmental monitoring has neces-
sarily preceded the design of good health effects
86
-------
studies. There is now little argument that the air of
urbanized regions has higher lead concentrations
than that of rural areas. Tentative results indicate
an amelioration of air quality in urban areas where
significant decreases in automotive lead emissions
due to reduction of lead in gasoline have occurred.
Finally, the contribution of airborne lead to ingested
lead via dustfall merits comment. As with ionizing
radiation and asbestos, lead represents a complex
environmental pollutant of several related pathways.
The spectrum of response for lead, unlike that for
ionizing radiation or asbestos, has been studied from
(1) pollutant body burden through (5) mortality.
Pollutant body burden is usually expressed as blood
lead concentration. In addition, however, consider-
able animal data are available to describe lead dis-
tribution in various body tissues and compartments.
The validity of extrapolation of these data to the
human is still not completely certain, but some cor-
roborating information from analysis of autopsy tis-
sues is becoming available. A change of uncertain
physiologic significance most clearly related to in-
creased lead body burden is found in,decreased
activity of red blood cell aminolevulinic acid de-
hydrase. This enzyme is associated with hemoglobin
synthesis, but is also present in other tissues such as
the central nervous system where its function is not
understood. Changes in activity of this enzyme cor-
relate well with changes in blood lead concentrations,
but this is of uncertain physiologic significance be-
cause lowered enzyme levels are not unambiguously
related to any specific diesase state. Changes in renal
function associated with higher exposures represent
a specific precursor of morbidity. These changse are
usually associated with industrial exposure or with
lead ingestion. Animal studies, however, suggest that
very low exposure to lead results in observable
changes in renal microstructure, and that there is a
dose-response relationship continuing down toward
zero with no demonstrable threshold. Moreover,
there is evidence suggesting that repeated low-level
exposure may alter pathways of detoxification and
excretion without any signs of toxicity. Finally, of
course, the overt morbidity and mortality of central
nervous system and renal disease related to lead
intoxication is well-known. Less well appreciated or
understood at this time is the possible subtle inter-
action of lead with many other environmental fac-
tors, including ionizing radiation, to produce mental
retardation in children. One sees for lead, therefore,
a wide spectrum of response irrespective of the path-
way into the human.
Whereas there has been fairly good understanding
of the relationship between dietary lead and result-
ing blood lead or pollutant burden level, only re-
cently have reasonably good data for respiratory
lead uptake become available. Studies exposing men
to carefully controlled low air lead provided data for
a rather well-defined respiratory lead uptake model.
Population studies have resulted in comparable esti-
mates of the relationship of air lead to total body
pollutant burden.
Females residing in urban area and exposed to
higher air lead levels were found to have significantly
higher blood lead levels than those living in suburban
areas where air lead levels were lower. In Phila-
delphia, 11 percent of women living in the urban
area had blood lead levels of 29 /ig/100 gm or
greater. The average urban air lead concentration
there was 1.67 /*g/m3 during the period of the study.
Similar findings in Chicago were: 3.4 percent of the
urban women studied had blood lead levels of 29
n%/100 gm or greater with an average air lead of
1.76 /ig/m3. In New York, however, only 1.4 per-
cent equaled or exceeded the 29/ig/100 gm level at
mean air concentration of 2.08 /*g/m3. Extrapolating
these percentages to the entire population base of
these urban areas would, of course, involve a large
number of people with blood levels that can be con-
sidered excessive.
Another recent study has found significantly in-
creased blood lead levels in women living near a
well-traveled roadway compared to those living fur-
ther away. Air lead exposures were particularly well-
quantitated in this study because measurements were
made on the front porches and inside the homes
where the women lived. The average air lead just
outside the homes nearest the highway was 4.6
Mg/m3 and that inside the home was 2.3 i»g/m3.
The air lead levels just outside and inside the homes
122 meters (400 feet) from the highway were 2.24
Mg/m3 and 1.57 Atg/m3, respectively.
Epidemiologic evidence indicates that while mean
blood lead levels in populations are not significantly
elevated by exposure to aid lead of about 2 ng/m3,
the frequency distribution of these blood lead values
do differ so that significant urban-suburban differ-
ences do exist, and large numbers of people do have
blood levels equaling pr exceeding 30 ^g/100 ml
level at average air lead exposure levels of 2/ig/m3
and less.
With respect to range of susceptibility, in the case
of lead it is becoming evident that children, the fetus
and newborn in particular, constitute a specially
susceptible group. This observation is certainly re-
lated to the increased incidence of overt lead poison-
ing in children due to their eating lead-based paint.
Not equally appreciated, however, is the importance
of their eating lead-containing dust and dirt from
atmospheric fall-out. Recent evidence indicates that
children may absorb considerably more of their oral
lead intake than do adults. Although opinion is far
from unanimous on the mechanisms of lead absorp-
tion from the respiratory tract, a significant fraction
of that inhaled and deposited there may eventually
87
-------
find its way to the gastrointestinal tract via mucocil-
liary tracheobronchial clearance. Greater enteric ab-
sorption of lead by children is thus important from
several aspects. Finally, the respiratory dose per unit
body weight may well be greater for children than
adults because their level of ventilation is greater at
rest and their level of activity is higher. There is no
evidence that their detoxification and excretory
mechanisms for lead are proportionally accelerated.
Neither is there any to suggest it is not. Taking into
account reported overt clinical toxicity in children
of lead at blood levels of 40—50 //.g/100 ml, the
other considerations discussed above, and the most
minimal of safety margins, an acceptable blood lead
concentration for children of 30/ug/lOO ml seems
prudent and reasonable. Because of the reported
high correlation between newborn and maternal
blood levels, expectant mothers should also be pro-
tected from achieving blood lead levels of 30 /ug/100
ml or higher.
It can be concluded, therefore, that a prudent and
reasonable upper limit for blood lead concentration,
one with a certain margin of safety, that can be used
as a criterion for setting an air quality standard is
30 /ig/100 ml. This level applies to a large fraction
of the population; that is, newborns, children and
expectant mothers. Adults may tolerate, with a
margin of safety, levels somewhat higher than this.
Because it is not feasible to implement different
ambient air quality standards for children and ex-
pectant mothers as opposed to those for the rest of
the population, the lower value must be chosen.
The exposure-response matrix for lead can be
completed with a degree of confidence not yet pos-
sible for most pollutants. Imperfect as it is, blood
lead concentration nevertheless serves as an im-
portant index of integrated lead exposure from dif-
ferent pathways over relatively long periods of time.
Comprehensive environmental standards for air,
food and water can be established to maintain the
quality of the environment above that related to ac-
ceptable risk.
VIII. CONCLUSIONS
Assessing the influence of environmental agents on
public health is a difficult multifactorial problem.
Composition and concentration, sources, distribu-
tion, and bioavailability of pollutants must be de-
termined before meaningful estimates of maximum
permissible human stress or acceptable risk are pos-
sible. As comprehensive environmental analysis gives
improved estimates of existing dose and dose-rate
factors, biological effects data become more relevant.
Spectrum of response and range of susceptibility
provide the tools for establishing an exposure-re-
sponse matrix from which some objective cost/
benefit analyses can be drawn. The problem of as-
sessing combined effects due to pollutants with
multiple pathways has barely begun. It will be best
pursued using firm principles and precedents estab-
lished in evaluating single pollutant effects.
88
-------
Complex Analysis of the Environment, Approaches to a
Determination of the Permissible Loads on the Natural
Environment and to a Justification for Monitoring
Yu. A. Izrael *
Two interrelated questions arise, given the ex-
tremely varied anthropogenic effect on the natural
environment:
(1) What is the state of the biosphere today (on
a global, ecosystem, regional, or local scale); is this
a crisis state?
(2) What is the possible, permissible load of the
different effects, taken in the existing complex, on
the environment on different scales?
Only the sequential development of a system of
observations, of predicting the state of the environ-
ment, and of estimating the existing and predicted
state of the environment, can answer the first ques-
tion.
This report, and the symposium as a whole, are
devoted to an attempt to answer the second question.
The result of the answer to the second question is
primary to estimating the state of the environment,
and is, therefore, the scientific basis of monitoring;
it is precisely this result that should be the basis for
the optimization of the relationship between man
and the environment, and the basis for monitoring
and controlling the quality of the environment.
Most of the work in determining permissible loads
up to this point has been limited primarily to the
sanitary and hygienic approach (in developing
permissible loads on man), and to the economic
approach (the fisheries approach, for example).
Secondary norms and standards too have been de-
veloped in which the possible consequences of effects
on vegetation have been taken into consideration.
However, it is apparent that generalized ap-
jHoaches to the development of permissible effects
on the population, of the assemblage of the system
as a whole must be considered; individual sources
must be taken into consideration, as must sources
in this region, or system, as a whole, and the com-
plexity of their effects, possible interactions. More-
over, it is desirable to find criteria to use to select,
or define, the different variants of permissible effects
(loads).
We shall, in making a complex analysis of all
effective factors, estimate the final result (from the
point of view of its acceptability) in terms of the
magnitude of the damage. The latter can be broken
down into ecological, economic, and moral (es-
thetic). Recreational damage can be included in the
latter two kinds of damage. The kinds of damage
indicated are interrelated to a considerable degree.
Approaches to determining the extent of the damage
will be described in what follows. It is desirable to
adopt in advance certain strict limitations (for ex-
ample, a total ban on any ecological or esthetic dam-
age to unique objects, or, in general, any changes
despite the fact that damage from a change at the
particular point in time is not in evidence), as well
as the unacceptability or unjustified (irreversible)
ecological changes in most natural objects, and so
forth.
Table 1 explains the kinds of damage that can
be done, and establishes certain links between them.
Obviously, there is little chance of estimating es-
thetic damage at this time. There are certain ap-
proaches associated with an estimate of the time and
money man spends to satisfy his need for esthetic
values that can be taken to make such an estimate.
Such expenditures are growing steadily, and will
continue to grow in the future. Improvement, or
preservation, of man's health will, at the same time,
yield to evaluation.
The moral (esthetic) damage is determined to a
considerable degree by an upset in man's habits, in
his perceptions, emotions, and in the concepts he has
formed during his upbringing, and as result of the
education he receives during his lifetime.
We shall try to establish a link between the magni-
ture of the source (emission, emissions) and the
tude of the source (emission, emissions) and the
scope of the ecological damage (factual A*) within
the framework of some limited area.
•First Deputy Chief of Main Administration of Hydrometeorological Service, Moscow
89
-------
Table 1
Ecological damage
A
Economic damage
B
Moral (esthetic) damage
I. Unjustified—irreversible con-
sequences
Direct losses in the
losses in the future
present;
II. Reversible consequences
a. deterioration in the con-
dition of individuals, pop-
ulations, ecosystem
b. response without deteri-
oration in condition
—animals, plants
—man
Direct or potential losses in in-
crease in biomass (harvest)
Possible migration with possible
economic and esthetic losses
Disruption of usual perceptions
Recreational damage with possible economic and esthetic losses
For this we must know:
(1) the function of the sources (in space and
time):
R, t*>= y y y
(1)
where
j 1
is the different factors, ingredients, of
ith orders within a homogeneous class;
is the nature of the medium (air, soil,
fresh, sea, or subsoil water, and so
forth);
is the nature of the heterogeneity (pol-
lutant, radiation, heat);
(2) the geophysical laws of distribution (function
F) in different media. The concentration (or value
of some factor) ^jn in the different media is
J
1
= F[Q(R, t*), v, KT, wz, t-t*] (2)
where
is the rate of horizontal movement of
an impurity (wind speed, flow, and so
forth);
is a coefficient characterizing turbulent
diffusion;
w is a coefficient characterizing the verti-
cal rate of transfer from one medium
to another (sedimentation, precipita-
tion, and so forth);
t-t* is the elapsed time between emission
and time;
The possible transfer of part of a substance from one
medium to another during distribution (migration)
is
Jjli
(3)
(3) the chemical (physical) laws tor the conver-
sion of a given ingredient (factor) into another
jli
(4)
where
is the concentration (or value of the
factor) of the ingredient at which the orig-
inal ingredient, or factor, can be converted
during distribution, or accumulation;
is is a coefficient characterizing the conver-
sion;
(4) the distribution of organisms in space
N»(R,t);
m is their level; it is desirable to know Mm, the
distribution of their weight relationship from' the
point of view of their ecological value, importance,
determining their role in system stability. Nm(R,t)
takes into consideration the distribution of individual
parts of the organism of a given specimen (popula-
tion) in space (sometimes this is very important, as
would be the case, for example, for plants, the roots
of which are in the ground, while the above-ground
organs are in the air).
The magnitude
(R, t)dRdt =
m
(5)
is an integral magnitude characterizing an m-popula-
tion as a whole.
The importance of the factor Nm(R,t) must be
stressed, because the factual distribution or organ-
isms in time and space determines to a considerable
degree the ecological damage of an effect on the en-
vironment (because this distribution characterizes
the possibility of contact between the organism and
pollutants, or other factors).
For example, in estimating the factual situation in
cities (determination of damage to the population)
knowledge of the distribution of pollutants is not
enough. This distribution must be compared with the
90
-------
distribution of the organisms in time and space in the
particular city (including according to height);
(5) t is the measure of the biological effect
(damage) for the given organism (level m) for the
effect of a factor (ingredient) (1), entering the or-
ganism (or making contact) from the j medium. At
the same time, it is very important to take into con-
sideration the entrance of the particular ingredient
fram all j media. The ingredient can affect different
organs, so can differ in the effect produced as well.
The magnitude t, generally speaking, depends on
the magnitude a, on the rate of increase dcr/dt, and
on its peak values, as well as on the duration of the
effect.
In a more general form
m
(6)
where
Sj(t) is the percentage of the substance enter-
ing the organism in unit time (or of
radiation irradiating the organism)
from the medium including the geo-
metric factor);
rm is the percentage of deposition, of im-
mobilization in critical organs;
xm is the effect of the biological effect on
any organ (on the organism as a whole)
from the unit of immobilized substance
(transmitted energy).
It is obvious that <7j8j(t)T is the quantity of the
substance in a critical organ (organs), or the quantity
of radiation interacting with the organ as a function
of time, and entering from medium j.
It should be pointed out that the magnitude < de-
pends on the condition of the organism. The latter
is determined by a great many external effects, and
undoubtedly by previous history (by the unhealthy,
or weakened, condition of the organism) as well.
This condition should be determined directly, or in-
directly (by measuring the factors that influence the
state of the physical and physicogeographical data,
living conditions, and state of health), and should be
introduced into the formula (the variable *) used
to characterize the condition of the organism, or
population, or more correctly its deviation from the
norm, attributable to the influence of external condi-
tions, and including some distribution of the sus-
ceptibility of individual organisms in the given popu-
lation.
Under normal, ordinary, conditions, the magni-
tude *m can be taken as equal to unity (on the as-
sumption of equal sensitivity of all organisms in the
given population);
(6) M is the effect of the interaction (the mutual
effect factor (of two factors, i, k (ingredients), acting
on the given organism;
The simultaneous effect must be estimated when
several factors are acting (generally speaking, for all
factors, including um) .
The < dependence is nonlinear, as a rule. The
function can have dimensions (especially in the case
of large doses, in the case of the destruction of the
organism, of the onset in the organism of irreversible
changes), possibly even in the case of very small
doses (when the effect has a clearly expressed thresh-
old). However, we are convinced that the effects of
the action have no threshold (accumulation of data
for small doses). This is important for summing
effects from different ingredients;
(7) Xim,n is the result of the intensification (or
weakening) of the effect on the given population,
m, taking into consideration the effect of the dif-
ferent factors on several (n) populations simultane-
ously. The result of the complex effect may be writ-
ten as £m where £m is a complex function of the
values of £m and n.
Thus, the result of the summed biological effect
At is
j 1 i
£m>k)dRdt
(7)
It is desirable, in order to sum the biological effect
of heterogeneous effects, to express the magnitude
(mm in units of permissible «mjn(p), or critical
tmju(cr) values. In other words, in terms of their
relative values with respect to the permissible or
critical effects for each type of effect, or ingredient
(probably in «.P units if c^cp and ccr, if ccr^e>cp).
If the exact distribution, Nm, is unknown, the area
of distribution, S(t), can be used conditionally.
Incidentally, this factor can have a value, even if
Nm is known (it may be useful here to introduce the
probability function for the potential "encounter"
between the organism and the polluting impurity).
As already noted, what we have to know in order
to resolve the question of the permissibility of the
load on the system under examination, is the relative
value of the biological (ecological) effect, not its
absolute magnitude; that is, its relation to some
permissible, or critical value. There are several ap-
proaches in the case of a comparison such as this:
each magnitude tmjij can be compared to a pre-
viously processed value of emj«(p) or tmju(cr), which
are magnitudes for use in establishing the values of
the maximum permissible concentration (MFC) of
ingredients for each organism and factor, or of the
maximum permissible intake (MPI) (of the in-
gredient), if it is a question of the integral magnitude
(mm (t)dt in the case of a permissible load, or of
values dangerous, or fatal, to the particular organism
91
-------
(population) in the case of a comparison with the
critical magnitude, ««.
It is obvious that exceeding e« for this population
is absolutely unacceptable if we are talking about an
individual population, but this assertion is not ob-
vious for an individual population when cm(cr) is
exceeded for the system as a whole.
This is followed by summing these relative values,
-»
using Eq. (7) (and taking Nm(R) into considera-
tion).
The result is the obtaining of AP and ACT (per-
missible and critical) values, developed for a specific,
concrete, natural object, or system, that take into
consideration all the peculiarities of the particular
ecosystem (system), as well as the climatic and
concrete (at a given time) natural conditions.
It would appear to be obvious that when we speak
of the stability of biological systems (or of other
systems), we have in mind criteria for establishing
the critical effect (that can lead to upsetting stability,
and then to the conversion, or destruction, of the
system).
What we mean when we talk about developing
an MFC, or MPI, for an individual organism, or
population, is some definite threshold, or initial path-
ological (exceeding the limits of physiological norms)
action, or effect, which usually is reversible, and
which certainly does not upset the stability of the
particular system.
Consequently, the formulation of an approach
such as this for the entire system is extremely im-
portant. At this point it would appear that we can
talk about MFC value for just a single, critical link.
However, it is more important to formulate an in-
tegral approach, even if no single value reaches the
magnitude of the MFC, but the sum of the effects
exceeds some magnitude, AP, for a system as a whole,
acted on by a whole complex of effects simultane-
ously. This approach, in our opinion, should be basic
to a comprehensive analysis of the environment.
The initial approach here can be the use of the
assumption of the lack of a threshold in the accumu-
lation of data (the occurrence of certain effects) and
the determination of the magnitude fp, by summing
the relative values, 5
j i
"A" will change by bounds, depending on whether
the system is stable, or will become unstable.
This formula also is suitable for determining dam-
age when the effect is on a major system, when not
only the ecological damage is critical (but perhaps
not the major aspect, but so too is the possibility
of upsetting the stability (or approaching this critical
state) of a major physical (geophysical) system,
when estimating damage that can result from pos-
sible climatic changes, for example.
Determination of the relative ecological damage,
Ar, undoubtedly must be made successively, with
possible approximation at first (selection of a limited
number of factors for the effect of polluting in-
gredients, and assumptions, where possible). v=l,
*-l.
Systems analysis approaches are necessary when
solving similar problems.
The Scientific Justification for Monitoring
In order to estimate the effect on a population,
community, or ecosystem by the scheme cited, we
must have the values of all the variables contained
in this formula. There is no question of the fact that
the solution to the problem cannot be obtained in
its entirety so we must proceed to simplify things by
selecting the most important factors of the effect, and
by searching for the links that are critical, or that
govern the state of the systems.
Thus, regular observation must be made (variables
Q, a, N, «, and others of the state of the medium, its
future state must be predicted, we must be able to
single out changes caused by anthropogenic action,
and we must be able to properly estimate this action,
by using the formulas cited above, for example. The
values of v, \, and of some of the others, are found
through research, and by setting up special experi-
ments.
Table 2 shows the scheme for observation, for the
production, and for the estimate, of the state of the
environment (1). This system is, in essence, the
scientific justification for monitoring. Let us explain
the data in Table 2 in more detail.
Observations (I) of changes in the state of the
environment include:
A. 1.1. Observations of the state of the environ-
ment, and of changes in this state, characterized by
geophysical data, are made by means of successive
measurements of the corresponding parameters de-
termining the instantaneous state of the environment.
Such observations already are being made by a
number of geophysical services.
A. 1.2. Physicogeographic data, including data on
the distribution of land and water, on the relief of
the land surface, on natural resources (mineral, land,
vegetable, water, fauna resources), population, ur-
banization, and others, also are most important in-
formation to have on the state of the natural environ-
ment.
A.2. Observations of the state of the environment
(and of changes in this state) characterized by geo-
chemical data, observations of matter and energy
cycles in nature, observations of the composition of
alien impurities in the biosphere (including radio-
active ones), and observations of different specific
types of pollution, including observation of noise,
thermal pollution, and different types of radiation
(the concepts of pollution in these latter cases are
conditional, but are generally accepted).
92
-------
Table 2
Observation
n
Prediction
m
Estimate of existing
and predicted states
A. State of environment
A. 1. State of the environment, characterized by
geophysical and physicogeographic data
A. 2. State of the environment, characterized by
geochemical data and data on the composi-
tion and nature of pollution
(AP in unique ob-
jects, that is, for A to exceed some permissible
value, AP, which was determined above;
(3) moral damage is not now subject to estima-
tion as a practical matter, so will not be considered
here, or in what follows.
Let us designate as E conventional systems with
a reserve (an "ecological" reserve for ecosystems)
E=Acr-Af
(8)
93
-------
where
is the factual state of the system, the eco-
logical load on the system; then
A ..A (9)
where
At, is the state of degradation of the particular
system;
Ai is the load attributable to the particular
effect.
In our opinion, C should depend on E; that is,
the damage increases with approach to the boundary
of the ecological reserve, C=E(Ao—Ar); that is, a
greater effort is required (including that of an eco-
nomic nature) to return the system to a normal state
from an unhealthy, or changed, one.
It is appropriate to introduce the concept of mag-
nitude of damage per unit of production in order to
make practical decisions
1 D,
where
(10)
D:
Di is a product of the i"1 type, manufactured
in the particular region, leading to damage
Bi; the subscript i should be replaced by
the subscript 2 when manufacturing the
summed product to estimate the damage.
This formula is useful in making a decision as to
the desirability (and feasibility) of manufacturing
the particular product as a function of the permissible
load, and of the relative ecological (or economic)
damage.
The economic effectiveness of the measures taken
to protect, or to restore, the quality of the environ-
ment may be estimated by comparing the assumed
expenditures, G, with the economic damage, B
-------
Defining "No Significant Deterioration" of Air Quality
Ralph A. Luken *
I. INTRODUCTION
The Clean Air Act of 1970 requires the Environ-
mental Protection Agency (EPA) to protect existing
levels of air quality in non-degraded areas. Taken
literally, protection means no change or deterioration
in the existing levels of air quality. Since no deterior-
ation would severely constrain economic growth in
less developed areas of the United States, the court
has interpreted protection to mean prevention of "no
significant deterioration" (NSD) of air quality.1
However, the court did not define significant in the
NSD ruling and has left the task to EPA.
Since discussion in a short paper of all the com-
plex issues associated with defining NSD is infeasible,
I have chosen to concentrate on (but not limit my-
self to) developing an economic definition of NSD.
The paper is divided into four major sections:
• Description of the Federal Law relevant to NSD
• Economic definition of NSD
• Critiqueing on the basis of the economic definition
the operational definition of NSD preferred by
• The role of rigorous Environmental Impact State-
ments in determining the significance of environ-
mental deterioration
II. RELEVANT FEDERAL LAWS
State Air Quality Implementation Plans
Under the Clean Air Act of 1970, EPA is charged
with establishing two types of national ambient air
quality standards: the national primary air quality
standard, which is that level of air quality required
to protect public health, and the national secondary
ambient air quality standard, which is that level of
air quality required to protect the public welfare
from any known or anticipated adverse effects as-
sociated with the presence of such air pollution in
the ambient air (Table 1).
EPA is also required to protect against significant
deterioration of air quality even where that air
quality exceeds the minimums prescribed in national
standards. The statutory basis for a NSD policy in
the 1970 Act is found in Section 101, which states
that the purpose of the Act is "to protect and en-
hance the quality of the Nation's air resources." To
allow deterioration of existing air quality in non-
degraded areas is considered to be antithetical to
achievement of the mandate to protect the quality of
the Nation's air resources.
Table 1
Primary and Secondary Air Quality
Standards for SO2 and TSP'
Primary
(Health)
Secondary
(Public
Welfare)
SO.
Annual . 24 hrs. 3 hrs.
80 365
1300
TSP
Annual 24 hrs. 3 hrs.
75 260
60 150
• There are also national ambient air quality standards for Carbon
Monoxide, Photochemical Oxidants, Hydrocarbons, and Nitrogen
Oxides.
The Act requires each State to submit to EPA a
plan providing for the implementation, maintenance,
and enforcement of such ambient air quality stand-
ards within each of its air quality control regions.
As a result of the Sierra Club citizen suit, the plan
must now provide for the prevention of significant
deterioration of air quality in clean areas. Each plan
must be approved by EPA before it becomes effec-
tive.
In summary, the Clean Air Act of 1970 establishes
different standards for degraded and non-degraded
areas. For degraded areas, the requirements are for
achievement of primary and secondary air quality
standards. For non-degraded areas, the requirement
is for the protection of the existing level of air quality
even though the quality level exceeds the standard.
ffl. ECONOMIC DEFINITION OF NSD
Two heuristic economic definitions of NSD offer
some insight into the problem of interpreting NSD.
A general equilibrium definition would address the
issue of whether national growth is over or under-
constrained by a NSD policy. The definition in this
•Chief, Water Programs Branch, Office of Planning and Evaluation. Environmental Protection Ai
1 Sierra Club vs. Ruckelshaus, 344 F. Supplement 523 (D.D.C. 1972) affirmed, 4 ERC 1815 (
[S!c' Circular 1972).
95
-------
case would be discussed in terms of a national linear
programming model which seeks to minimize the
costs of production subject to restraints specified for
environmental standards and demands of the do-
mestic market. The advantage of this approach is
that it recognizes that development is of national as
well as local concern and fully accounts for the inter-
dependencies among several (if separate) locations.
A partial equilibrium definition would address the
issue of whether a new source would be permitted in
a specific location. The definition in this case would
be discussed in terms of marginal damage and cost
functions associated with the residuals generated by
a plant. The advantage of this approach is that it
focuses on the value of environmental damages as
well as the cost of control. The partial equilibrium
definition is used in this paper because of the ad-
vantages noted in the previous sentence as well as the
fact that the damage and cost functions can be de-
signed to account for some national as well as local
costs and benefits.
Environmental economists have given some
thought to the "efficient" level of discharge reduc-
tion. For example, they have examined the residuals
discharged by industrial, municipal, agricultural, etc.
activities and have concluded that the discharger must
control his residual discharge up to the point where
the additional costs of reduction begin to exceed the
benefits from further curtailment of discharge.
The "efficient" level of discharge reduction for a
degraded area 3 is diagrammed in the upper half of
Figure I.4 The amount of residual discharge is re-
corded on the horizonal axis. In the absence of any
control efforts by the polluter, the maximum level of
discharge equals B. Also, the corresponding change
in air quality is recorded on the horizontal axis; the
highest level of environmental quality is 0. The cost
of controlling residuals and the dollar value of
damages associated with increasing levels of dis-
charge are recorded on the vertical axis. Total
damage cost (TDC) is shown for various levels of
discharge along the curve OC. The curve has zero
slope (essentially no marginal damages) from O to
A on the assumption that there is no damage below
the standard5; it increases at an increasing rate from
A to C on the assumption that the damages (for ex-
ample, health effects) increase with rising levels of
discharge.6 Total abatement cost (TAC) for a given
level of product output is shown for various levels of
discharge reduction along curve DB. The curve in-
creases at an increasing rate because of the increas-
ing marginal cost of controlling residuals. The cost
of control is D if all residuals are controlled. Curve
DHC is the total cost attributable to pollution and
abatement, that is, it is the sum of TDC and TAC. It
combines the cost of controlling residuals with the
damage cost associated with the remaining residuals
discharged. The efficient level of discharge reduction
for a given level of output is discharge level E,
where the sum is a minimum.
The efficient level of abatement may also be de-
termined in the lower part of Figure 1 by comparing
the marginal damage curve (MDC) and the marginal
abatement curve (MAC). The efficient level of
abatement is obtained when MDC = MAC as it is
at E. To the right of E, the plant would be required
by a society interested in efficient resource alloca-
tion to curtail its residual discharge either by regula-
tory actions or by the imposition of an effluent
charge. To the left of E, the plant would be en-
couraged by a society interested in efficient resource
allocation to discharge its residuals because the
damage to the airshed is less than the cost of dis-
charge reduction. However, the plant would be pre-
vented from discharging its residuals in the range
from G to E because the standards are based on pro-
tecting all populations against all known adverse
health effects rather than a balancing of adverse
health effects and the cost of discharge reduction.
The efficient level of abatement for a non-degraded
area 7 is diagrammed in Figure 2. The units on the
axes are the same. The shape of the TAC curve for
a projected level of output is also the same. How-
ever, the shape of the TDC curve is fundamentally
different as is the initial level of environmental qual-
ity. The TDC reflects the assumption on the part of
many environmentalists that the initial deterioration
or damage to a non-degraded area often results in a
significant loss to aesthetic amenities and, if it is an
ecologically fragile area, a significant loss in the
regenerative capacity of the natural system. The
shape of the curve suggests that the air quality stand-
ards are inadequate for protecting against all known
long-term adverse effects. Thus the TDC curve in
Figure 2 is a composite of damages because it reflects
both aesthetic and ecological effects resulting from
minute changes in ambient concentrations and the
health and welfare effects normally associated with
the TDC.
"The diagram follows A. M. Freeman, III, R. H. Haven, and A. V. Kneese, The Economics of Environmental Policy (New York: Wiley
4 A deteriorated or degraded area is defined as an airshed where the level of ambient air quality violates the national standards (Table 1).
"The TDC is drawn in Figure 1 as a hockey stick (with a kink at A) to reflect the fact that all populations in all areas must be given the same
level of protection against adverse effects.
0 The TDC as depicted in Figure 1 is a short-run damage curve for one pollutant. A long-run damage curve which took Into account cumulative
effects, or a damage curve for several pollutants would probably have a gradual slope from O to A in Figure 1.
7 A non-degraded area is defined as an airshed where the initial level of ambient air quality does not violate the national standards.
96
-------
TOTAL
ABATEMENT
COST
(TAG)
TOTAL
DISCHARGE
COST
(TDC)
G E
UNITS OF WASTE DISCHARGED
•LEVEL OF AIR QUALITY
D
t
(9
D
\
V)
DC
MARGINAL ABATEMENT COST
(MAC)
MARGINAL DAMAGE COST
(MDC)
G E
UNITS OF WASTE DISCHARGED
LEVEL OF AIR QUALITY
B
FIGURE 1. POLLUTION ABATEMENT IN A DEGRADED AREA
97
-------
G E
UNITS OF WASTE DISCHARGED
•LEVEL OF AIR QUALITY
AESTHETIC EFFECTS
HEALTH EFFECTS
MOC
B
UNITS OF WASTE DISCHARGED
_EVEL OF AIR QUALITY
FIGURE 2. POLLUTION. ABATEMENT IN A NON-DEGRADED AREA
98
-------
The efficient level of abatement is determined in
the lower part of Figure 2 by comparing the marginal
damage curve (MDC) and the marginal abatement
curve (MAC). The section of the MDC curve be-
tween I and A reflects the aesthetic and fragile eco-
logical effects, while the section between A and C
reflects the health and welfare effects. As long as the
initial level of discharge is to the left of F, the
efficient level of discharge is near zero. However, to
the right of discharge level F, the efficient level of
discharge would be in E in spite of the fact that
damages to a non-degraded area would occur in the
range from F to G and national standards would be
violated in the range from G to E.
The MDC curve might shift from MDC to MDC'
in Figure 3a if it reflected national as well as local
concern for non-degraded areas. Many urban resi-
dents in the U.S. have become increasingly interested
in maintaining the environmental quality of non-
degraded areas for their future use, for their descend-
ants or just for the pleasure of knowing a non-de-
graded area is protected from environmental deter-
ioration.8 MDC', which indicates that a higher value
is placed on environmental damage, can be seen as
reflecting both local and national concerns and
thus is a more accurate indicator of societal value.
Now as long as the initial level of discharge is at F'
or to the left of F', the efficient level of abatement is
to curtail residual discharge to near zero. However,
to the right of discharge level F' the efficient level of
discharge reduction would be E because costs of
control exceed demages in the range from F' to E.
Similarly, the MAC curve would shift if there were
opportunities foregone, such as electrical energy con-
sumption by individuals outside the region. The argu-
ment is that cost curves ought to reflect both onsite
and external costs rather than only on-site cost. If
the users of electrical energy consumption outside
the region were denied the benefits of lower cost
electricity generated in the region, then their loss
ought to be reflected in the MAC. MAC' in Figure
3b, which indicates a higher cost of residual control,
sums both on-site and external costs. In this case,
there might not be a low level of discharge for which
MDC would exceed MAC. The efficient level of
abatement would be E' rather than some low level of
discharge for a non-degraded area as suggested in
Figure 2.
The absence of certain types of information limits
the operational usefulness of an economic definition.
The most difficult problem is measuring the MDC
function or, more simply, the value people place on
clean air. Moreover, estimation of MDC is more than
assigning values. For example, estimation of a MDC
requires three sequential steps:
Step 1: construction of a transfer function which
relates changes in residuals to changes in
ambient conditions. Control of residuals
can affect the concentration of such pol-
lutants as particulate matter and sulfur
oxides;
Step 2: estimation of the impact of changes in
ambient concentrations on human health
and welfare. Linking changes in ambient
concentrations to human health and wel-
fare is the most difficult step in estimating
MDC;
Step 3: Identification of the economic value of
induced changes in human health and
welfare.
On first thought one would think about asking
people how much they value changes in clean air to
obtain information about Step 3. Since people are
likely to understate their preference for jointly-shared
social programs, several indirect techniques have
been devised to overcome this limitation. These tech-
niques measure losses in wages, residential property
values, materials, and vegetation. While these tech-
niques have provided some insight into the benefits
of pollution control (damages avoided), they usually
do not generate sufficient monetary information on
which to base a MDC function.
Measuring the cost of control is sometimes as
difficult as measuring the damages avoided. If several
plants are required to control their residuals simul-
taneously, then the change in demand for pollution
control equipment may alter the price of this equip-
ment for all plants. If a plant controls its residuals
by transferring them from water to the air, then the
control cost must also include the damage imposed
on the other medium. If there were an opportunity
cost associated with a high level of abatement, that
is, there are benefits foregone by inhabitants in other
regions because the final output is restricted, then the
control cost must reflect the value of these foregone
benefits.
Also, there are the usual problems of separating
pollution control costs from process change costs and
determining if the value of the controlled pollutants
as a secondary product will offset the cost of control.
Information about all these conditions is important
for a correct assessment of costs.
Another difficulty is estimating the initial level of
environmental quality. In many non-degraded areas
there are few if any monitoring stations for TSP and
SO;.. Vast numbers of measurements would be re-
quired to establish the baseline level of environmental
quality. Data variability is still another problem.
Normal variations in pollution concentration in clean
areas, especially TSP, exhibit considerable variability.
"There is considerable literature on the non-user values for environmental quality (Krutilla, 1967). Weisbord (1964), for e:
has made a
thly „'
99
-------
MDC
CO
a
<
V)
b
8
I MDC
FIGURE 3a
SHI FT IN MDC
F F' E
UNITS OF WASTE DISCHARGED
EVEL OF AIR QUALITY
CO
O
O
z
p
8
F E E'
UNITS OF WASTE DISCHARGED
B
•LEVEL OF AIR QUALITY
FIGURE 3. EFFECT OF SOCIETAL EVALUATION ON
MARGINAL COSTS
100
-------
For example, the 1968 TSP maximum concentration
at the Grand Canyon for participates was 126 /ig/m3,
and the annual average was 31 ^g/m3. In 1969, the
maximum concentration was 32 /tg/m3, and the an-
nual average was 12 ^g/m3. These differences are
caused by random variations due primarily to normal
meteorological factors.
Even if essential data are unavailable, framing the
issue of NSD as an economic problem is useful be-
cause it indicates the conditions under •which varying
degrees of deterioration of environmental quality are
efficient and are inefficient. Formulation of the issue
as depicted in Figure 2 suggests that permissible
changes in air quality in non-degraded areas ought to
vary with the original level of ambient quality and
the shape of the costs and damages curves. Essenti-
ally, permissible changes will fall into three cate-
gories:
• no change from current ambient concentration
would be permitted under conditions where
there is exceptionable ambient quality and
MDC is significantly greater than MAC (Zone
1 in Figure 2).9
• some change from current ambient concentra-
tions would be permitted under conditions
where there is high ambient quality, but un-
certainty about whether MDC is greater or less
than MAC (Zone 2 in Figure 2). As indicated
above, there may be considerable uncertainty
about the shape and thus intersection of the
MDC and MAC curves.
• as much change from current ambient concen-
tration as necessary to accommodate new de-
velopment under conditions where there is good
ambient quality and relative certainty that MAC
is greater than MDC (Zone 3 in Figure 2). The
permissible increment would vary depending
upon whether national standards dictated the
upper limit (point H, Figure 2) or whether an
efficient level of abatement dictates the upper
limit (point E, Figure 2).
In summary, a definition of NSD consistent with
efficient resource allocation must meet two condi-
tions. First, the definition must reflect both the costs
and damages associated with a change in residuals
discharged. A definition based on either costs or
damages alone would not result (unless by chance)
in an efficient level of residuals discharged. A defini-
tion of NSD based only on a change in ambient
quality (damages avoided) is inadequate because it
ignores the costs of achieving that change. A defini-
tion of NSD based only on a change in emission rates
from a plant is inadequate because it ignores the
change in damages associated with that change. Both
changes in ambient conditions (damages) and emis-
sion rates (costs) are central to establishing a defini-
tion of NSD consistent with economic theory.
Second, the definition must reflect the unique
characteristics of the situation. Both the original level
of ambient quality and the shapes of the costs and
damages curves will probably vary from region to
region. The permissible, i.e. non-significant, change
will vary within a non-degraded area and among
non-degraded areas.
IV. OPERATIONAL DEFINITION OF
NSD
The operational definition of NSD of air quality
preferred by EPA is partially but not entirely con-
sistent with the economic definition of NSD. The
definition provides that:
• States divide their territory into three zones;
• the Federal Government establish for each zone
deterioration changes which cannot be ex-
ceeded.
Zone Designation
A Zone I designation would be areas where little
if any change from current air quality patterns is
permissible. States would probably use Zone I desig-
nation where:
• Development of industry within the zone is in-
consistent with desirable uses for the area.
• The area should be protected from degradation
due to pollution sources outside the area.
• Development of polluting sources would signifi-
cantly degrade an area with very clear air.
A Zone II designation would be relatively unde-
veloped areas where some change is permissible. A
Zone II designation would be a recognition that the
point of transformation between Zone I and Zone
III in a non-degraded area is often difficult to dis-
tinguish because of the uncertainty about the cost
and damage functions and the absence of physical
data. This designation would prevent the arbitrary
designation of an area as either Zone I or III until
sufficient data were available to make a decision.
A Zone II designation would be areas where major
industrial growth is desired and where increases in
concentrations up to the national standards would be
insignificant. (Again, note that the national standard
as a limitation is inconsistent with the economic
definition, which sets the point where MAC=MDC
as a limitation.) States probably would use Zone III
designations where:
• an existing urbanized region has air quality
which does not violate secondary standards.
• This definition appears to be more rettrictive than the statements of the Sierra Club on the issue of NSD. The Sierra Club recognized that
NSD does not prevent all increases in pollution. "If the bed available technological developments are utilized and if numerous pollution-produc-
ing sources are not concentrated in one place, most industry can enter clean areas without causing significant deterioration." Ftderal
July 16, 1973, p. 18987.
101
-------
• counties are urbanizing.
• areas are designated for concentrated develop-
ment.
Zone Changes
The Federal Government has indicated permissible
air quality changes for Zone I and II for SO2 and
TSP (Table 2). (A permissible change for Zone III
is not necessary because it is the difference between
initial ambient quality and secondary standards.) The
permissible changes for Zone II were calculated
backwards from the impacts on the ambient environ-
ment in an open terrain of an 850 MW power plant
and a 200,000 bbls/day petroleum refinery with the
application of best practicable control technology.
(See Appendix A for more detail.) The permissible
changes for Zone I were very small numbers, arbi-
trarily chosen, which would allow little or no change
from the initial ambient quality.
Table 2
Permissible Increments for Zones I and II
Zone I
Zone II
SO,
(u8/ms)
Annual
2
15
24hrs.
5
100
3brs.
25
300
TSP
(ug/m§)
Annual 24 hrs.
5 15
10 30
Source: Harbridge House, "The Impact of Proposed Non-degradation
Regulations on Economic Growth," EPA, BOA #68-01-1561, Novem-
ber, 1973, Volumes Mil. Also, U.S. Federal Register (July 16, 1973),
pp. 18990-18993.
The permissible changes for Zone I and II are not
absolute limits on the potential changes in ambient
concentrations, in that State government can alter the
final designation of a given area if the resulting air
quality does not correspond to significant degrada-
tion.
The area classification definition is partially but
not entirely consistent with the economic definition
for the following reasons. First, it incorporates the
idea that the permissible deterioration in air quality
ought to vary within a non-degraded area and
among non-degraded areas. It allows for variability
within a non-degraded area with the three zone clas-
sification which takes into account the initial ambient
concentration and the physical, biological, and eco-
nomic development status of an area.
It allows for variability among non-degraded areas
by permitting States to modify the level of deterior-
ation given adequate consideration of costs and bene-
fits. Second, the permissible changes established by
EPA are based on the cost of control. The Federal
definition is based upon technology which has already
been installed by industrial establishment without
undue financial hardships. In addition, an economic
analysis of the impact of the Zone II requirements
suggested that industrial growth would not be unduly
restricted in upper New York State, Dallas/Fort
Worth area, and the Four Corners region.10
The definition is inconsistent with the economic
definition in that the operational definition was not
based on a comparison of the benefits and costs as-
sociated with the zonal limitations. The zonal limita-
tions incorporated only the first step in benefit analy-
sis, i.e., relating residuals discharged to changes in
ambient concentrations. It did not incorporate the
effect of the changes in ambient concentrations on
human health and welfare and the economic value of
the changes in human health and welfare. Thus, there
is no information about the physical and monetary
damages associated with the permissable changes for
Zone I and II. Whether they are of greater or less
damages than the associated costs of control is still
an unanswered question.
In summary, the proposed operational definition
is only partially consistent with the economic defini-
tion of NSD. The operational definition accounts for
the need for variation in permissible air quality
changes in non-degraded areas and is based on the
costs of controlling residuals. However, the permitted
changes in air quality for each zonal type are not
based on the damages or the absence of damages
associated with the permissible changes; this will
restrict the ability of the States to balance costs and
benefits to some extent in their zonal designation.
V. THE IMPORTANCE OF
ENVIRONMENTAL IMPACT
STATEMENTS
As a practicable matter, estimation of a national
MDC function associated with the permissable
changes is an impossible empirical task. The data are
not uniformly and consistently available. However,
Environmental Impact Statements (EIS), which are
micro-evaluations of environmental consequences of
proposed actions, are a potential source of data about
local damages. They potentially offer data for esti-
mating the local MDC associated with the residuals
generated by a new industrial source.
The EIS, which was introduced into the Federal
decision-making process by the National Environ-
mental Policy Act of 1969, must cover five points: "
Part 1—the environmental impact of the proposed
action,
Part 2—any adverse environmental effects which
cannot be avoided should the proposal be imple-
mented,
M Harbridge House, "The Impact of Proposed Non-Degradation Regulations on Economic Growth," EPA BOA #68-01-1361.
11 The following material is based on a piper prepared by Neil Orion1, "The Environmental Impact Statement Process," Director, Regional Liai-
son Staff, EPA, Washington, D.C. February 1973.
102
-------
Part 3—alternatives to the proposed action and
the results of not accomplishing the proposed
action,
Part 4—the relationship between local short-term
uses of man's environment and the maintenance
and enhancement of long-term productivity, and
Part 5—any irreversible and irretrievable commit-
ments of resources which would be involved in the
proposed action should it be implemented.
The proposed actions under Part I are limited pri-
marily to three types of action—licensing activities,
funding activities, and activities directly and wholly
undertaken by the Federal Government. More re-
cently the courts have ruled that reconstruction of a
segment of highway was a Federal action because,
even though that segment was not Federally-funded,
the overall project was Federally-funded.
The environmental effects under Part 2 have been
defined very comprehensively in the Federal guide-
lines. The environment is not to be considered just
in terms of air or water pollution or physical changes
in the land. The environment also includes the social
environment. The environmental effects must take
into account the adverse effects in excess of those
created by existing uses in an area. Thus the effects
in a non-degraded area could be recorded as being
very different than the effects of a similar ambient
change in a degraded area. Also, environmental ef-
fects must take into account the cumulative harm to
an environment.
Consideration of alternatives under Part 3, as this
process has come to be defined by Federal guide-
lines, could contribute significantly to estimating a
MDC:
"Agencies must indicate in their procedures that
all reasonable alternatives and their environmental
impacts are to be discussed, including those not
within the authority of the agency. Examples of
specific types of alternatives that should be con-
sidered in connection with specific kinds of actions
should be given where possible. Such examples
should include, where relevant:
(1) the alternative of taking no action;
(2) alternatives requiring actions of a significantly
different nature which would provide similar
benefits with different environmental impacts
(e.g. a fossil fuel vs. a nuclear power plant);
(3) alternatives related to different designs or de-
tails of the proposed action, which would
present different environmental impacts (e.g.
pollution control equipment on a nuclear
plant).
In each case, the analysis of alternatives should
be sufficiently detailed and rigorous to permit in-
dependent and comparative evaluation of the
benefits, costs and environmental risks of the pro-
posed action and each alternative." 12
Thus, the process of analyzing alternatives has the
potential for generating not only physical data about
a project, but also data about the value associated
with a change in the environment.
However, the greatest potential for providing in-
formation about how the public values environmen-
tal damages reported in an EIS is in the public com-
ment period, assuming that there is well-prepared
EIS. For the first time in many instances, concerned
citizens will have the technical information on the
environmental consequences from the introduction
of a new source of residuals and an opportunity to
state formally their opinion. Draft EIS's must be
available to interested public and private groups for
comments and any substantive comments must be
included in the final EIS. This process allows them
to record their valuation of the environmental im-
pacts.
Since public and private groups in all parts of the
U.S. can comment on EIS's, the comment process
has the potential for soliciting local, State, regional
and national concerns. If there are numerous ex-
pressions of non-local concern, they would provide
at a minimum an indication that significant values
are being lost and at a maximum a basis for a shift
in the MDC depicted in Figure 3 a. The new MDC
would be a more accurate indicator of value because
it would be a summation of the value of all interested
parties for a jointly shared good.
This type of information for estimating MDC will
be generated for more projects in future years be-
cause there is a growing trend in the States to adopt
requirements for EIS's like those of the Federal
Government. Fifteen States and Puerto Rico now
require impact statements for a wide range of activi-
ties (sometimes including local and private projects)
significantly affecting the quality of the environment,
and several others apply the process to limited classes
of projects. At least twenty other States have such
requirements under consideration.13
In concluding this section, I do not want to leave
an overly optimistic impression about how much
data EIS's will provide for estimating MDC. Esti-
mation of a total damage cost curve, from which the
MDC is derived, requires physical and economic
data over a range of abatement levels. Physical and
economic data over such a range are very difficult
if not impossible to obtain. However, EIS's do have
the potential for generating a few physical data
points and a proxy for the value of the environmental
damages. As limited as this information might be, it
would be more information than is usually available
« U.S. Council on Environmental Quality, "Guidelines on the Preparation of Environmental Impact Statements," 40 CFR 15.8A4.
is A more ««ns]j' dff£LT,tion of State programs Is available in Thaddeus Trzyna, "Environmental Impact Requirements in the States," EPA
Contract #68-01-l»l» U»74).
103
-------
for determining whether a new source should be
permitted in an area.
VI. CONCLUSION
The proposed economic definition of "no signifi-
cant deterioration" suggests that the permissible, that
is non-significant, changes in ambient air quality
vary within and among non-degraded areas. The
size of the permissible change would depend upon
the initial level of ambient quality and the costs and
damages functions.
The operational definition of NSD preferred by
EPA would require States to classify their territory
into three zones and the Federal Government to
specify permissible changes in each zone. The area
classification definition is partially consistent with
the economic definition. It is consistent because it
incorporates the idea that the level of permissible
changes ought to vary within and among different
areas and it is based on the cost of control. It is not
entirely consistent because it is based on a limited
comparison of the benefits and costs associated with
the zonal limitations.
As a practical matter, estimation of a national
damage function associated with the permissible
changes is an impossible empirical task. One viable
option, however, is to estimate the local MDC re-
sulting from the residuals generated by a new in-
dustrial source. Environmental Impact Statements,
which are micro-evaluations of environmental con-
sequences, are a potential source of data for this
task.
BIBLIOGRAPHY
Anderson, F., NEPA in the Courts (Baltimore: The
Johns Hopkins Press, 1973).
Freeman, A. M. Ill, Haveman, R. H. and Kneese, A. V.,
The Economics of Environmental Policy (New York:
Wiley, 1973).
Harbridge House, "The Impact of Proposed Non-
Degradation Regulations on Economic Growth," EPA,
BOA #68-01-1561, November 1973, Volumes I-III.
Krutilla, J., "Conservation Reconsidered," American
Economic Review, 57 (December 1967).
Orloff, N., "The Environmental Impact Statement Proc-
ess," speech, February 1973.
Sierra Club vs. Ruckelshaus, 344 F. Supplement 423
(D.D.C. 1972) affirmed, 4 ERC 1815 (D.C. Circular 1972).
Trzyna, T., "Environmental Impact Requirements in the
States," EPA Contract #68-01-1818 (1974).
U.S. Department of the Interior, "Navajo Project En-
vironmental Statement," September 1971, draft.
"Navajo Project Environmental
Statement," February 1972, final, Appendix I (Review
Comments).
Statement on Anti-Degradation,
September 1968.
U.S. Federal Register (July 16, 1973), p. 18987.
Weisbrod, B., "Collective Consumption Services of Indi-
vidual Consumption Goods," Quarterly Journal of Econom-
ics, 78 (August 1964).
APPENDIX A
TECHNICAL BASIS FOR ZONE II
PERMISSIBLE CHANGES IN
AMBIENT AIR QUALITY
Table A-l shows the impact of five point sources
on four Air Quality Control Regions. Concentrations
were computed after the application of Best Prac-
ticable Control Technology. High concentrations
caused by the power plant are due to several as-
sumptions. These assumptions called for an 850
Table A-l
Predicted Impact ( g/m3) of Point Socrces
in Four ACQR's*
Source Type
By-Product
Coke Oven
Cement Plant
Petroleum
Refinery
Power Plant
Municipal
Incinerator
RURAL
Open Terrain
SO
24- An-
hr. nual
1.6 0.3
110 23
95** 13
TSP
24- An-
hr. nual
0.7 0.6
0.7 0.1
15
0.6 ..
0.1 ..
Valley Terrain
SO
24- An-
hr. nual
2.4 0.3
122 26
437** 16
TSP
24- An-
hr. nual
2.1 0.6
0.5 0.1
25
0.8 ..
0.1 ..
URBAN
Open Terrain
SO
24- An-
hr. nual
0.6 0.2
100 16
89** 8.7
TSP
24- An-
hr. nual
0.8 0.6
0.7 0.1
20
0.6 ..
0.1 ..
Valley Terrain
SO
24- An-
hr. nual
2.4 0.2
122 26
447** 9.9
0.1 ..
TSP
24- An-
hr. nual
2.1 0.7
0.4 0.1
50
0.8 ..
0.1 ,.
•AQCR: Air Quality Control Region
••High concentrations due to short stack (122 meters) and targe size (850MW)
NOTE: The figures in this table represent those maximum ground level concentrations estimated to accompany the Ben Ltvil of Technology.
104
-------
MW plant, burning 2% sclfur coal with a stack of
122 meters. If the stack height had been greater, the
contribution would have been less.
Table A-2 shows SO2 and TSP concentrations
resulting from 14 power plants in the midwest United
States. Twenty-four hour results are based on ob-
served stability wind readings obtained from meteor-
ological observations at nearby airports for a period
of 1 year. Existing output of SO2 and TSP were
modified to reflect compliance with New Source
Performance Standards (NSPS).
Table A-3 shows 1-hour SO2 concentrations
around five coal-fired power plants. Column 4 shows
measured SO2 and column 5 modified SO2 concen-
trations which are likely if the plants were in com-
pliance with NSPS.
Table A-2
Maximum 24-Hour Concentrations for Selected Power Plants
Using Observed Stability Wind Roses1"
Plant
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Capacity
(MW)
444
414
53
75
908
133
221
450
531
100
396
308
59
1,304
Volume
(M8/sec.)
177
182
25
60
130
88
100
456
1,218
51
210
70
200
440
Stack Gas
Temp. (°F)
430
410
500
430
400
440
436
415
361
405
460
450
450
422
Stack
Height (m)
91
76
31
60
91
44
92
168
152
69
76
65
83
208
24-hr.
SO (ug/ms)
52.8
52
24.4
9.5
106.4
54.3
18.5
21
36.1
17.3
29.3
57
9.5
56
3-hr.
SO (Mg/m')
160.03
227.36
75.4
52.78
513.3
309.72
88.16
48.72
27.0
71.0
175.16
356.7
55.1
137
24-hr.
TSP (ug/m11)
7.5
7.8
10.2
2.4
13.7
5.6
3.5
1.9
5.1
4.9
8.1
0.59
1.2
•Existing coal burning plants, maximum estimated ground level concentrations assumed to meet NSPS by burning 0.7% sulfur coal.
Table A-3
Impact of Selected Power Plants on 1-Hour Air Quality'1'
Fuel Mtasured
Source Description
1000 MW Plant
Flat Terrain
2-450 ft. Stacks
1500 MW Plant
2-265 ft. Stacks
10-250 ft. Stacks
120 MW Plant
2-265 Ft. Stacks
455 MW Plant
Rolling Terrain
3-400 ft. Stacks
1700 MW Plant
Flat Terrain
2-600 ft. Stacks
1-800 ft. Stack
Present
1.5%S
Coal
3%S
Coal
1.3%S
Coal
2.2% S
Coal
3.8% S
Coal
Assumed 1-hr. SO
0.7 %S 1048 ug/m'
Coal
0.7%S 2358 u8/m'
Coal
0.7%S 733 ug/m'
Coal
0.7 %S 825 ug/m'
Coal
0.7%S 2600 ug/m'
Coal
Modified Distance to Max
3-hr. SO Concentration (km)
273 u8/m'
319 u8/m'
273 ug/m'
150.8 ug/m'
304 ug/m'
5-6
4
3
3
18
Maximum estimated average ground level concentrations
105
-------
Determining Acceptable Levels of Health and Environmental
Damages
Fred H.Abel*
I. INTRODUCTION
This conference is concerned with the process of
developing comprehensive environmental quality
management plans. These plans include a determina-
tion of the maximum pollutant environmental load-
ing (MPEL), i.e., a determination of the maximum
emissions allowed in any part of the region at any
point in time. The determination of the MPEL, in
turn, depends upon the level and kind of economic
activity allowed, the cost society is willing to pay to
obtain the desired level of environmental quality,
the level of damages from pollutants society will
accept, the distribution of costs (who pays costs)
and the distribution of benefits (who receives bene-
fits).
Societies must realize that the determination of
MPEL does depend upon these factors and so these
factors should be explicitly considered in the decision
processes. In deciding to control pollution, societies
have decided to trade off some economic growth, less
expensive products, or better quality products for a
better quality environment. This process of con-
sidering and then trading off the advantages of ad-
ditional pollution control against the cost of ad-
ditional control is called cost-benefit or trade-off
analysis.
Whenever a decision is made, there is, either im-
plicitly or explicitly, consideration given to the ad-
verse effects (costs) and desirable effects (benefits)
that will result from the decision. Our objective as
scientists is to make these trade-off considerations as
explicit and as full of knowledge as possible.
The presentation will proceed as follows. The con-
cept of damage functions is developed with attention
being given to the kinds of environmental pollution
costs and health damages. Problems of quantifying
damage functions and in turn valuing them in eco-
nomic terms are discussed. The concepts of social
costs of control and cost functions are developed.
The principle of comparing costs to benefits is dem-
onstrated in terms of economically valued functions.
This process of balancing costs with benefits is ex-
tended to include complications arising when many
costs and benefits cannot be valued in economic
terms. Finally, the status of cost-benefits as a cri-
terion for determining acceptable levels of pollution
in the United States is presented.
II. DAMAGE FUNCTIONS
There are many adverse effects of emitting pol-
lution into the environment. Most societies are con-
corned with the health and environmental damages
from pollution.
The number and kind of effects society suffers
from pollution are a function of the level of pollution
and the exposure of people to it. Curve A in Figure
1 shows the number of effects increases as the level
-of pollution increases. The effect function measures
the rate of incidence of a specific effect, such as a
premature death, an asthma attack, wildlife de-
stroyed, crops destroyed, etc. Similar functions would
be determined for each kind of effect.
Curve A in Figure 1 is a hypothetical curve: the
actual curve could be quite different for each pol-
lutant and each kind of effect. If a pollutant causes
few effects below a given level (threshold) but many
effects above the level, then the damage function
would look more like Curve B in Figure 1. The ac-
ceptable level of pollution depends upon the exact
slope of the damage function and in turn, on the
location of point PI, if there is a threshold. In prac-
tice, we may have a function like A but, because only
the part of the function above point P3 has been de-
termined due to the difficulty of monitoring and de-
termining effects at low levels of pollution, point
P3 may be misinterpreted as a threshold.
The functions A and B in Figure 1 imply a very
precise relationship between the level of exposure
and the number of effects. Actually, a statistical pro-
cedure is needed to estimate these curves. Thus, a
point on a curve, like point Z, is only the average or
the expected number of effects. There is a frequency
distribution of number of effects from exposure level
P2 around the point Z as indicated by the curve H.
There are two reasons for the frequency distribution
•Economic Analysis Branch, Washington Environmental Research Center, U.S. Environmental Protection Agency
107
-------
H. One is the natural variation in individuals' or
environmental response to pollution exposure P2.
Second, there are errors in the data used to quantify
the relationship. The variance becomes relatively
large at low levels of exposure, primarily because of
the data problem.
If the variance of the function A is known, then
confidence limits can be estimated and confidence
bounds like A' and A" can be displayed as in Figure
1. The existence of variance around the damage
function makes the process of determining the ac-
ceptable level more complex and less certain.
The damage function is based on exposure and
not on emissions. This has a significant impact on
regional environmental management because ex-
posure can be reduced in several ways. One way is
to reduce the level of pollutants emitted into the
environment. A second way is to shift the location
of either the receptors (the people) or the polluters,
such that exposure is reduced. A third way is to re-
duce exposure by removing the pollutants from the
environment at the receptor, i.e., air conditioning
buildings to avoid air pollutants, or pretreating water
to remove pollutants.
III. ENVIRONMENTAL POLLUTION
COSTS
Pollution imposes burdens (pollution costs) upon
society by causing deterioration in materials and liv-
ing things (damages), by causing concern and dis-
satisfaction with the quality of the environment
(psychic costs), and by causing extra expense
(avoidance costs) to reduce the potential damage
and psychic costs.
Damage Costs
Included in this category are those effects on
physical things, including human health, that result
from exposure to pollutants. Effects on materials,
buildings, vegetation and direct expenses involved
with adverse health effects are examples. These dam-
ages result in decreased value of an item, decreased
useful life of a structure or item and direct out-of-
pocket costs associated with the damage (i.e., doctor
or medicine costs). Health damages will be dis-
cussed in some detail below.
In estimating damage from pollution, dose-re-
sponse relationships that show how much more
Pi P2
LEVEL OF POLLUTION
FIGURE 1. DAMAGE FUNCTION
108
-------
quickly things deteriorate and become nonfunctional
as a result of being exposed to different levels of
pollution are needed. Dose-response relationships
will allow estimates of the extra costs to industry
and to society for replacing equipment and buildings
more frequently. In the United States, estimates have
been made of the level of materials damage from
several air pollutants. These estimates are shown in
Table 1. (Blank cells indicate that estimates are not
available.) This was done by determining the levels
of exposures and quantities of materials exposed
under the current situation and estimating how much
less deterioration there would be if pollution levels
were reduced to meet the standards. The materials
damage estimates are for unpainted metals, paint,
rubber, and textiles. It was determined that the eco-
nomic values of damages to many other materials
such as concrete, glass, stone, and leather are very
small. Estimates of damage to vegetation were de-
termined in a similar way. That is, estimates were
obtained of: (a) how the yield of agricultural crops
or the growth of horticultural plants are related to
levels of pollution, (b) the total acreages of these
commodities exposed to pollution and (c) the value
of the damage. The damage value was assumed to be
the loss in revenue to producers because of reduced
yields.
Psychic Costs
This category includes the value of dissatisfaction
and reduced quality of life that results because pol-
lution exists. It includes the value of the dissatis-
faction that results because (a) pollution interferes
with enjoying the natural environment, and with vis-
ibiilty of scenic vistas, (b) populations of wildlife
and natural vegetation have been reduced, (c) pol-
lution causes discolored vegetation, dirty buildings
and littered waterways and (d) people know the
environment is being violated by indiscriminant
dumping of wastes. It includes the value people as-
sociate with maintaining as many natural or high
quality areas as possible because this gives them
more choices for recreation. It includes the value
people place on leaving future generations a better
quality environment, and it includes the value of fear
that people have for the welfare of the environment
and the welfare of mankind. People are willing to
live with some risk. At some point, they will be will-
ing to give up some goods and services to reduce
risk. Determining this level of acceptable risk is a
major need in environmental decision making.
Estimating psychic costs is very difficult. It is al-
most impossible to obtain a dose-response relation-
ship because the response is primarily in the state
of mind of a person. A dose-response relationship is
obtained either by observing the behavior of people
as changes in environmental loadings occur, or get-
ting them to express their changed state of mind in
some terms that can be quantified.
For example, if paniculate matter is emitted into
the air, it can cause soiling of buildings and prop-
erty; and if people find this reduction in aesthetic
quality of the buildings objectionable, they would
spend additional energy, time and resources in clean-
ing up the properties. The level of effort expended
would be a measure of the damage or social costs
from particulate matter emissions. Another method
of estimating psychic costs associated with aesthetics
in urban areas is concerned with the value of prop-
erty or the level of rent. In either case, the value of
the property or the value of the rent will be reduced
if people perceive an aesthetically less pleasing en-
vironment. The aesthetic values in Table 1 were
derived in this manner.
The extra cost of traveling to alternative recre-
ation sites would be an estimate of the value of the
psychic costs of pollution. The value of water recre-
ation costs in Table 1 were derived in this manner.
Table 1. United States Annual Pollution Costs, 1970 '
(billions of dollars)
Effect
Aesthetics and Soiling '
Human Health
Materials
Vegetation
Animals
Recreation
Water Supply
Pollutant
Air
SO,
Best
2.9
1.9
0.6
Range
1.7-4.1
0.7-3.1
0.4-0.8
Particulates
Best
2.2
2.7
0.2
Range
1.7-4.1
0.9-4.5
0.1-0.3
Ox
Best Range
0.9 0.5-1.3
0.2 0.1-0.3
Water
Best
0.1
0.7
2.6
0.8
1.7
Range
0.03-0.18
0.3 -1.0
2.5 -2.7
0.6 -1.0
1.2 -2.2
1 Blank cells indicate that values have not been estimated.
* Based on property value changes.
Source: Waddell, Thomas E., "The Economic Damages of Air Pollution" unpublished report, National Environmental Research Center, Environ-
mental Protection Agency, Research Triangle Park, N.C., February 1974. Tihansky, Dennis P., and Richard Walsh, "A Status Report on the Na-
tional Benefits of Water Pollution Control" Washington Environmental Research Center, Environmental Protection Agency, Washington, D.C.,
November 1973.
109
-------
People's fear of environmental catastrophies or their
dissatisfaction with the knowledge that the environ-
ment is not being preserved in the best possible
quality is determined by a survey which would
ask them to make tradeoffs among many different
things they want. For example, would they rather
have cheaper electricity or more miles of pollution-
free water? In practice, these survey techniques have
only recently been tried and are only of limited suc-
cess.
Even though estimating the psychic costs as-
sociated with pollution may be very difficult, this
does not mean that it is unimportant. In fact, many
of the environmental goals and many of the objec-
tives of a comprehensive environmental management
plan are derived from the fact that people desire
these goals that affect their state of mind and their
ability to enjoy the world.
Avoidance Costs
Avoidance costs, the third kind of costs resulting
from pollution, are out-of-pocket expenditures in-
curred to avoid damage and/or reduce psychic costs.
Examples of this are people incurring extra ex-
pense to paint metals so that they will not deteriorate,
people moving from a polluted location to an un-
polluted one to avoid getting sick, people moving
from a polluted environment to a non-polluted one
to enjoy the increased aesthetics. It includes the de-
cisions to stop growing crops that are affected by
pollution and growing otherwise less profitable crops
that are resistant to the pollution. It includes people
incurring costs by pretreating water or air so that
the pollutants are removed before the medium is
consumed. Many of these costs can be estimated
if the share of the total expense that is incurred be-
cause of pollution can be determined. Only that
share should be counted as pollution avoidance costs.
IV. POLLUTION HEALTH COSTS
Most societies are very concerned with the effects
of pollution on the health of its citizens. This effect
deserves and has received a good deal of attention.
There are five ways that health effects impact upon
the citizen, his family and community.
1. Increased doctor, medicine and hospital costs
2. Value of loss from premature deaths
3. Value of time lost from work or other activity
4. Reduced productivity of work if performed
5. Value of suffering from sub-clinical effects
Some or all of the above costs would apply to any
given health effect. Some of these components of
damages can be estimated in economic terms, al-
though it is difficult. The biggest problem with esti-
mating the value of medical costs, lost activity days
and reduced productivity is determining the share
of the total observed effect that is attributable to the
pollutant when other factors also affect these costs.
The problem of placing an economic value on pre-
mature death has often been described as putting a
value on life, when it is the length of life that is
affected. Determining a value for reduced length of
life is very difficult, particularly if age, occupation,
health status, and family status obtain different
values for individuals. There is no way to determine
the correct or true value of an additional day of life,
and none will be proposed here. However, it is clear
that the value cannot be infinity, which is the value
people assign if they insist that a life should be saved
(actually increase length of life) regardless of costs.
It is also clear that the value is not zero, which would
imply that nothing would be done to extend length
of life, even if it were very cheap or free. Given the
importance societies attach to premature deaths and
the difficulty of placing a value on them, it is often
necessary to present the number of days lost because
of premature deaths to the decision maker and let
him make the tradeoff.
Determining the value of suffering is also difficult.
In principle, people are willing to put up with some
suffering rather than give up goods and services they
enjoy. At some point, however, they would be will-
ing to give up some goods and services in return for
lower levels of suffering. The value of goods and
services people would give up (their willingness to
pay) for a given reduction in pollution exposure to
reduce suffering would be the correct value to use.
In the absence of good estimates of these values, the
total number of hours of suffering as a function of
pollution exposure should be presented as the dam-
age function.
Because of the difficulty of putting economic values
on premature death and suffering, the health dam-
ages from pollution can be presented as three dam-
age functions: one showing the number of days lost
because of premature deaths, one showing hours of
suffering, and the third showing the economic value
of all other health impacts. One health damage func-
tion that has been estimated is shown in Figure 2.
The damage function shows the percentage of a
normal, healthy population between ages 10 and 65
experiencing some ill effects as oxidant levels in-
crease. The upper and lower curves are the 95%
confidence bounds.
Also, functions of mortality rate versus paniculate,
sulfate, and socioeconomic variables were estimated
with regression analysis, and a significant, positive
linear function for each pollutant was obtained
[Lave, L. B. and E. P. Seskin. Air Pollution and
Human Health. Science 769(3947) :723-733, Au-
gust 21, 1970]. The number of persons to die per
100,000 population increased 0.041 per microgram
110
-------
a
UJ
u
DC
100 r-
90
80
70
60
50
40
30
20
10
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
OXIDANTS, PARTS PER MILLION
FIGURE 2. PERCENTAGE OF NORMAL POPULATION AGE 10-56 AFFECTED BY OXIDANT EXPOSURE
per cubic meter increase of average annual par-
ticulate reading, and 0.071 each microgram per
cubic meter increase in sulfate level. A 95% confi-
dence bound for the particulate coefficient is 0.009
to 0.073, and for sulfate coefficient is 0.027 to 0.115.
These rates are reliable only for pollution levels near
the mean of 4.7 micrograms sulfates and 120 micro-
grams particulates per cubic meter and with the level
of other variables held constant at their means. In-
cluding both pollutants in the analysis is superior to
estimating the effects of each and ignoring the level
of the other. If synergism exists, it can be incor-
porated by the inclusion of interaction terms. The
results would then indicate that the effect of an in-
crease in one pollutant depends upon the level of
the other. Thus, the decision about the acceptable
level for both pollutants would have to be made at
the same time.
Information presented to the decision maker when
synergism is present must be different than that dis-
cussed above. For example, instead of a single dam-
age function for a pollutant, a set of damage func-
tions must be presented. Each function in the set
would assume a different level for the synergistic
pollutant.
Health effects damages of water pollution have
also been determined. This was done by determining
the number of illness outbreaks in the U.S. and de-
termining the share of those attributable to water
pollution. A value per illness was then used to ob-
tain the estimate of $0.7 billion per year as shown
in Table 1. These estimates are crude and dated.
It is expected that more accurate and comprehensive
estimates would be several times larger.
V. COST FUNCTIONS
When a decision is made to control pollution by
reducing the level of current emissions, by prevent-
ing the increase of emissions or by relocating the
source of the emissions or receptors, it imposes costs
on society. There are costs of administering and
enforcing the decision, costs to plants for add-on
111
-------
equipment, process changes, or changed input mix,
as well as costs to society of poor quality products,
increased unemployment and reduced economic
growth.
These costs can be quantified in economic terms,
except for the costs (burdens) to society from de-
creased employment and reduced economic growth.
Thus, it is necessary to supply the decision maker
with information on three kinds of adverse effects:
employment, economic growth and economic costs.
The costs of obtaining a given level of pollution
emission control can vary substantially, depending
upon the implementation strategy used. If the time
period for obtaining control is lengthened, the costs
would be reduced as plants have time to phase in
new equipment and because new, more efficient tech-
nology may be developed. If control applies to all
plants, only large plants, only new plants, etc., costs
will vary widely. Thus, cost analysis requires that the
implementation strategy be pre-determined, several
alternatives be studied or the least-cost (econom-
ically most efficient) method be determined. In the
least-cost case, the required level of control is based
on the cost of control. The optimum solution has the
cost of an additional unit of control equal for all
plants. In all cases, it will be necessary to determine
the number, distribution and kinds of sources, the
control options and their associated costs available
to each source, and the probability that the control
costs would be so high as to force the closing of the
plant.
Studies have shown that the least-cost strategy has
significantly lower costs than strategies designed to
impact each source uniformly. A least-cost solution
is obtained by linear programming, where each kind
of source is considered, available control options
are included as activities and total emissions are the
constraints. The solution specifies the control bption
and emission level for each kind of source such that
total costs are a minimum. Alternative solutions are
obtained as the constraints (total emissions) are
varied and a least-cost cost function is derived.
Cost functions have been estimated for selected
regions. The curves in Figure 3 show the difference
between the least-cost and a uniform plant strategy
that was being implemented. Even though the least-
cost strategy is economically most efficient, there
may be political or institutional constraints that pre-
vent it from being used. Nevertheless, least-cost
functions should be derived in all studies so the
decision maker realizes what is given up to accom-
modate these other factors.
VI. COMPARABILITY OF COSTS
AND BENEFITS
As stated earlier, when a decision is made, the ad-
vantages and disadvantages of that decision and
alternatives to that decision have been compared,
either explicitly or implicitly. In this process of com-
paring advantages (benefits) and disadvantages
(costs), it is useful if they are measured in the same
units. Although many different units can be used,
monetary units have many advantages. One ad-
vantage is that many of the effects are measured in
monetary units such as the costs of add-on control
equipment and medical expenses from pollution dam-
age. Another is that the monetary units are easily
understood. The difficulty of measuring all the ad-
verse and beneficial effects in economic terms has
been discussed. In practice, it may be necessary and
desirable to present the decision maker with several
functions, measured in different units, i.e., number
of premature deaths, hours of suffering, reduced
employment, etc. Here it is assumed that they can
all be measured in economic (or other unit) terms
so that determining the socially desirable level of
pollution can be explained. Later, this assumption
will be relaxed and the process of making trade-off
decisions when effects are measured in different units
will be explored.
In making decisions about the acceptable level of
control, the change in damages must be compared
to the changes in costs for a proposed change in
ambient quality. These changes are called marginal
damages and marginal costs. Marginal damage func-
tions and the negative of the marginal cost functions
with their associated variance are presented in Figure
4. The negative of the cost function is used to get
both marginal functions in a single quadrant. Both
sets of data, the costs and the benefits, have been
annualized so the figure depicts costs and damages
per year. Point P, in Figure 5 represents the socially
optimal level of pollution control. Except for very
unusual situations, that level of control will not be
zero emissions, zero effects, or uncontrolled emis-
sions, zero effects, or uncontrolled emissions. At
Point P, the marginal costs are just equal to the
marginal damages. It can be demonstrated that the
point where the marginals are equal is the optimum
level. The procedure for determining the optimum
point is not different for a damage function with a
threshold like function B in Figure 1.
The A', A" and C, C" functions reflect the fact
that neither the cost function nor the damage func-
tion is known with certainty. The shaded area in
Figure 4 shows the area within the confidence
bounds of both functions. Because of uncertainty,
any pollution level corresponding to a point in the
shaded area could be the optimum level. Thus, the
optimum level could, with the confidence implied by
the confidence bounds, be somewhere between P2
and P3. The more uncertainty we have about the
damage and cost functions, the larger will be the
range that contains the optimum level and the more
112
-------
"
1 .
K
u
_J
o
3 2
iu
oc
_1
10 20 30 40 SO 60
PARTICIPATES, MICROGRAMS/M3
FIGURE! COST FUNCTIONS
70
80
90
K
O
u
to
e
C' MARGINAL COSTS (LEAST COST)
P2 PI
LEVEL OF POLLUTION
A (MARGINAL DAMAGES)
FIGURE 4. MARGINAL DAMAGE AND MARGINAL COST FUNCTIONS OF
CHANGING THE LEVEL OF POLLUTION
113
-------
PA
EXPOSURE
us
z
E
GO
cc
SB
so
SA
PA PO
EXPOSURE
PA PO PB
EXPOSURE
DC
•3
<
a.
u.
O
6
FO.FB
PA PO PB
EXPOSURE
PA PO PB
EXPOSURE
"A
, "B
PA
EXPOSURE
cc
to
u
o
0
r
j GB
g
2
K
PA
EXPOSURE
PA PO
EXPOSURE
FIGURE 5. DAMAGE AND COST FUNCTIONS
PA po PB
EXPOSURE
difficult for decision makers to choose an acceptable
level of pollution with confidence that it is near the
correct level.
VII. DECISION MAKING WITH
MULTIPLE BENEFIT AND COST
FUNCTIONS
The current state-of-the-art in estimating both
costs and benefits does not allow the ideal of getting
all the effects collapsed into a single functional form
so that the optimum decision point can be easily
determined. In fact, many of the effects simply can-
not be measured at all. As stated earlier, many of
the costs, particularly when emission control is ob-
tained by relocation or by changing the process or
the product quality, also have not been measured.
So, the decision maker has to settle for estimates of
some key effects that are not in comparable units.
114
-------
For health damages there are three key relationships:
number of years/days lost because of premature
deaths, the total health costs including doctor and
hospital costs and lost days of work, and the hours
of suffering from sub-clinical effects, each as a func-
tion of air pollution levels. Four key variables on
costs are the total costs of add-on equipment with
the associated operating and maintenance costs, the
number of plants forced out of production, the de-
creased employment in the region, and the foregone
benefits of economic growth. The two key variables
for environmental pollution costs are psychic costs
and economic value of avoidance and damage costs.
Hypothetical functions for these nine functions are
shown in Figure 5.
The level of exposure P0 results from the existing
level of pollution. This becomes the base-line case
for the analysis. Decision makers could use a set of
functions as depicted in Figure 5 to make their own
trade-off decisions. However, there are many trade-
offs involved, and it may be easier if selected results
were displayed in a table. Results for several alter-
native control levels can be presented in such a table.
Examples are shown in Table 2. Alternatives A and
B are shown. The values in the cells are the changes
from the base case.
Alternative A considers a reduction in exposure
levels. Thus, in comparing it with the base case, the
tradeoffs are fewer premature deaths, (Do-Da),
fewer hours of suffering, (So-Sa), lower health costs,
(Ho-Ha), increased psychic benefits (Bo-Ba) and
reduced avoidance and damage costs (Ao-Aa);
traded off for higher control costs (Co-Ca), reduced
regional growth (Go-Ga), increased plant failures
(Fo-Fa), and increased unemployment (Uo-Ua).
A decision maker would have to weigh these different
effects to determine overall costs and benefits of
Alternative A.
Alternative B presents similar tradeoffs from al-
lowing pollution levels to be worse. Many alterna-
tives could be presented to the decision maker in this
format. Also, the number of effects can be increased
to include important subsets. If it were important
to the decision maker to know how premature deaths
were distributed between young and old, separate
functions would be generated and separate rows pre-
sented in the table.
The procedure described above demonstrates how
information for cost-benefits (trade-off) analysis can
be presented to decision makers when there is a
problem of weighing the different effects. It assumes
that functions are available for these effects. In fact,
very few functions have been estimated.
VIII. STATUS OF COST-BENEFIT
ANALYSIS IN THE UNITED STATES
Cost-benefit analysis has been used for many years
as a decision tool for water resource projects. Much
of the current state-of-the-art of cost-benefit analysis
for environmental objectives has come from this ex-
perience. However, environmental cost-benefit analy-
sis requires estimates of many benefits not important
to water resource decisions.
Cost-benefit analysis is required in many of the
laws governing the establishment and enforcement
of acceptable levels of control. However, require-
ments for its use vary considerably.
In the laws governing the primary ambient level
of air pollutants, the level is to be determined by
threshold health effects, and costs and benefits are
not to be considered. As it becomes clear that thres-
holds do not exist or are very low, consideration is
being given to shifting to a cost-benefit criterion for
determining levels.
Special laws were passed requiring emissions from
mobile sources to be rolled back 90%, with limited
Table 2. Selected Results From Damage and Cost Functions
Alternative Control Targets
Effects
Number of Premature Deaths
Hours of Suffering
Health Costs ($)
Number of Plant Failures
Control Costs ($)
Unemployment
Regional Growth
Other Benefits
Base
D
o
s
o
H
o
F
o
C
o
u
o
G
o
g
o
A
D -D
0 A
S -SA
O A
H -H
0 A
F -F
O A
C -C
O A
U -U
O A
G -G
O A
B -B
O A
B
D -D
0 B
S -S
O B
H H
O B
F -F
O B
C -Cn
O B
u -u
O B
G -G
O B
B B
O~ B
115
-------
consideration of the costs or benefits in setting this
level and date of attainment. Similarly, very stringent
transportation controls are being relaxed and post-
poned, in cases where it has become apparent that
the social costs would be very high.
The laws considering water quality vary in their
requirements for cost-benefit analysis. All plans by
states to implement water pollution programs are re-
quired to be evaluated in a cost-benefit analysis and
others require such analyses to establish effluent dis-
charge limits. If individual firms believe the costs of
complying with the limits exceed the benefits, they
can request a cost-benefit analysis; and if costs ex-
ceed benefits, they do not have to comply at that
limit.
Limits on toxic substance emissions to the en-
vironment are to be established on the basis of cost-
benefit analysis. Determination of whether a pesti-
cide can be used has to be made after consideration
of the costs and benefits.
Although requirements for cost-benefit analysis
vary among pollutants and change over time, increas-
ing consideration is being given to using it more in
making decisions. The need for improved methods
and improved data, particularly dose-response re-
lationships, is increasing rapidly.
116
-------
Methods for Abatement of Water Pollution
Robert B. Schaffer *
I. INTRODUCTION
The comprehensive environmental analysis of a
denned area requires the assessment of all sources
of pollution, their interrelationships and ultimate
control possibilities. This paper has two objectives:
1. to provide the participants of this symposium
necessary information on water pollution
sources and associated control strategies; and
2. to look at some of the more conceptual strat-
egies that may be termed Operational Con-
trol.
The former information includes the description of
levels of control and their cost for some represent-
ative industrial categories and the municipal or do-
mestic waste source. The diffuse or non-point
sources cannot be as precisely scoped. The infor-
mation, therefore, will only provide data for de-
cision-making and not make the decisions. The
strategies employed in Operational Control must
integrate the technologies described and incorporate
them in the decision-making process through the
development of basin or region plans. Specific data
on the relative costs and socio-economic benefits
have not been included, but they must be weighed
with all other aspects of the overall environmental
plan. Many social and legal/regulatory factors must
also be considered, and some of the obvious will be
briefly discussed.
II. POINT SOURCES
Point sources of pollution from municipalities, in-
dustry, commerce, and agriculture are the most con-
trollable and best lend themselves to the improve-
ment of water quality in a cost effective manner.
Technology is available to achieve the reduction of
most pollutants, to the limit of detectability, if cost
is no object. However, this option is not generally
available,'and other alternatives must be considered.
The single variable that provides the most poten-
tial for achieving the greatest removal from point
sources is time. A long-range strategy must be de-
veloped for the region to achieve maximum control
at the lowest cost. The objective is to do the best
job you can economically, not to get by as inexpen-
sively as possible, which is the tendency when op-
posing interests attack a common problem.
The optimum point source control strategy then
should have as its goal zero discharge of pollutants,
defined as the removal of undesirable constituents of
wastewater to levels of detectability. Zero discharge
does not mean the removal of all constituents nor
the elimination of the liquid discharge itself. In these
instances primarily, but in all others as well, the
pollutants so removed from discharges to waterways
become materials of concern in other environmental
areas. They may become air pollution problems
through incineration, solid waste problems through
conversion and disposal to the land, or a problem
due to their presence in products produced for con-
sumption and use. A water-borne discharge may
turn out to be the best ultimate disposal option for
the overall environment.
The most complex and difficult problems in de-
velopment of an effective environmental assessment
of an area are associated with an industrialized so-
ciety. Treatment or abatement technologies for the
varied and necessarily complex manufacturing proc-
esses stretch the limits of existing technology when
viewed in the context of zero pollution. Whether
municipal or industrial, a certain amount of that
which results from man's industrial society must be
tolerated.
One basic, but often forgotten, law in relation to
pollution control is the conservation of matter. Many
would have those materials that present environ-
mental problems just disappear. Unfortunately, they
will not, no matter how hard we try. There is no
such thing as infinite recycle or reuse. Therefore,
difficult technical, social and economic decisions
must be made with regard to each specific problem.
The goal of zero discharge must be approached,
therefore, in a step-wise fashion; Public Law 92-500
outlines the pragmatic, technology-based approach
now underway in the United States.
1. The application of Best Practicable Control
Technology Currently Available to all exist-
ing point sources other than publicly owned
•Permit Assistance and Evaluation Division, U.S. Environmental Protection Agency
117
-------
treatment works, i.e., the average of the best
practiced technologies to be in operation by
July 1977;
2. The application of the Best Available Tech-
nology Economically Achievable, i.e., the best
that could be built using possible scale-up of
pilot technology; to be in operation by 1983;
3. The application of the Best Demonstrated
Technology, including use of different manu-
facturing processes, to all new sources;
4. The application of pretreatment requirements,
which must control incompatible pollutants,
on discharges to municipal treatment systems;
5. The application of secondary treatment to all
existing municipal systems, this step on the
same time frame as Step 1 for other types of
discharges;
6. The application of Best Available Technology
to municipal treatment works, this step to co-
incide with Steps 2 and 3.
The application of this strategy is complicated by
the diversity of industrial categories and the ability
of technology and economics to cope with these
problems. In some categories it will be impossible
to achieve complete recycle and may well be un-
desirable. This paper examines, for purposes of
illustration, the application and costs of various
levels of pollution abatement in different industrial
and municipal situations. Both major industrial seg-
ments shown, Iron and Steel and Petroleum Refin-
ing, do not provide for the goal of zero discharge of
pollutants. Petroleum refining effluents can be con-
trolled more easily, primarily due to the technical
sophistication associated with the processing of the
raw materials into the various products. The Iron
and Steel Industry requires high volumes of water
for cooling and quenching and presents a less con-
trollable option, from a water recycle standpoint.
The Metal Finishing category presents the "no dis-
charge" option, primarily because of its size and
product recovery economics. In this instance, Step
2 (Best Available Technology) will be equal to
Step 3 (New Source Requirements) which is zero
discharge.
The treatment of municipal wastes using existing
technologies takes us up to the renovation of water
for beneficial reuse. Effluent quality may be tailored
to the use required, from water quality maintenance
through various stages of water reuse.
Petroleum Refining
A petroleum refinery is a complex combination of
interdependent operations, and the characteristics of
the wastcwater differ considerably for different proc-
esses. Each process is itself a series of unit operations
which causes chemical and/or physical changes in
the feedstock or products, and each operation may
have drastically different water usages associated
with it. Table 1 shows the discharge concentration
averages for both Best Practicable and Best Available
levels of treatment. The difference in the amount of
pollution discharged in this case is related to a sig-
nificant reduction in the volume of water used per
unit of production for Best Available Treatment.
Iron and Steel Industry
The industry is comprised of a complex of manu-
facturing processes and wastewater treatment oper-
ations, both biological and physical-chemical, which
may be used. The proposed Best Available levels for
the Iron and Steel industry indicate that it requires
and will continue to require large volumes of water
for the manufacturing process. Therefore, in long-
range planning, this consideration must be integrated
into the ultimate location of the manufacturing facili-
ties. Tables 2 and 3 show the discharge concentra-
tion averages for the industry for Best Practicable
and Best Available levels.
Table 1. Petroleum Refining Industry
Attainable Concentrations from the Application of
Best Practicable Control Technology Currently Available
and
Best Available Treatment
Parameter
Annual Average
(mg/1)
Biochemical Oxygen Demand
Chemical Oxygen Demand
Suspended Solids
Oil and Grease
Phenol
Ammonia Nitrogen
Chromium, Total
Hexavalent Chromium
Zinc
15.
80.
10.
5.
0.1
Removal
0.25
0.005
0.5
Table 2. Iron and Steel Industry
Attainable Concentrations from Application of
Best Practicable Control Technology Currently Available
Parameter
Concentration
(mg/1)*
Biochemical Oxygen Demand (BOD8)
Suspended Solids
Oil and Grease
Phenol
Ammonia (as Nitrogen)
Cyanide, Amenable to Chlorination
Chromium, Total Dissolved
Iron, Total Dissolved
Zinc, Total Dissolved
150.
50.
10.
3.
125.
7.
0.05
1.2
1.2
•With Recycle
Metal Finishing
The basic treatment model is that of unit process
stream precipitation. For essentially all of the param-
eters, Best Practicable involves precipitation, which
includes coagulation, sedimentation, flotation, and
118
-------
Table 3. Iron and Steel Industry
Attainable Concentrations for Application of
Best Available Control Technology Economically Available
Concentration
Parameter (mg/) *
Table 4. Reclaimed Water Quality
First Year of Plant Operation
South Tahoe Public Utility District
Biochemical Oxygen Demand
Suspended Solids
Oil and Grease
Phenol
Ammonia (as Nitrogen)
Cyanide, Amenable to Chlorination
Chromium, Total Dissolved
Iron, Total Dissolved
Zinc, Total Dissolved
20.
10.
10.
0.5
10.
0.25
0.05
1.2
1.2
•Treatment of Recycle Slowdown
finally, filtration. Evaporation recovery and ion ex-
change offer other effective approaches, and in some
cases, chemical oxidation (for cyanides) and chem-
ical reduction (for chromium) are needed as part
of the treatment process.
The costs of achieving effluent reduction to "no
discharge" have been examined for the industry.
The cost may be about the same as for conventional
chemical treatment because it may be possible to
recover and reuse sufficient chemicals and water to
offset the costs of treatment.
Municipal Wastewater Treatment
In making determinations with regard to handling
of domestic wastewaters in the context of area plan-
ning, certain considerations become more important
than others. First, the combination of compatible
industrial wastes may present economical solutions
and must be considered. Secondly, the existence of
defined areas of development may limit the choice
available, usually on social and economic grounds.
Thirdly, where new systems may be designed, there
are fewer constraints on renovating water for reuse.
Of immediate interest in the handling of waste-
water management problems in areas already under
development is the upgrading of present treatment
plants. Technology in the past decade has provided
many innovative upgrading procedures to meet de-
ficiencies in existing processes. Performance and cost
data on the upgrading of facilities have been
developed.
The consideration of advanced waste treatment
processes for water renovation and potential reuse
can result in a viable, economical alternative. Suhr
has described cost and performance data for the
South Lake Tahoe Public Utility District. The re-
sults of treatment meet or exceed the stringent re-
quirements imposed. These results are summarized
in Table 4.
All areas do not require the same degree of waste
treatment, and all of the tertiary processes utilized
at Lake Tahoe are not required at each proposed
tertiary treatment plant. Because they are unit chem-
Quality Parameter
Median
Maximum
BOD mg/1 1.0
COD mg/1 10.8
Phosphate (as Phosphorus) mg/1 0.42
Jackson Turbidity Units 0.6
Suspended Solids mg/1 0.5
Coliform MPN/100 ml 2.2
7.8
24.7
4.8
6.0
7.8
No two consecutive
samples greater
than 16.
ical processes, the Tahoe processes can be adapted
to provide treatment to meet almost any practicable
pollution abatement criteria.
The overall consideration of point source control
is summarized in Figure 1. Considerations for the
implementation of pollution abatement beyond Best
Practicable, which is assumed to be the base accept-
able level, must consist of an area's water use re-
quirements, volume available, water quality, and the
assimilative capacity of these waters to maintain de-
sired quality. In general, the costs for achieving the
next level, that is, Best Available Treatment, will be
roughly three times the cost of Best Practicable.
To go beyond, to approach zero discharge of pol-
lutants, will increase the cost significantly. The de-
sirability of this option, where technically feasible,
will depend directly on the value of the water in that
specific location.
III. NON-POINT SOURCES
Characteristics
The fact that nearly all non-point sources are
rainfall/runoff dependent and are associated with
highly variable natural phenomena has profound
implications for their control. It is extremely diffi-
cult to characterize a non-point source or evaluate
control performance dependent on such variables
as rainfall and growing season which cannot be pre-
determined, controlled or accelerated. Accurate as-
sessment of the problem may require a significant
period of time (one to three years). Because there
is no typical year or season, a multiyear study may
be required.
Non-point sources have enormous diversity. Dis-
charges are noncontinuous and highly dependent on
location and local conditions. No single set of con-
ditions would be representative, and an extrapola-
tion of data from specific studies is risky. These
sources have no defined discharge point and cannot
be conveniently or economically measured. The dis-
charge or load must be related to some measurable
source feature or activity. Because control by ef-
fluent treatment is impractical in most instances, we
must control the activity by causing a change in the
pollutant's manufacture or use. Source control prob-
lems involve signivcant institutional, legal, and socio-
economic interfaces.
119
-------
>
UJ
_J
>•
o
o
QC
BAT
(10 Years
BPT
(5 Years)
TOOl
LU
POLLUTION %
FIGURE 1. RELATIVE COST OF IMPROVING
WATER POLLUTION CONTROL
Proposed Control Strategies
Relief from non-point source pollution requires,
as a first step, the assessment of the problem in order
to set proper program priorities. The leading models
for non-point sources and representative land use
patterns should be developed. The probably effec-
tiveness of available control options must be assessed.
Basin load allocation models must be developed to
allow trade-offs in individual situations between the
control of point sources and non-point sources.
The following is a brief discussion of six specific
non-point source areas.
Irrigation Return Flows
This water quality problem involves all irrigated
agriculture, including water and crop management
procedures and accompanying use of chemicals.
Major pollutants from return flows include salts,
120
-------
sediments, nutrients, toxic and other bioactive ma-
terials. The magnitude of the problem is indicated
by the fact that approximately 10 percent of the
cultivated acreage in the United States is under irri-
gation, producing 25 percent of the national food
and fiber. Doubling of irrigated acreage by the year
2000 with a 10-20 percent increase in total water
use is projected. With existing practices, total salt
discharged could double. Salt has extensive effects on
agriculture to the extreme of rendering farm land
no longer productive. Fifty to seventy percent of
man-induced salt sources in receiving waters can
come from irrigation return flows. This pollutant
will present a major interbasin problem and can af-
fect international uses of water. Control strategies
must consider soil and water management practices
in each climatic/geographic location and the strategy,
therefore, will vary with location. Source manage-
ment is the most cost-effective general control option
now available, entailing water management as well
as the management of the application of various
chemicals used in irrigated agriculture.
Animal Feeding
All animal raising operations produce pollution
from the oxygen-demanding materials, nutrients,
salts, and pathogenic organisms present in waste
waters from the operations, with significant impact
in any basin where such activity is a major factor.
Present trends relate to the economics of larger
operations. In the United States sixty percent of the
marketed beef cattle are produced by fewer than two
percent of the operations. Control strategies for many
operations include land disposal of residues pro-
duced, and containment and spray irrigation with
water resulting from such operations. The economics
of hauling wastes, and climate are problems here.
An ultimate solution is likely to be by-product re-
covery and land disposal of effluents.
Agriculture Runoff
This segment includes crop raising, animal pas-
turing, forestry and other rural land uses from all
nonirrigated agricultural operations. Major pol-
lutants include sediments, nutrients, oxygen-demand-
ing materials, organic and inorganic residues from
agricultural chemicals, toxic and other bioactive
materials, heavy metals, and pathogens. This activity
in any given basin may cover at least fifty percent
of the land area, resulting in significant amounts of
pollution due to runoff from precipitation. The trend
towards sharply increased agricultural production
both in acreage and in productivity is likely to ac-
celerate pollution by use of marginal lands and even
greater amounts of chemicals. Control strategies
focus primarily on source control management, but
highly variable and complex conditions make con-
trol of this diverse situation very difficult.
Hazardous Material Spills
Hazardous materials production, storage, trans-
portation, or utilization in close proximity to a water
source provide a potential hazard to water quality.
Major pollutants from spills are toxic and other bio-
active chemicals, acids, alkalies, corrosives, oxygen-
demanding materials, and petroleum oils and related
products. The significance of the problem is mani-
fest, but its magnitude is difficult to assess. Control
requires detailed, comprehensive planning and action
response capabilities to minimize harmful effects.
Mining
All surface and underground mining activities,
including milling, benefication, and other site oper-
ations, are sources of pollution. Major pollutants
include sediments, acids, toxic metals, and other
salts. Increased demands for energy and all types of
minerals will result in increased impact of active
mining operations. Long-term problems will con-
tinue as our requirements continue and increase.
Control strategies, depending on the source, include
surface reclamation and specific collection and treat-
ment of discharges. Coal mine drainage treatment,
for instance, is costly but not unreasonably expen-
sive. An associated problem is pollution from aban-
doned mines. Nonoperating surface and underground
mines, including refuse and tailing piles, provide
significant pollutant amounts. Control is often dif-
ficult because specific sources cannot be identified.
Land Modification
All of man's activities which alter land surface
and thereby facilitate runoff (construction, land-
filling, etc.) produce significant amounts of pol-
lutants, primarily sediment carried by surface runoff.
The impact of these activities is a universal problem
related to overall area economic development. Sedi-
ment is both a physical pollutant and a carrier for
other pollutants. Source controls are the only ap-
parent promising operations for the control of runoff
from construction. Conditions vary widely, but they
must be seriously considered in the development of
any type of regional plan that expects to significantly
modify existing land use patterns.
IV. OPERATIONAL CONTROL
Wastewater treatment technology describes alter-
natives available to achieve a desired result through
its application to point sources. Non-point sources
may limit the potential for achieving all desired
water quality goals through point source control
alone. Other avenues for the achievement of the
121
-------
optimal water quality must be assessed, and these
fall into the area of operational control. Some of
these alternates and how they may be incorporated
into the overall plan are examined here.
Non-Point Source Alternates
Since there are no easily or rapidly achievable
means of overcoming non-point source pollution, it
seems most important to incorporate some basic
principles into the long-range objectives. Some of
these basic principles or concepts are:
1. the establishment of land use requirements to
minimize the concentration of industry and
other associated development;
2. modification of agricultural practices to mini-
mize runoff and the associated pollutional
problems of silt, fertilizers, pesticides, herbi-
cides, and organics; and
3. the control of area runoff (stormwater) which
can be as much of a pollutional load from
urbanized areas as the wastewater produced.
Land Use
The establishment of land use requirements may
well be the most difficult and certainly most com-
plex problem facing an areawide planner. An opti-
mum system for one purpose, water pollution con-
trol, is not adequate. Other environmental concerns,
such as air pollution control and solid waste dis-
posal, are important as well as considerations of
traffic flow, recreational and aesthetic requirements.
Other less obvious factors are the location of natural
resources, i.e., one must locate a mine where the
minerals are, transport the product to the market,
supply resources, and provide support functions.
The benefits of establishing a good land use plan
from a water quality standpoint are tremendous.
Minimizing the stress of pollutants on water re-
sources through selective locations will give this part
of the ecosystem a chance to stabilize, that is, reduce
the rate of degradation.
Agripractice
Because the choice of location for agricultural
activity is normally limited in any given basin by the
availability of suitable lands, the potential for opera-
tional controls becomes more important. Agricultural
non-point source water pollutants come in large part
from natural runoff. Pollution prevention must be
achieved through the management of agricultural
practices in cropping pasture, forestry and other
rural land uses. The basic problem, to manage for
proper operational changes to minimize pollution
without decreasing productivity, is further compli-
cated by such uncontrollable natural factors as
climate, topography, etc. Nevertheless, significant
progress can be made. Contour farming methods and
the proper application of fertilizers and other chem-
icals can materially reduce pollutants reaching the
water. As in all operational control areas manage-
ment models can effectively be used for characteriza-
tion and load reduction predictions. The degree of
control attainable will be somewhat less than that
for other sources because of the diverse nature of the
problem. Even so, the potential water quality gains,
especially in areas affected by toxics, eutrophication,
and silting, can be well worth the effort.
Area Runoff
Area runoff, water resulting from rainfall con-
taminated by contact with man's urbanized and sub-
urbanized activity, results in more diverse pollutants
than other non-point sources. Street runoff (con-
taining oil, motor vehicle exhaust chemicals), park
and lawn area runoff (fertilizers and toxic organics),
and construction site runoff (silt from disturbed
areas) will equal or exceed pollution from municipal
wastewaters. The volume of water can be extremely
high, depending on the location and degree of cover-
age (e.g., concrete).
Flows of this type collected in storm or combined
sewers present a significant point source of pollution
that is rarely effectively controlled. The only prac-
ticed technology on these large volume point sources
is brief settling and skimming. Lower flows or the
"first flush," can be routed to municipal or industrial
treatment plants.
Manufacturing facility runoff adds to the complex-
ity of the wastes, with material spills, and with run-
off and/or leaching of storage piles capable of con-
tributing extremely high pollution loads. Available
control methods include management of flow (e.g.,
dikes), routing of certain types of flows (e.g., roof
drains), and covering of refuse or storage piles, etc.
Combined sewer separation, and in developing
areas, the construction of separate storm drains and
sanitary sewers, is the best solution from a pollution
standpoint. The economics of separation in urban-
ized/industrialized areas may be prohibitive.
Effective operational control of non-point sources
will require time and good management. Some of
the management techniques will include long-range
planning, turning non-point sources into more con-
trollable point sources, and significant predictive and
management modeling.
Water Management
Optimum water management in urban areas re-
quires the serious consideration of water renovation
and reuse, particularly in water-short areas. The
supply of water is static geographically, unless di-
verted by man. Will man move to water or develop
122
-------
the techniques to supply water where he lives? Tradi-
tionally, man has remained relatively immobile. Pol-
lution patterns in 2070 will probably be similar to
today's. The total supply of water is about the same
as it was when the earth was very young and it will
be the same tomorrow!
The ultimate in recycle or "closing the loop" is
recycle of public wastewater directly back to the sup-
ply. Psychological factors inhibit this type of reuse
despite technical capabilities. Reuse of another city's
effluents by a downstream city doesn't seem to bother
us as much, though. In other words, the use of an-
other's treated sewage is acceptable, whereas the use
of one's own treated sewage is unacceptable. Ad-
vanced water purification technology destroys the
"logic" of this traditional attitude. A clear, tasteless,
odorless water can be produced. This "reuse reluct-
ance" has caused many cities to go to great expense
for alternative water sources, such as building large
reservoirs and transporting water hundreds of miles.
When faced with a decision on where to go for
increased water supply, the existence of these many
alternative schemes makes selection of the best one
extremely difficult. Many technical, economic, and
policy questions must be answered. Actual conditions
may cause many of the options to "drop out," mak-
ing the best alternative obvious by inspection.
V. NON-TECHNICAL
CONSIDERATIONS
A brief mention of some important non-technical
aspects of integrating water pollution control into
an environment plan is an appropriate way to con-
clude this portion of the overall assessment. Probably
one of the most frustrating problems may be termed
legal or jurisdictional. Conflicts will arise, for in-
stance, when one part of an optimum water system
must suffer due to the need of a transportation system
or vice versa. Bureaucratic battles can be predicted
when one local jurisdiction must, for instance, be-
come the repository for the waste materials from
another jurisdiction.
Depending on the existing form of government,
personal property rights, resources rights, water laws,
and industrial interests can serve to delay and change
the optimum plan.
Consequent legal challenges are as much of a
stumbling block to the orderly progression of a plan
as any technical problem.
VI, CONCLUSIONS
Technology is available to make great strides to-
ward improving water quality in a cost effective
manner through point source control. Methods for
the control of non-point sources provide less of an
opportunity to achieve control in the immediate
future and will require continued research and de-
velopment to develop and perfect control measures.
The longer range goal of zero discharge of pollutants
will be dependent upon the development of new and
more sophisticated pollution abatement techniques.
123
-------
Methods for the Control of Air Pollution
K.E.Yeager*
I. INTRODUCTION
The control of air pollutant impacts is a complex
problem because of the variety of sources and source
characteristics. Technical and economic factors fre-
quently make it necessary to use different control
procedures for different types of sources. Many
emission control techniques are still in the develop-
mental stage, and practical near-term control strat-
egies may call for the use of interim methods until
these techniques are perfected. In these situations,
alternatives, such as the use of design and siting
criteria which distribute rather than limit emissions,
may be required as a means for achieving air quality
goals. Therefore, fundamental to the development of
either Air Quality Management or widespread source
control strategies for the achievement of air quality
goals is an understanding of the performance, cost,
applicability and availability of potential control al-
ternatives. This paper will consider these basic
factors as they bear on both technical and opera-
tional emission control as well as emission dispersion
methods.
These considerations will begin with a discussion
of the air emission inventories associated with sta-
tionary and mobile source categories and their rela-
tionship to ambient air quality and proceed to discuss
the status of control methods for (1) sulfur oxides,
(2) total suspended particulate, (3) fine suspended
paniculate, (4) nitrogen oxides, (5) hydrocarbons
and oxidants, and (6) carbon monoxide.
n. CONTRIBUTION OF SOURCES TO
EMISSIONS
Table 1 summarizes the contribution of source
categories contributing at least one percent to the
inventory of six major air pollutants in the United
States. With the exception of fine suspended par-
ticulate, health related ambient air quality standards
have been established for each of these pollutants.
Fine suspended particulate (less than 3 microns) on
the other hand has been shown to be potentially the
most offensive particulate size fraction because it is
more easily respirable and more readily retained
in the lungs as well as having the ability to persist
for extended periods in the atmosphere.
Sulfur Oxides
Over 60 percent of the total sulfur oxides emis-
sions result from the combustion of coal, primarily
in large electric generating facilities. About two-
thirds of the nationwide emission occurs in urban
areas with power plants, industries, businesses and
homes all contributing to the burden. Much of the
pollution in rural areas is emitted by a relatively
small number of large individual sources. For ex-
ample, individual smelters, generally located away
from metropolitan areas, may emit over 100,000 tons
per year. In addition, almost half of the Nation's
power plants, including seven of the eight largest
sulphur oxide emitters, are located in rural areas. An-
nual emissions from these seven range from 200,000
to 300,000 tons per year.
Nitrogen Oxide
Motor vehicles and the statutory combustion of
fossil fuels account for about 90 percent of the na-
tional total. Nitrogen oxides and carbon monoxide,
probably more than other pollutants, are closely re-
lated to the distribution of population. It is not sur-
prising, therefore, that over 60 percent of nitrogen
oxide emissions occur in urban areas. In terms of
sources, only about 45 percent of the nitrogen oxide
emissions from motor vehicles are in urban areas
because, unlike carbon monoxide and hydrocarbons,
they are relatively unaffected by travel speed.
Carbon Monoxide
Petroleum-powered motor vehicles are the largest
single emitters of carbon monoxide. Although ve-
hicular travel is evenly divided between urban and
rural areas, traffic in urban areas, where slower
driving speeds prevail, accounts for 70 percent of
total motor vehicle carbon monoxide emissions.
•Chemical Engineer, Office of Research and Development, U.S. Environmental Protection Agency
125
-------
Table 1
1970 United States Air Pollutant Emission Contributions
by Source Category
(Percent)
Stationary Combustion
Electric Utilities
Indusrtial Combustion
Residential & Commercial
Pipeline Pumping
Motor Vehicles
Incineration
Industrial Processes
Carbon Black Manufacturing
Chemical Manufacturing
Grain Handling
Iron and Steel
Mineral Products
Nonferrous Metallurgy
Petroleum Refining
Pulp & Paper Manufacturing
Sulruric Acid Manufacturing
Solvent & Petroleum Product Evaporation
Natural Sources
Forest Fires
Controlled Burning
Natural Dust
Million Tons/Year Total Emissions (all categories)
Sulfur
Oxides
75
52
15
7
1
7
0
18
1
10
5
2
0
34
Nitrogen
Oxides
46
18
14
3
11
52
2
0
0
23
Carbon
Monoxide
0
65
8
10
3
4
2
1
17
8
9
149
Hydro-
Carbons
3
1
2
51
5
28
4
1
1
10
12
13
7
6
35
Total Fine
Suspended Suspended
Paniculate Paniculate
4 9
3 7
1 2
1 6
1 0
8 15
1
1 4
5 8
1
1 2
86 70
26 21
14 11
46 38
136 5
Hydrocarbons
Of the sources surveyed, in the absence of con-
trols, motor vehicles would again be the largest single
emitter category. There are fundamental difficulties
involved, however, in simultaneously eliminating
hydrocarbon, carbon monoxide and nitrogen oxide
emissions from mobile sources. Specifically, the car-
bon monoxide and hydrocarbon emissions could be
reduced by increasing the ratio of air to fuel to the
point where more air is present than is required for
complete combustion of the fuel to carbon dioxide
and water. Maximum emissions of nitrogen oxides,
however, would occur under such conditions. On
the other hand, at very low air-fuel ratios, the nitro-
gen oxide emissions could be reduced but high con-
centrations of carbon monoxide and hydrocarbons
would be produced. At the extremely high air-fuel
ratios where all three emissions could theoretically
be low, operating difficulties such as misfire and
stalling would be encountered with most commer-
cially available internal combustion engines. This
would result in poor performance and high emissions
of carbon monoxide and hydrocarbons.
Total Suspended Particulates
As indicated in Table 1, the largest sources of
particulates are natural dusts and forest fires. These
sources account for an estimated 86 percent of the
national atmospheric primary particulate loading on
a mass basis, and are a substantial portion of back-
ground levels. Emissions from these sources are es-
sentially beyond the scope of present air pollution
control methodology. These natural emissions differ
from that found in urban industrialized areas, in that
a generally smaller percentage of particulates are in
the respirable range (less than 3 microns). Similar
fugitive or unconfined sources directly related to
man's activity involve agriculture, mining, construc-
tion and transportation.
The mass of emissions from most of the industrial
sources may be highly controlled by the installation
of presently available control equipment. In some
cases, such as combustion of coal by electric utilities,
over 90 percent collection efficiency is routinely
achieved. Nevertheless, because of the large amounts
of coal burned and its inherently large noncombust-
ible particulate content, emissions from coal com-
bustion remain a major portion of the total par-
ticulate emissions from man's activity.
FINE PARTICULATES
The overall impact of fine particulate pollutants,
broadly denned as particles less than 3 microns in
size, on man's environment is not well known. How-
ever, evidence continues to point to the many nega-
tive aspects of fine particulate pollution. Research
on health effects of air pollution indicates links be-
tween fine particulate pollution and health effects of
varying severity. In addition to particle size, the
126
-------
chemical composition of participates is an important
factor in determining the effects of this type of air
pollution. For example, the health hazard associated
with inhaled airborne particles depends on: (1) the
site of deposition in the respiratory tract, which is
determined by particle sizle; and (2) the effect on
biological tissues at the deposition site, which de-
pends on chemical composition. The hazardous pol-
lutant problem (e.g., mercury, lead, cadmium,
vanadium) is also directly linked to fine particle
pollution because many of the industrial processes
that emit hazardous pollutants liberate them in the
form of micron or submicron particles. On both a
regional and global scale, fine participate pollutants
play a principal role in the transport through the
air of a variety of hazardous substances. It is empha-
sized that the hazards posed to human health by trace
amounts of toxic materials in the form of fine par-
ticulates can be disproportionate to the mass in-
volved.
Fine paniculate air pollutants may be classified
into two major classes based on their origin. These
are (1) primary fine particulates which are emitted
or immediately condensed as fine particulates from
a specific source, and (2) secondary fine particulates
which are the products of atmospheric reactions to
form particulate air pollutants.
III. LOCAL EMISSION / AMBIENT AIR
QUALITY RELATIONSHIPS
The major concern with source emissions is their
impact on maximum ground level pollutant concen-
trations in the vicinity of the site. Several meteoro-
logical diffusion models exist which can be used to
provide satisfactory engineering estimates of sulfur
oxide and total suspended particulate concentration
on any given time reference of concern. The results
of these analyses can be used to define control and
siting criteria based on the actual contribution of
each source to ambient air quality.
Figure 1 summarizes this typical variation in con-
tribution to ground level ambient sulfur oxide con-
centration for four emitter categories. For example,
while utilities typically contribute over 50 percent of
all sulfur oxide emission in a local region, well-
known factors such as stack height may limit the
corresponding impact on the ground level sulfur
oxide concentrations within a few kilometers of the
emitter to less than 30 percent of the total. Con-
versely, area and smaller industrial combustion
sources typically contribute more heavily to the
ground level ambient concentrations than would be
inferred by their percentage of total regional emis-
sions.
This ability to relate emissions to ambient air
quality can be a very powerful and cost-effective tool
for (1) determining the level of control required by
source to achieve a desired ambient air quality goal,
(2) allocating limited emission control capabilities
to those sources having the greatest impact on am-
bient air quality and (3) defining new facility siting
options compatible with achieving or maintaining
ambient air quality goals.
Unfortunately, this technique has several limita-
tions. First, it requires an accurate source-by-source
emission inventory and regional meteorological data
base. Second, considerable additional research is re-
quired before these meteorological diffusion models
can accurately include the effect of atmospheric
transformation primarily affecting hydrocarbons,
oxides of nitrogen, oxidants and sulfates. Third, it
may be concluded from the emission/ambient air
quality relationships indicated in Figure 1 that peak
ground level concentrations of sulfur oxides and total
suspended particulates may be reduced to levels con-
sistent with ambient air quality goals thorugh the
relatively simple expedient of increasing effective
stack height. Indeed, such considerations have
formed the basis for "supplementary" control
methods designed to achieve air quality goals in the
absence of sufficient low sulfur fuel or control tech-
nology. This should be considered only as a near-
term tactical expedient however. For example, our
current knowledge of the significant health effect
thresholds for sulfates and related fine suspended par-
ticulate matter infers the possible need for ground
level ambient concentrations that may be eight to
ten times lower than current practice. Under these
conditions, the use of tall stacks is inadequate for
achieving the necessary ground level ambient condi-
tions and indeed results only in subjecting a larger
ground area to excessive concentrations. This in-
adequacy is further accentuated in the long term as
a result of a continued growth of society and its
associated energy needs. Unless improved control
techniques continue to be developed and applied, by
the year 2000 it is forecast that SOx and fine sus-
pended particulate emissions from industrial activities
will increase fivefold. These emissions will not only
dominate regional conditions, but could produce
global concentrations exceeding acceptable health
thresholds.
IV. STATUS OF CONTROL METHODS
Sulfur Oxide Control
Beginning in 1966, both industry and government
started working seriously to develop processes cap-
able of removing sulfur oxides from fossil fuel com-
bustion flue gases. Because of the operating com-
plexity and cost associated with these essentially
chemical processing facilities, their practical appli-
cation at the present state-of-the-art is limited to
127
-------
K)
00
111
U.
1L
Ul
>
z
5
8
ATMOSPHERIC
ELECTRICTY
ATMOSPHERIC
VISIBILITY
CONDENSATION NUCLEI
FOR PRECIPITATION
SOILING PHENOMENA
(HORIZONTAL SURFACES)
UPPER RESPIRATORY
TRACT DEPOSITION IN MAN
PERIPHERAL AIRWAYS AND
ALVEOLAR DEPOSITION-MAN
SOILING PHENOMENA
(VERTICAL SURFACES)
ATMOSPHERIC
CHEMISTRY (GAS-SOLID)
W///////////
10-4
10-3
10-2 10-1 10° 101
PARTICLE DIAMETER -J» m
102
SOURCE: CONTROL TECHNIQUES FOR NITROGEN OXIDES FROM STATIONARY SOURCES,
AP-67, NATIONAL AIR POLLUTION CONTROL ADMINISTRATION, MARCH 1970.
FIGURE 1. EFFECTS OF PARTICULATE AIR POLLUTION IN THE
COMMUNITY AS RELATED TO PARTICLE SIZE
-------
large industrial and utility steam generators (greater
than 100,000 Ibs. steam/hr—equivalent to about 10
MW). As a result of this activity, about a dozen
processes have been developed to the point where
they are being studied on full-scale industrial facil-
ities. Nine of these processes are variations of the
slurry scrubbing concept using ash, lime, limestone,
or magnesium oxide as the reactant. Development
results to date generally indicate a sulfur dioxide re-
moval capability of 80-90 percent. As of January
1974, 44 commercial and demonstration flue gas
desulfurization installations were under construction
or on order for use on existing and new electric
utility steam generators in the United States. These
installations encompass 18,000 MW of generating
capacity at an average contract cost of $60 per KW
of controlled capacity. This represents 9 percent of
the coal and oil fired electrical generating capacity in
the United States.
The most serious inherent limitation in the broad
commercial application of these processes is disposal
of the unmarketable waste gypsum by-product. Aside
from the immediate spatial problem associated with
the storage and handling of this waste product, dis-
posal must be performed in a manner which does not
cause water pollution.
The immediate question is the extent of com-
mercial application which can be achieved by effluent
treatment technology and the rate at which this ap-
plication can be achieved. The extent of application
is essentially a question of economics. Based on field
surveys, it is theoretically possible to retrofit as much
as 80 percent of all existing utility and large in-
dustrial coal and residual oil-fired combustion
sources with effluent treatment processes for the re-
moval of sulfur dioxides. In reality, however, con-
siderations such as process packaging within avail-
able space, by-product transport and disposal, re-
actant availability, plant capacity use factor, gas, and
size will limit the degree of application. For example,
the range of annualized SOX control process capital
and operating costs (incremental costs of control)
required to satisfy 80 percent of existing utility and
industrial combustion facilities extends from 100/M
BTU of fuel consumed to over $1.00/M BTU. A
practical upper bound on the ultimate extent of flue
gas control application would be about 40 percent of
large utility and industrial coal and residual oil
capacity. This level of application would imply an
incremental effluent control cost on the order of
600/M BTU.
It is clear that the combination of natural low
sulfur fuels and flue gas desulfurization technology
will not provide the quantity and size range of com-
bustion capacity required to achieve existing national
ambient air quality standards. Thus, a third category
of control technology, fuel desulfurization, must also
be considered. These processes for not only desulfuri-
zation, but total paniculate cleanup of coal and
residual oil are required to achieve ambient air qual-
ity standards. Their basic role is twofold:
(1) Provide control technology for combustion
sources to which other alternatives are not technically
or economically applicable;
(2) Provide synthetic, low pollution fuel aug-
menting the limited domestic supply of natural gas.
These approaches range from the application of
currently commercialized mechanical coal cleaning
technology, through chemical coal and oil desulfuri-
zation technology which maintains the clean fuel in
its original physical state, to coal conversion tech-
nology for the synthesis of low- and high-BTU gas.
The development cycle for these latter efforts, how-
ever, restricts commercial application until 1978-
1985.
Given the current limitations in the availability of
low sulfur fuel and control technology, the near term
achievement of sulfur oxide ambient air quality stand-
ards can be achieved by application of the previously
discussed ability to predict emission/ambient air
quality on a source-by-source basis. This under-
standing can significantly reduce the number of
sources requiring control. A long term benefit of such
an approach is the ability to limit the very expensive
implementation of sulfur oxide control technology
to a few critical sources until a better appreciation
of the technical implication of sulfate control can be
developed. The effects of these considerations has
been examined in three separate urban regions in
which sufficient data were available.
For example, analysis of the 10 worst receptors in
the Philadelphia region indicated an average annual
sulfur oxide concentration ranging from 166 to 255
micrograms per meter3. As shown in Table 2, if
controls could be selectively applied in the Phila-
delphia region, it would be necessary to control, at a
75-percent effectiveness level, only 53 of the emitters
encompassing 19 percent of the total regional emis-
sions to achieve the existing sulfur oxide air quality
standard.
Extrapolation of the trends indicated in the urban
regions examined suggests that the existing sulfur
oxide air quality standard could be achieved in the
eastern United States by the selective control of a set
of sources using 30 percent or less of all coal and
residual oil consumed. This extrapolation has been
supported by relating actual air quality measurements
to source fuel modifications.
Fine Particulate Control Technology
Four major categories of industrial paniculate con-
trol equipment (electrostatic percipitators, fabric
filters, scrubbers and inertial or centrifugal mechan-
ical collectors) each having special characteristics
that determine the suitability for a particular gas
cleaning application, have been available for many
129
-------
Table 2
Selective Reduction of SOX Emission to Achieve 80 /ig/m3—Philadelphia
Emitter Category
Industrial Combustion
Industrial Processes
Utility Power
Area Sources
TOTALS
Present Emissions
Sources Tons/Day
238 555
165 300
21 1345
276 4378
700 6578
Number of Sources Requiring Control at:
100%
Sources Tons/Day
3 172
5 41
6 519
11 148
25 880
75%
Sources Tons/Day
10 293
13 124
9 578
21 222
53 1217
years. This presently available control equipment
has achieved only limited effectiveness in fine par-
ticulate collection capability and i$ generally not ap-
plicable because of material limitations in situations
where co-contaminants must be dealt with.
Efficiency data indicate that the only conventional
control devices capable of significant collection of
fine particulates from industrial sources are high-
efficiency electrostatic precipitators, venturi scrub-
bers and fabric filters. Existing data, as well as
theory, on fine paniculate collection efficiency (both
on a mass and number basis) of conventional equip-
ment indicate that inefficiency in the fine particulate
range is an inherent characteristic of the equipment.
Therefore, definite constraints will be imposed on
the types of emission standards for fine particulates
that can be implemented in the near term. These
inherent limitations may require that process modi-
fications or other alternatives be used to meet near-
term regulations.
These are three basic avenues that might lead to
improved control of fine particulates: (1) droplet
condensation and impaction; (2) augmentation of
commonly used collection mechanisms by additional
forces such as thermal, diffusiophoretic and electrical
which do not approach zero in the less than 3 micron
size range; and (3) utilization of particle condition-
ing or agglomeration techniques producing particles
compatible with conventional collection devices.
Substantial improvements in fine particulate emis-
sion could be achieved using currently available con-
trol equipment as indicated in Table 3. This table
also indicates the relationship of emission level to
plume opacity. Table 4 indicates that in most cases,
installation of this best available technology will have
relatively minor effects on overall production costs,
generally amounting to less than one percent of the
value of the product produced.
The collection of fiine particulate emitted directly
from combustion or other high temperature processes
Table 3
Estimates of Fine Particle Emissions as a Function of Emission Standard *
Source
I. Coal combustion
(I. Iron and steel
A. Sinter machines
B. Basic oxygen furnace
C. Electric arc furnace
III. Cement plants, rotary kilns
[V. Asphalt plants, dryers
V. Ferroalloy plants
A. Closed electric furnace
B. Hooded open electric
furnace
C. Unhooded open electric
furnace
VI. Lime plants, rotary kilns
VII. Municipal incinerators
VIII. Iron foundry, cupola
Estimated Fine Particle Emissions
Uncontrolled
Lb/Hr Ton/Yr
796,000 3,184,000
7,100 28,400
400,000 1,600,000
27.400 109,600
243,000 972,000
3,260,000 1,956.000
92,000 368,000
129,500 518,000
72,000 288,000
75,800 303,200
15,000 37,500
22,900 22,900
Present
Lb/Hr Ton/Yr
243,000 972,000
1,400 5.600
43,600 174,400
3,600 14,400
44,300 177,200
257,000 154,200
11,300 45,200
84,100 336,400
60,800 243,200
22,000 88,000
13,100 32,800
13,100 13,100
BICD'
Standard
Lb/Hr Ton/Yf
15,000 60,000
64 256
2,560 10,240
645 2,589
2.000 8,000
25,600 15,350
11,300 45,200
15,800 63,200
28,800 115,200
1,200 4,800
490 1,225
356 356
10% Opacity
Standard
Lb/Hr Ton/Yr
5,050 20,200
722 2,888
1,540 6,160
1,036 4,144.
25,200 100,800
200,000 120,000
NCk NC
NC NC
NC NC
10,500 42,000
3,400 8,500
5,900 5,900
5% Opacity
Standard
Lb/Hr Ton/Yr
2,510 10,040
354 1,416
800 3,200
683 2,732
12.200 48,800
167,000 100,200
NC NC
NC NC
NC NC
4,800 19,200
1,900 4,750
4,000 4,000
BICD—Best installed
NC—Not calculated.
controls (not neccuarily the highest efficiency device available, but rather the beft that is generally being installed at
130
-------
Table 4
Summary of Fine Particulate Control Costs ut BICD-Level Control
($/Unit of Production)—at Capital Charge Rate of 0.20
Source
Coal-fired electric plant
Municipal incinerator
Cement plant (rotary kiln)
Asphalt plant (rotary dryers)
Iron and steel
Basic oxygen furnace
Mine plant (rotary klin)
Annual Production
Rate for Model
Plant
2.4 X 10" kwh
80,000 tons
3.0 X 10° bbls
90,000 tons
1.0 X 10' tons
87,500 tons
Fine Particle
Control
Efficiency
98.13
96.71
99.17
99.21
99.70
98.43
Unit Cost
or Value
$
0.015
10.00
4.32
23.50
187.00
18.00
Control Cost/
Unit of
Production
0.00012
0.550
0.049
0.177
0.259
0.240
is only one portion of the fine paniculate problem.
Atmospheric fine particles which contain sulfate,
nitrate and organic aerosols are largely formed dur-
ing secondary atmospheric reactions involving the
controlled pollutants subject to ambient air standards.
Conversion of sulfur dioxide into sulfate is con-
sidered to be responsible for one-half to three-fourths
of the atmospheric particles resulting from man's
activities. The conversion of nitrogen oxide to nitrate
particles accounts for 7 to 14 percent of the par-
ticulate burden produced by man, and the conversion
of hydrocarbons has been estimated to account for
much of the remainder. The relationships are under-
stood in a general way, but they have not yet been
placed on a quantitative basis. The nature of such
formation may be through catalytic oxidation with
heavy metal ions, ammonia-sulfate dioxide reaction
in the presence of water, photochemical oxidation,
and on heterogeneous gas reactions. As a result, it
is anticipated that the control of this secondary fine
paniculate fraction may involve further control of
direct links in the chain of precursors or modification
to the physical and chemical parameters ultimately
affecting formation. Table 5 summarizes current best
judgment as to the short- and long-term thresholds
for paniculate sulfate as compared to total suspended
paniculate and sulfur dioxide. These data indicate
that paniculate sulfate has a threshold health effect
an order of magnitude less than sulfur oxide and total
suspended paniculate.
The capabilities of current sulfur oxide control
technologies are targeted at achieving moderate (30-
70 percent) reductions in sulfur dioxide concentra-
tion in urban areas. Unless other controllable links
in the precursor chain can be defined, the sulfate
levels in Table 5 imply the need for a high efficiency
control technology (greater than 95 percent) on
large industrial sources and increased use of clean
fuels in smaller combustion sources. Similarly, nitrate
control may require improvements well beyond the
current state-of-the-art in nitrogen oxide control
technology.
Mobile sources contribute only about 3 percent
of the total paniculate. However, studies on the size
distribution of automotive exhaust indicate that 90
percent or more of the mass emitted consists of par-
ticles smaller than 1 micron. Estimated emission rates
for gasoline-powered vehicles burning fuel containing
usual quantities of tetraethyl lead (2.5 gm.Pb/gal.)
are on the order of 0.08 to 0.25 grams/mile. Lead
free gasoline will be required in cars equipped with
hydrocarbon control catalysts since lead emissions
quickly deactivate catalysts. With these lead-free
fuels and catalysts, paniculate emissions will be in
the 0.02 to 0.05 gram/mile range and may be dom-
inated by sulfates and nitrates.
Emissions from aircraft turbines are significant in
some areas. Estimates of Los Angeles turbine par-
ticulate emissions are one-third to one-fifth of auto-
mobile-generated particulates. Near approach and
take-off corridors, aircraft particulates dominate the
atmospheric aerosol. Particulate emissions may cover
a wide range from 0.1 to 2 percent of the fuel
burned, depending on engine type and operating
mode.
Nitrogen Oxide Control Technology
The control of nitrogen oxides, as indicated earlier,
must consider the impact of both stationary and
mobile combustion sources. The combustion of fossil
fuels in air results in the formation of oxides of
nitrogen both as a result of chemically bound nitro-
gen in the fuel and high temperature fixation of
atmospheric nitrogen. In high temperature combus-
tion processes, atmospheric nitrogen and oxygen play
a dominant role in nitrogen oxide formation. Ther-
modynamic equilibrium highly favors the formation
of NO over NO2. Once formed at high temperatures,
the NO can react with excess oxygen to yield NO2.
However, the residence time usually available in
combustion equipment is too short for more than
5-10 percent of the NO to be oxidized to NO2.
In principle, nitrogen oxide emissions from sta-
tionary sources can be controlled by stack gas clean-
ing methods or by combustion modification. Un-
fortunately, however, flue gas treatment methods are
131
-------
Table 5
($/Unit of Production)—at Capital Charge Rate of 0.20
(SHORT TERM)
Effect
Mortality Harvest
Aggravation of Symptoms in Elderly
Aggravation of Asthma
Acute Irritation Symptoms
Present Standard
24-Hour Threshold, wg/m '
Sulfur
Dioxide
300 TO 400
365
180 TO 250
340
365
Total Suspended
Participates
250 TO 300
80 TO 100
100
170
260
Partic-
ulate
Sulfate*
No data
8 to 10
8 to 10
No data
No
standard
(LONG TERM)
Effect
Decreased Lung Function of Children
Increased Acute Lower Respiratory Disease
in Families
Increased Prevalence of Chronic Bronchitis
Present Standard
Annual Threshold, ttg/m •
Sulfur
Dioxide
200
90 TO 100
95
80
Total Suspended
Particulates
100
80 TO 100
100
75
(Geometric)
Partic-
ulate
Sulfate*
11
9
14
No
standard
• There is a growing body of evidence that adverse health effects are related to some form of sulfates. However, inadequate data on con-
centrations and compound-specific health effects make highly uncertain even a "best judgment" estimate of a threshold.
faced with major technological and economic ob-
stacles and at present, no treatment process, either
catalytic or scrubbing, appears viable. Thus, combus-
tion modification is the only demonstrated method
for the control of nitrogen oxide emissions from sta-
tionary fossil fuel combustion.
Control technology developments to date have
demonstrated that flue gas recirculation is the most
effective combustion control technique for nitrogen
oxide emissions from nitrogen fixation and that
staged combustion is the most effective technique for
controlling fuel nitrogen conversion. The application
of this technology has controlled the nitrogen oxide
emissions from gas and oil fired utility boilers by
more than 50 percent to a level of 150-250 ppm.
Reduction in emissions of up to 80 percent can be
achieved by the combination of reducing excess air
from 50 to 20 percent and recirculating 50 percent
of volume of the flue gas. The mechanism for the
reduction of nitrogen oxide emissions via flue gas
recirculation is primarily a result of decreasing flame
temperature and dilution of the available oxygen in
the flame zone. Unfortunately, in large installations
the amount of flue gas recirculation desired poses
major redesign problems for existing gas handling
equipment and ducting.
In limited short-term testing, the above combus-
tion modifications, combined with burner designs
which prevent burner flame interaction, have also
resulted in nitrogen oxide reductions of up to 50 per-
cent in coal fired utility boilers. Experimental burner
designs have resulted in further reductions of up to
90 percent. These burner design modifications are
extremely attractive because they are very inexpen-
sive and have little or no operating cost penalty.
In smaller households, commercial and industrial
combustion sources, although the same design and
operating principles apply, very little experimental
data are available on the control of nitrogen oxide
emissions from such sources. On a more long-term
basis, work has been initiated on advanced combus-
tion modification concepts, such as catalytic burning
of natural gas and coal-derived synthetic fuels at
temperatures near 1400°C. Ultimately, utilization
of such new concepts will allow essentially pollution
free energy conversion from fossil fuels.
Low excess air firing of new and existing units
costs about S0.82/KW depending upon unit size, age
and design. This amounts, however, to less than one
percent of the average $225/KW cost of a new coal
fired generating plant.
Mobile Source Control of Nitrogen Oxides,
Hydrocarbons and Carbon Monoxide
The standards for these motor vehicle related pol-
lutants have been exceeded in a number of major
urban areas. In the United States, urban areas repre-
senting nearly 55 percent of the Nation's population
132
-------
exceed the ambient air quality standards for one or
more of these pollutants. Table 6 indicates the gen-
eral range of relative contributions of emission
sources in our urban areas.
Table 6
Mix of Emission Sources in the
Urban Areas—1971
Table 7
Cost Effectiveness of Hydrocarbon
Control Techniques41
Pollutant
CO
HC
NOX
Percent of Total Emissions
Automobile
77-87
50-65
40-50
Trucks, Buses
& Motorcycles
8-10
5-10
8-13
Stationary
Sources
3-15
25-45
37-52
Because of the fundamental problems discussed
earlier in simultaneously controlling nitrogen oxides,
hydrocarbons and carbon monoxide from current
designs of moter vehicle internal combustion engines,
the strategy in the United States depends on control
of stationary sources to achieve nitrogen oxide air
quality standards. Mobile source controls on the
other hand will be applied to achieve hydrocarbon
(oxidants) and carbon monoxide air quality stand-
ards. These controls may take the form of (1) emis-
sion controls on new and in-use vehicles, (2) the
reduction of vehicle miles traveled through the use
of traffic controls, mass transit, parking taxes, etc.,
and (3) the development and implementation of new
power systems having improved emission character-
istics.
In addition, uncontrolled hydrocarbon vapor
losses from service stations will make this source as
important a contributor to hydrocarbon emissions
as the vehicles they serve. Translated into grams/
mile, the hydrocarbon emissions from service stations
exceed the 1976 new car hydrocarbon standards.
Vapor displacement control techniques currently
being developed can partially reduce these emissions
by over 75 percent by 1977.
1. Emission Controls
The cost effectiveness of these control techniques
in terms of pounds of pollutant controlled per dollar
expended on control are described in Tables 7 and
8 for hydrocarbon and carbon monoxide, respec-
tively. The data in Table 7 show that control tech-
niques designed to reduce hydrocarbon emissions
from pre-controlled vehicles and service stations are
clearly the most cost-effective. However, these are
not necessarily the most important in eliminating a
regional air pollution problem. The figures in paren-
theses in Table 7 show the reductions in total hydro-
carbon emissions from the implementation of each
control technique.1 These data show that the use of
pre-controlled vehicle retrofits in 1977 (Air Bleed or
Inspection/Maintenance
(all cars)
Catalyst (1972-1974)
Catalyst (1968-1974)
Air Bleed (pre-1968)
Vacuum Spark Advance
Disconnect (pre-1968)
Gasoline Station Controls
Date of
Program Implementation**
1977
1.64 (3.6 %)
1.79 (3.77%)
1.90 (5.5 %)
5.48 (1.0 %)
6.91 (1.25%)
10.67 (7.1 %)
1980
0.54 (1.5%)
1.31 (3.0%)
1.23 (3.6%)
5.20 (0.5%)
8.36 (0.6%)
11.56 (9.6%)
•Cost-effectiveness increases with the size of the figure.
••The cost-effectiveness of the 1976 hydrocarbon standard is 3.7.
Vacuum Spark Advance Disconnect) will reduce
total hydrocarbon emissions in the air quality region
by only approximately 1 percent. Therefore, the cost-
effectiveness of pre-control vehicle retrofits is very
high, but they achieve relatively little in the way of
improvement in air quality. Conversely, measures
such as inspection/maintenance and catalytic retro-
fits, which are applicable to controlled vehicles, have
relatively low cost-effectiveness but higher air quality
impact. Unfortunately, as indicated earlier these
catalytic devices may have serious and unacceptable
impact on sulfate and on fine particulates. Gasoline
marketing controls prove to be both important and
cost-effective in eliminating regional hydrocarbon
emissions.
Finally, it should be noted that inspection/main-
tenance is a prerequisite for all retrofit measures
owing to the need to keep the retrofit devices in good
operating condition. Hence, it is not possible in
practice to select an approach such as Vacuum Spark
Advance Disconnect, for example, based on its high
cost-effectiveness and implement it without inspec-
tion/maintenance.
The cost-effectiveness calculations for carbon
monoxide controls follow a progression similar to
that for hydrocarbon controls. Retrofit devices for
pre-control vehicles again prove to be the most cost-
effective but measures for controlled vehicles have
the greatest impact on air quality. However, the
carbon monoxide emission reductions (shown in pa-
rentheses) achieved by those control techniques are
considerably higher than those achieved for hydro-
carbon. Air bleed retrofits, for example, yield a 7.1
percent reduction in total carbon monoxide emissions
in 1977, while air bleed only yielded a 1 percent re-
duction for hydrocarbons. The relative importance
of carbon monoxide control techniques is generally
higher because regional carbon monoxide problems
are caused primarily by motor vehicles. Again, in-
spection/maintenance is a prerequisite for retrofit.
» Calculations based on a region with 30 percent stationary lource hydrocarbon contribution such as Philadelphia, Baltimore, and Indianapolis.
133
-------
Table 8
Cost-Effectveness of Carbon Monoxide
Techniques'"
Inspection/Maintenance
Catalyst (1972-1974)
Catalyst (1968-1974)
Vacuum Spark Advance
Disconnect (pre-1968)
Air Bleed (pre-1968)
Date of
Program Implementation**
1977
17.5 ( 7.7%)
20.0 ( 8.4%)
24.7 (14.5%)
31.8 ( 1.1%)
194.0 ( 7.1%)
1980
8.8 ( 6.3%)
15.2 ( 9.2%)
15.5 (11.8%)
40.9 ( 0.7%)
192.7 ( 4.9%)
•Based on a region with 5 percent stationary source contribution such
as Seattle, Phoenix and Minneapolis, .
« "The cost-effectiveness of the 1976 new car carbon monoxide is 32.9.
2. Vehicle Mileage Reduction
Increasing public awareness of the adverse effects
of the automobile on the urban environment, and
legislation such as the Clean Air Act that has re-
sulted from this awareness, make it clear that urban
development policies that have encouraged and relied
upon unrestricted use of the automobile must be
changed. Controls must be placed on automobile
use; transit must be found to prevent future urban
growth from generating large volumes of traffic. The
need to reduce urban area auto use is no longer at
issue. The problem is how to do so without exces-
sively restricting the mobility of urban area residents.
Among the possible approaches to the solution of
this problem, increased transit usage and quality
seem to offer the greatest potential for success over
the long run. Other possible approaches include in-
creased carpooling, reducing trip frequencies or trip
lengths, and direct vehicular restraints (e.g., vehicle
free zones).
Planning and implementing substantial transit im-
provements are likely to present severe problems in
the period 1973-77 owing to the difficulties of de-
signing suitable transit systems and acquiring the
necessary vehicles. Existing transit does not have
the capacity to achieve large reductions in auto use.
Most urban transit systems operate at more than 75
percent of capacity during periods of peak work
travel. EPA calculations indicate that with this level
of capacity usage, the maximum reduction in auto
use that can be achieved by existing transit fleets is
about 5 percent. Achieving a 10 to 20 percent re-
duction in auto use could require expansions of cur-
rent transit fleets of at least 50 percent and possibly
over 300 percent. Threefold fleet expansions in many
urban areas could exceed the short run production
capacity of the bus manufacturing industry. In addi-
tion to bus production problems, there appear to be
significant short-run planning problems. Most exist-
ing urban area transit plans are projected to achieve
decreases in auto use of less than 10 percent. The
most ambitious transit plan that has come to EPA's
attention, that for the Washington Metro System, is
projected to be capable of achieving a 20-percent re-
duction in auto use if it were fully implemented in
1976. In fact, the system will not be completed until
1983. These production and planning problems sug-
gest that although the potential for reducing auto use
through improved transit is large, it may be unreal-
istic to expect reductions greater than 10 to 20 per-
cent by 1977.
The cost of bus transit depends on the detailed
characteristics of the bus system, notably on vehicle
occupancies. Transit buses cost roughly $1.00 per
mile to operate compared with $0.07 per mile for
cars. Hence, a transit system that carries roughly 40
riders per vehicle round trip will cost about the same
as the auto. Higher occupancy systems might achieve
net savings of $100 per rider per year. However,
with low occupancies cost could reach $900 per com-
muter per year. There is clearly a potential for
achieving substantial emission reductions at a net
cost savings through increased use of bus transit.
However, precise cost estimates will not be possible
until detailed plans for emission-control oriented
transit systems have been developed.
Car pooling programs appear capable of achieving
net cost savings. A car pool program for the Wash-
ington, D.C. area based on a locator system and in-
creased parking fees has been estimated to require
an initial investment of $1.3 million and to have
operating costs of $0.6 million year. If this system
achieves a three percent increase in auto occupancies
for peak period downtown work trips, the savings it
achieves in auto operating costs will equal the an-
nualized costs of the system.
3. Mobile Source Control Effectiveness
The projections of emissions reflect reductions in
emissions due to in-use vehicle controls; Federal new
vehicle emission standards for automobiles, trucks,
and buses; and Federal new source performance
standards for certain categories of stationary sources.
They also reflect substantial reductions in emissions
from new and existing stationary sources due to
present and planned State and local regulations.
Figure 2 gives a comparison of hydrocarbon and
carbon monoxide ambient air concentrations now
and that expected in 1977 based on the cumulative
control capability of (1) inspection and maintenance,
(2) control devices on new and existing vehicles,
(3) 20 percent reduction in auto use, and (4) ap-
plications of best available hydrocarbon control tech-
nology to stationary sources.
V. CONCLUSIONS
The preceding discussion of methods for the con-
trol of air pollution leads to the following conclusions
concerning the application of these methods to con-
trol strategies.
134
-------
NATIONAL
STANDARDS
CO
Z
o
CO
g
5
Ul
cc
o
a.
O
o.
120
100
80
60
40
20
OXIDANTS
1970-72
CUMULATIVE POPULATION
.08 .16 .24 .32 .40 .48 .56 .64
MAXIMUM CONCENTRATION (PPM)
O
CO
O
O
UJ
CC
u.
o
o
100
80
60
40
20
0
NATIONAL
STANDARDS
CARBON MONOXIDE
1970-72
CUMULATIVE POPULATION
18 27 36
MAXIMUM CONCENTRATION (PPM)
FIGURE 2. AMBIENT AIR CONCENTRATIONS
45
135
-------
1. The achievement of sulfur oxide ambient air
quality improvements is fundamentally dependent on
the control of fossil fuel combustion, primarily coal
and residual oil. In the near term (1975-1985), the
availability of low sulfur content fuel and control
technology limits reductions in urban sulfur oxide
air quality to levels approximating current standards
(60-80 micrograms per meter3 on an annual basis).
Even the achievement of this improvement must de-
pend on the allocation of limited low sulfur fuel and
controls to those sources imposing most heavily on
ground level sulfur dioxide concentrations on a
limited regional basis.
The potentially more serious problem of sulfate
control may require widespread application of much
more efficient sulfur oxide emission control methods
than those currently being developed and com-
mercially introduced. The apparently very low, safe
ambient concentrations for this pollutant imply that
plume dispersion considerations will not be appro-
priate for limiting source control requirements. As a
result, until the interrelated questions concerning (a)
what sulfate compounds impact seriously on health;
(b) what are the atmospheric transport and trans-
formation phenomena which produce these com-
pounds; and (c) what precursors must be controlled
to what levels are answered it may be very counter-
productive to force the widespread installation of
sulfur oxide control technology fundamentally inade-
quate to solve this problem at costs which often ex-
ceed 25 percent of the annualized cost of the con-
trolled facility.
2. The widespread commercial installation and
proper maintenance of relatively low cost existing
particulate control technology can reduce the emis-
sion of fine particulate (less than 3 micron) from
industrial processes and large combustion sources by
95 percent relative to current practice. This capabil-
ity can be achieved at an increase of about 1 percent
to 5 percent in the annualized cost of the controlled
facility. The achievement of ambient air concentra-
tions complying with probable future health stand-
ards for trace metals and carcinogens will, however,
require further improvement in controls particularly
for combustion sources, either through fuel pretreat-
ment or flue gas conditioning. In either case, the
technologies and energy requirements associated with
further improvement imply a tenfold increase in
control cost for every one percent further reduction
in fine particulate emissions. Further, these improved
capabilities are at early stages of development and
will not be commercially available until at least 1980.
3. Current nitrogen oxide ambient air concentra-
tions in the United States are generally consistent
with current health based standards. As a result,
emphasis is placed on a strategy of maintaining these
levels. Combustion modification techniques capable
of reducing nitrogen oxide emissions by at least 50
percent from all classes of stationary combustion
sources appear to be either commercially available
or in an advanced state of development. These tech-
niques appear capable of at least maintaining current
ambient air quality levels for the foreseeable future
if adequate incentives for commercial implementa-
tion are applied. The costs of these combustion modi-
fication techniques are about one percent of the an-
nualized cost of industrial and utility combustion
facilities, but may be as much as 10 percent of the
cost of smaller space heating equipment.
4. The control of hydrocarbons, oxidants and
carbon monoxide is primarily a function of mobile
source control strategies. If fully implemented in
1977, mobile source emission control techniques can
produce a 20-percent reduction in hydrocarbon emis-
sions and a 40-percent reduction in carbon monoxide
emissions. This control capability will increase motor
vehicle cost by about 5 percent. While these emission
reductions and the corresponding urban ambient air
quality improvements which result are significant,
they are not adequate to universally achieve ambient
air quality standards for oxidants and carbon mo-
noxide. As a result, urban areas encompassing more
than 40 million people in the United States will re-
quire at least 10 percent reductions in the vehicle
miles travelled in addition to direct emission control
techniques. Reductions of this magnitude will require
expansion of urban mass transit capabilities extend-
ing past 1980 and incentives to shift urban commuter
patterns to mass transit.
BIBLIOGRAPHY
I. Nationwide Inventory of Air Pollutant Emissions 1970,
U.S. Environmental Protection Agency, October 1972*.
2. Control Techniques for Paniculate Air Pollutants
AP-51, National Air Pollution Control Administra-
tion, January 1969.
3. Control Techniques for Carbon Monoxide, Nitrogen
Oxide, and Hydrocarbon Emissions from Mobile
Sources AP-66, National Air Pollution Control Ad-
ministration, March 1970.
4. Control Techniques for Nitrogen Oxides from Sta-
tionary Sources AP-67, National Air Pollution Control
Administration, March 1970.
5. Control Techniques for Hydrocarbon and Organic Sol-
vent Emissions from Stationary Sources AP-68, Na-
tional Air Pollution Control Administration, March
6. Abatement of Particulate Emissions from Stationary
Sources COPAC-5, National Academy of Engineering,
7. Abatement of Nitrogen Oxide Emissions from Sta-
tionary Sources COPAC-4, National Academy of En-
gineering, 1972 .
8. Feasibility of Emission Standards Based on Particle
Size, U.S. Environmental Protection Agency Contract
68-01-0428, Midwest Research Institute, November
9. A Survey of Emissions and Controls for Hazardous
and Other Pollutants, EPA-R4-73-021, U.S. Environ-
mental Protection Agency, February 1973.
136
-------
10. Proceedings-Symposium on Control of Fine Paniculate
Emissions from Industrial Sources, U.S.-U.S.S.R. Work-
ing Group Stationary Source Air Pollution Control
Technology, lanuary 1974.
11. The Economics of Clean Air, Report of the Adminis-
trator of the Environmental Protection Agency to the
Congress of the United States, Document No. 92-6,
March 1971.
12. Environmental Considerations in Future Energy
Growth, Vol. I, II, and IV, U.S. Environmental Pro-
tection Agency, Office of Research and Development,
April 1973.
13. National Public Hearings on Power Plant Compliance
with Sulfur Oxide Air Pollution Regulations, U.S. En-
vironmental Protection Agency, January 1974.
14 Proceedings of Symposium on Multiple-Source Urban
Diffusion Models AP-86, U.S. Environmental Protec-
tion Agency, 1970.
15. Some General Economic Considerations of Flue Gas
Scrubbing for Utilities, U.S. Environmental Protection
Agency, Control Systems Laboratory, October 1972.
16. Pollution Control and Energy Needs, Advances in
Chemistry Series 127, Chapter 5-T6, Effect of Desul-
furization Methods on Ambient Air Quality, The Amer-
ican Chemical Society, September 1973.
17. The Clean Air Act and Transportation Controls: An
EPA White Paper, U.S. Environmental Protection
Agency, Office of Air and Water Programs, August
137
-------
Development of Maximum Permissible Environmental Loading
(IMPEL)-Water
Mark A. Pisano *
I. INTRODUCTION
Ideally the desired level of water quality improve-
ment and the date it will be achieved would be estab-
lished by taking into consideration the costs of
achieving various levels of improvement and select-
ing the level of control where the costs equate with
the benefits. To this date the water pollution control
program in the United States has not followed a
strategy establishing a control level that in all cases
would balance costs of control options with the bene-
fits of reduced health and ecological damage.
In the 1948 Federal Water Pollution Control Act
the United States established an enforcement con-
ference procedure whereby control levels on indi-
vidual dischargers were established by allocating pol-
lutant load reductions in proportion to the reduction
required to meet locally established water use (bene-
fit) determinations. The process did not result in an
adequate level of water quality improvement. Since
that time there have been successive changes in na-
tional laws, most notably the Clean Water Restora-
tion Act of 1966. This Act introduced the concept
that the level of water quality to be achieved would
be determined not by a local enforcement confer-
ence but rather national water quality standards were
to be jointly established by States and the Federal
Government. These efforts, however, also were not
found satisfactory, owing partially to the complexity
of the water pollution problem and the difficulty in
implementing the regulatory machinery. Implement-
ing a national pollution control program involves a
large number of organizations and institutions, many
of which have competing goals and objectives. To the
extent that the implementation does not specify the
decision rules and the target outcome of such a
process, the resolution of these competing interests
often fails. While national/state standards were es-
tablished, the implementation of these standards was
conducted on a case by case basis. The actual re-
quirements levied against dischargers became unclear
and the schedules for achieving the standards slipped.
In 1972, new legislation was enacted which at-
tempted to solve the problems of previous efforts.
This paper describes the method of establishing water
quality targets and dates of their achievement and
the design of the program for meeting these targets.
II. REGULATORY STRATEGY
Previous papers have dealt with stress of pollutants
on man and on ecosystems. These discussions con-
tained, to the extent possible, an assessment of the
risks and costs associated with stresses on the eco-
system, i-e-, damage functions. From these functions
target maximum permissible species stress could be
established. Another set of papers concentrated on
the control methods that are employed to achieve
pollution control. These discussions also included the
costs associated with these controls. Depending on
the costs of these control methods and the time
period required to implement them, a strategy speci-
fying the level and time period of implementation
of the Maximum Permissible Environmental Load-
ing (MPEL) could be established.
The regulatory strategy in water is not based on
such a balance where the costs of the level of control
would equate with benefits. Rather the central thrust
in the water program is that control measures them-
selves will establish the MPEL. Successive improve-
ments in technology will be required over time, until
a target of zero discharge is achieved. Thus, to a
large extent, the "raison d'etre" of the program is a
reliance on the advancing capability of technology.
The program is based on an assumption that any
discharge to water would upset the natural integrity
of water, therefore any increase in the removal of
pollutants is desirable.
The strategy does recognize that a uniform appli-
cation of technology might not result in an appro-
priate level of controls to protect health and eco-
systems in every case. Thus, the concept of a control
level dependent on the instream criteria which sup-
port the intended use of each specific body of water
•Director, Water Planning Division, U.S. Environmental Protection Agency
139
-------
is still maintained. If this water quality standard can-
not be met by the requisite technology standard for
the period, then a higher level of technology could
be required. Also, if over time it is found that it is
not technologically feasible and/or it is economically
infeasible to move to zero discharge then the water
quality standard can be used to determine a less strict
level of control than zero discharge.
It should be noted, however, that successively
higher levels of technological controls where feasible
will be required even though the water quality stand-
ards might be achieved with less stringent controls;
in which case the cost of control could be greater
than the apparent benefits to society. This apparent
inefficiency is permitted based on the assumption that
if a pollutant can be removed at "reasonable" cost
then it will be. Also, because of our historical in-
ability to adequately equate cost of removing pol-
lutants to the benefits achieved by cleaner water, a
second best solution of at least requiring what tech-
nology can economically achieve regardless of the
cost/benefit calculus has been an important part of
the U.S. regulatory strategy.
There are two parts to establishing the MPEL in
the U.S. The first is effluent guidelines which are like
performance specifications as to what technology can
do and when the technology can be implemented.
The second is the water quality standard which serves
as a guide or check point as to the adequacy of
technology.
Determining Effluent Guidelines (Table 1
provides a summary of these requirements.)
By 1977 all dischargers will be required as a mini-
mum to install best level technology—secondary
treatment for municipalities and best practical con-
trol technology for industries. The key concept is
that by 1977, every facility of a given type and size
must have its pollution release equivalent to all other
facilities of the same type and size. To the extent
possible, the objective is to develop guidelines that
may be applied to individual dischargers without
the necessity of treating each discharger as a unique
situation. Because of the heterogenity of dischargers,
industrial categories have been sub-categorized ac-
cording to their age and size, raw materials, manu-
facturing processes, products produced, available
treatment technology, energy requirements and costs.
Within each category the effluent amount is based
on the known performance of the better performers
having well designed and well operated treatment
systems. A key consideration in the determination
of the effluent limit for the category is a balancing
of total costs of treatment, including capital costs and
operation and maintenance, against the effluent re-
duction benefits achieved. The decision rule is to set
the limit at the point where the marginal cost of re-
moving additional amounts of pollutants significantly
increases. For example, as shown in Figure 1, the
incremental cost of removing additional amounts of
BOD increases rapidly beyond the secondary treat-
ment level for municipal discharges. This was a major
reason for establishing the base level technology for
municipal dischargers as secondary treatment and is
expressed as shown in Table 2.
For industrial categories the removal levels and
the parameters considered are not uniform between
categories. These variations are shown in Table 3
which presents summary data for six effluent guide-
lines. Because of technological advances in some in-
dustries, the cost/effluent reduction benefit test re-
sults in a base removal level of zero discharge, e.g.,
insulation fiberglass.
Table 1, Principal Statutory Considerations for Effluent Guidelines
Statutory Basis
Best Practicable Control
Technology Currently
Available
[Existing Sources]
Best Available Tech-
nology Economically
Achievable
[Existing Sources]
Standards of Performance
Best Available Demon-
strated Control Tech-
nology
[New Sources]
General Description
1. Achieve by 1977.
2. Generally average of best exist-
ing performance; high confidence
in engineering viability.
3. Where treatment uniformly in-
adequate a higher degree of treat-
ment may be required if practicable
[compare existing treatment of
similar wastes].
1. Achieve by 1983.
2. Generally best existing per-
formance but may include tech-
nology which is capable of being
designed, though not yet in place;
further development work could
be required.
1. Achieved by sources for which
"construction" commences after
proposal of regulations.
2. Generally same considerations
as for '83; more critical analysis of
present availability.
Process Changes
Normally does not
emphasize in-process con-
trols, except where pres-
ently commonly
practiced.
Emphasizes both in-
process and end-of-
process control.
Emphasizes process
changes.
Cost
Balancing of total cost of
treatment against effluent
reduction benefits.
Cost considered relative
to broad test of reason-
ableness.
Cost considered relative
to broad test of reason-
ableness.
140
-------
Figure 1. Typical Values, Biological
Oxygen Demand (BOD)
Raw
Primary
Secondary (Hi-Rate)
Secondary
(Conventional +)
Two Stage Nitrification
Advanced Waste
Treatment +
Relative
Cost
0.00
0.50
1.00
1.30
1.75
2.25+
Carbona-
ceous
300
180
45
23
23
?
Removal
0%
60%
85%
92%
98%
99%
Table 2. Base Level Technology Limits
Parameter
Biochemical Oxygen
Demand (5-day)
Suspended Solids
Fecal Coliform
Bacteria
PH
Measurement
mg/liter
mg/liter
number/ 100 ml.
units
Monthly Monthly
Average Average
30 45
30 45
200 400
within limits
of 6.0-9.0
By 1983 all industrial dischargers will be required
to install best available technology (BAT). The level
of control will be based on best performer of a
category and will not include just end-of-pipe treat-
ment, but process changes and product modifications
as well. In some cases BAT may reflect prototype
technology not yet demonstrated to be practicable.
The objective is to develop technologies with higher
removal rates, including some calling for "no dis-
charge."
In setting the BAT limits, sub-categorization will
also be employed. Within these categories costs will
be considered relative to a broad test of reasonable-
ness which includes the feasibility and cost of the
technology, i.e., can the control be put in place at a
cost that is not prohibitive. The economic test will
not necessarily be the cost/amount of effluent re-
duced.
For municipal dischargers the 1983 requirement is
not similar to the industrial standard. Instead of in-
creasing the amount of reductions to be obtained by
requiring more advance technologies, municipalities
are required to consider alternative waste water man-
agement techniques and choose the most cost-effec-
tive and environmentally advantageous alternative.
Treatment and reuse, and land treatment of wastes,
both of which provide either zero discharge or high
levels of treatment, must be among the alternatives
considered. The environmental and cost-effective
tests may demonstrate that a higher level of removal
above base level technology or an alternative man-
agement technique may be required. The test is not
uniform; rather a case by case assessment must be
made.
Setting Water Quality Standards
Water quality standards are established jointly by
the States and Federal government. Each of these
standards consists of a designated use criterion repre-
senting the acceptable limits of pollutants in receiv-
ing waters dedicated to that use. These criteria are
based on the latest scientific knowledge of the effects
of pollutants on human health and biological com-
munities. With respect to uses, the U.S. has set as a
goal that where attainable all waters will be swim-
mable and will provide for the protection and propa-
gation of fish, shellfish and wildlife by 1983. The
phrase "where attainable" recognizes that naturally
occurring conditions, or uncontrollable non-point
source pollution, could result in a failure to meet the
1983 goal. Table 4 provides an example of the 1983
use classification and criteria that are likely to be
needed to support these uses. The criteria in this
table are changing as new data are acquired. The
process for setting these standards is to first establish
the intended use of the water body and the intended
biological activity desired. For these uses and bio-
logical protection the appropriate criteria (MFC)
are defined, for example, temperature in a water body
Table 3. BPT for Industry
Industry
Cement
Meat Processing
Rubber
Feedlots
Flat Glass
Insulation
Fiberglass
No. of Plants
166
1200-4800 very
small plants
92
3500 large
1.6M-1.8M
47
19
Estimated Percentage
Raw Waste Load
Reduction Expected
BOD
• —
80-99
—
100
100
TSS
84-100
97- 99
80
100
75-100
100
Other Pollutant
Parameters Controlled
pH, Temperature
pH, Temperature, COD, TDS,
Phosphorus, Nitrogen, Oil and
Grease, Chlorides
COD, Oil
COD, Chlorides, Nitrogen,
Phosphorus, Fecal Coliform
Oil, COD, pH, Phosphorus, TDS,
Temperature
Phenols, COD, TDS, Oil and
Grease, pH, Temperature,
Ammonia
141
-------
Table 4. Example Water Quality Criteria Summary by Use Classification
Use Class.
Class A
Primary
Contact
Recreation
(swimming,
water ski-
ing, etc.)
Class B
Desirable
Species of
Aquatic
Life and
Secondary
Contact
Recreation
(boating,
fishing,
etc.)
Micro-
biological
Shall not
exceed a
geometric
mean of
200 fecal
coliform
per 100 ml.
Shall not
exceed a
geometric
mean of
10,000 total
coliform or
2000 fecal
coliform
per 100 ml.
(Fecal coli-
form counts
are pre-
ferred.)
Dissolved
Oxygen
Not less
than 5 mg/
1. Class B
levels also
apply.
Not less
than 5 mg/
1 (except
for 4 mg/1
for short
periods of
time within
a 24-hour
period).
Not less
than 6 mg/
1 in trout
waters. Not
less than 5
mg/1 in
marine
waters.
Temp.
90'F Max-
imum. Class
levels also
apply.
Cold Water
(Trout)
50°F rise.
Max. of
68°F.
Warm
Water
(Bass etc.)
5°P rise in
j r i toe 111
streams.
3 °F rise in
impound-
ments.
Max. 90'F.
Marine
Water
m-Frise.
Hydrogen
Ion
Hydrogen
ion concen-
trations ex-
pressed as
pH shall be
maintained
between 6.5
and 8.3.
Hydrogen
ion concen-
trations ex-
pressed as
pH shall be
maintained
between 6.0
and 9.0.
Dissolved
Solids
Shall not
exceed 500
mg/1 or
one-third
above that
characteris-
tic of nat-
ural condi-
tion (which-
ever is less).
Shall not
exceed one-
third above
that charac-
teristic of
natural con-
ditions.
Taste
and Odor
Producing
Substances
None in
amounts
that will in-
terfere with
water con-
tact use.
Shall con-
tain no sub-
stances
which will
render any
undesirable
tastes to fish
flesh or in
any other
way make
fish inedible.
Dissolved
Gas
Class B lev-
els apply.
Cold Water
Total dis-
solved gas
pressure not
to exceed
110 percent
of existing
atmospheric
pressure.
Color and
Turbltv
• «»»»»«ij
Producta*
Substance!
Secchi disc
visible at
min. depth
of 1 meter.
Cold
Waters
10 JU
Warm
Waters
50 JU
Marine
Waters
Secchi disc/
visible at
min.
depth of 1
meter.
will vary according to the level of biological protec-
tion. Finally, the flow regime defining the critical
periods when standards must be met are specified,
for example, 7-day, 10-year low flow in humid areas,
5-day, 5-year low flow in arid areas, etc.
III. IMPLEMENTATION STRATEGY
The immediate task of the U.S. regulatory pro-
gram is the classification of all hydrologic segments.
Where the application of best practicable technology
for industries and secondary treatment for municipal
plants will result in meeting 1977 water quality
standards, the segment will be categorized as an
effluent guidelines limited segment. Where this con-
trol technology will be insufficient to achieve the
necessary level of water quality, the segment will be
classified as a water quality limited segment. Where
there is significant doubt, a segment will be con-
sidered as a water quality limited segment, subject
to a later reclassification.
About half of the segments are so heavily polluted
the dischargers located on them will be required to
go beyond 1977-level pollution controls to enable
water quality standards to be met:
Effluent guidelines limited segments 1515
Water quality limited segments 1588
Total segments identified by States
3103
Short Run Control Program
All dischargers must obtain permits which dictate
the amount of discharge that is allowed to be dis-
charged from an individual plant or municipality, and
a schedule of remedial measures, including an en-
forceable sequence of actions or operations leading
to compliance with the effluent limitation. Failure to
comply with an interim or final requirement in the
permit constitutes a violation of the permit, for which
the permit holder could be fined up to $25,000 per
day.
Effluent Segments
The effluent conditions for dischargers in effluent
segments are based on the effluent guidelines. A hy-
pothetical example will be presented to illustrate
the application of these guidelines. A proven and
feasible technology for treatment of pickle-packer
wastes is an "activated sludge treatment plant"; this
is essentially just like a municipal sewage treatment
plant, and is a good treatment technology for food-
processing wastes. Furthermore, let us suppose that
20 percent of the dill pickle-packers already use this
treatment method, and that the average performance
of the best-run treatment facilities results in 1 pound
of BOD (oxygen-demanding organic wastes), 2
pounds of suspended solids, and zero (no)-discharge
142
-------
of dill seeds or garlic cloves per 1,000 pecks of
pickles packed. Thus our final effluent limitations
might well be stated as shown in Table 5.
Table 5. Hypothetical Effluent Limits
INDUSTRY: Pickle-Packers
SUBCATEGORY: Dill Pickles
Pollutant
Parameter
BOD
Suspended
Solids
Dill Seeds
Garlic Seeds
Effluent Limitations
Amount
Permitted
1 pound
2 pounds
Zero Discharge
Zero Discharge
Per Unit of
Production
Thousand pecks of
pickles packed
Thousand pecks of
pickles packed
(No units needed)
(No units needed)
Given this sort of effluent limitation, we could
easily calculate the pounds of BOD and suspended
solids a dill pickle-packer of any size could be al-
lowed to release and still be consistent with this
standard. For example, a pickle-packer processing
5,000 pecks of dill pickles per 5-day week would
receive a permit allowing release up to 5 pounds of
BOD and 10 pounds of suspended solids per week,
or 1 pound of BOD and 2 pounds of suspended
solids per day, from the dill pickle part of his facility.
The permit would also specify the final date when the
packer would be required to achieve this level of dis-
charge, and interim compliance milestones, e.g.,
begin construction.
Water Quality Segment Analysis
Violations of standards in water quality segments
could be caused either by point sources which must
be subjected to controls beyond best practicable
treatment (BPT)/secondary treatment or by non-
point sources which must be controlled. The imple-
mentation program in the U.S. currently concentrates
on point source discharges. Ideally the water quality
segment analysis would determine the most effective
allocation of waste leads between all point and non-
point discharges. In the short run, however, if the
contribution of the point source dischargers cannot
be distinguished from the nonpoint source dis-
chargers, then base level technology will be applied
to the point source discharge.
The analysis would be completed for each para-
meter which is in violation of water quality standards.
The analytical process for performing the load al-
location for these parameters is outlined in Figure 2.
Each source contributing that parameter to the seg-
ment should be identified and alternative remedial
measures considered. The final treatment/control
strategy for the segment should reflect a combination
of control methods which will meet water quality
standards for all water quality parameters.
As noted in Figure 2, modeling is generally the
appropriate method of ascertaining total maximum
daily loads and determining the effects of the pro-
posed alternative abatement strategies.
Water quality analysis through modeling enables
planners to predict water quality under adjusted con-
ditions of flow, temperature and pollutant loads.
Hence, it provides a basis for load allocation and
effluent reduction determinations.
Water bodies may be categorized in one of three
categories: flowing streams, estuaries, and lakes and
impoundments. Dominant transport mechanisms dif-
fer in each category; hence, different modeling tech-
niques are appropriate for different categories.
Further, the degree of sophistication of technique
within each water body category may also vary ac-
cording to conditions in the water.
The water quality analysis and prediction de-
veloped by any model can only approximate the
actual water quality which will occur under the
various suggested hypotheses. These assumptions will
never be entirely correct for the distinct water body
being analyzed. Hence, remedial measures (effluent
reductions) based on the model predictions will not
result in water quality exactly as predicted by the
model. Since unnecessarily stringent measures may
result in costly overbuilding and inadequate meas-
ures may fail to protect the aquatic ecosystem and
achieve established water quality goals, model selec-
tion must consider the degree of risk to be accepted.
Table 6 summarizes the criteria for the selection
of an analytical technique. The table lists criteria
for each of the four levels of complexity. For each
level of complexity, the table presents the type of
problems and water bodies for which the level is
appropriate, the planning characteristics (complexity
and risk) associated with that level and the time
required for a study.
• Type A simplified analysis generally has rela-
tively few waste sources and an assumption is
made that there are no changes in waste loading
and hydrology with time (steady state).
• Type B steady state linear kinetics has several
sources which have overlapping stream effects,
steady state conditions still hold, i.e., no change
in waste loading and hydrology.
• Type C transient linear kinetics has multiple
sources which have overlapping stream effects,
waste loading and hydrology vary linearly with
time, e.g., storm event diurnal variations and
variable waste discharges. Type D time variable
non-linear kinetics has multiple sources which
have overlapping two-dimensional stream, lake,
or estuary effects. The internal mixing process
represents a major problem.
Once a model is selected a set of worst conditions
is assumed and streamflows and temperatures are
fixed to arrive at a waste load allocation. By suc-
cessive approximations of waste inputs, the maxi-
mum allowable effluent levels for each of the dis-
143
-------
DETERMINE
DO
STANDARD
I
SELECT
BACKGROUND
DO DEFICIT
WATER
QUALITY
MODEL
1
INPUT "SECONDARY"
TREATMENT LEVEL
AT EACH SOURCE
1
MEETS
DO STANDARDS?
(INCLUDING BACKGOUND)
YES
i
NO
ALLOCATION IS
AS GIVEN BY
'SECONDARY" TREATMENT
i
INCREASE
DISCHARGE
LEVEL UNTIL
STANDARD IS VIOLATED
"FIGURE 2b.
"EQUIVALENT"
RESERVE CAPACITY
FIGURE 2a. SUGGESTED ALLOCATION PROCEDURES (BOD - DO EXAMPLE)
chargers can be established to meet the water quality
standard for a particular constituent. The goal may
consider a reserve margin for accommodating future
projected loads or uncontrolled natural background
levels. The total allocation for a segment is the sum-
mation of all individual allocations of the component
discharges.
In a system with a large number of pollutant
sources the chance of each discharge being at its ex-
treme allocated stress value simultaneously with all
of the other sources and with the rare low design
streamflow is statistically small. To issue permits
based on the daily averages would be unnecessarily
restrictive. It is reasonable therefore to define the
permitted waste load for each discharger as a weekly
average waste discharge.
144
-------
INCREMENT EACH
DISCHARGER BY
DISCRETE LEVEL
WATER
QUALITY
MODEL
I
CHECK FOR
STANDARD
ATTAINMENT
YES
I
NO
MAXIMUM ALLOWABLE
DISCHARGER LOAD
1
SELECT
RESERVE
FACTOR (R)
RESERVE
CAPACITY =
(R) MAX. ALLOWABLE
ALLOCATION =
R x MAXIMUM
ALLOWABLE
CHECK FOR
UPPER BOUND
CONSTRAINT
FIGURE 2b. SUGGESTED ALLOCATION PROCEDURES, ctd.
145
-------
Table 6. Criteria for Selection of Techniques
Model Complexity
Water Quality Problems
and Variable
Water Body
Planning
Characteristics
Time Required
for Study
Simplified analysis
(Type A)
D.O. (carbon and ni- One-dimensional
trogen) streams and estuaries
(completely mixed).
a. Low risk of capital
and/or environmental
quality degradation.
b. No alternate strate-
gies and control op-
tions available.
Days to weeks.
Steady state linear
kinetics
(Type B)
a. D.O (carbon and One or two-dimen-
nitrogen), temperature sional streams, estua-
and nonpoint source. ries, rivers, lakes.
b. Anticipated or ex-
isting water
problems.
quality
a. Low to moderate 2 to 9 months.
risk of capital and en-
vironmental quality de-
gradation.
b. Alternative strate-
gies and control op-
tions must be availa-
ble.
Transient linear kinetics
analysis
(TypeC)
a. Time varying D.O.,
nonpoint source anal-
ysis, and temperature.
Simple eutrophication
analysis. Full storm
water overflow analy-
sis.
b. Water quality prob-
lems.
Rivers, lakes and estu-
aries. One or two-di-
mensional.
a. Moderate to high
risk of capital and en-
vironmental quality de-
gradation.
b. Alternative strate-
gies and control op-
tions must be availa-
ble.
6 to 24 months.
Time variable non-linear
kinetics analysis
(Type D*)
a. Detailed eutrophica-
tion analysis, etc.
b. Water quality prob-
lems.
c. High growth of area
projected.
All bodies of water.
a. High risk of capital
and/or environmental
quality degradation.
b. Alternative strate-
gies and control must
be available.
12 to 36 months.
•Type D studies are to be performed only in very complex situations and should be regarded as research efforts,
Exceptions to this may be desirable when one or
two dischargers dominate the combined load or when
relatively large loads occur within short time periods
(as opposed to reasonably steady day-to-day dis-
charges). If there are only one or two dischargers in
a segment, the probability of all peak loads occurring
simultaneously is greatly increased. If the environ-
mental risk is high, shock loads of some constitu-
ents should be restricted. For such cases, permits
specified on a shorter term such as a daily mean
rather than a weekly average would be prudent.
Long Run Control Program
On some water quality segments the total loading
discharged into the stream may be such that the best
known technology may not be adequate to achieve
the water quality standard when applied to existing
point source discharges. On the other hand, the cause
of the problem may be a non-point source problem.
In many urban areas the control problem might be
urban run-off or combined sanitary and storm wastes
which bypass the treatment plant during periods of
high flow. Construction of roads, housing and other
facilities also creates problems of increased sediment
run-off.
Where serious water quality control problems
exist, areawide waste water planning and manage-
ment agencies are being established. The agencies
will develop control strategies that look not just at
technology solutions, such as treatment facilities, but
also will examine land use and land management
alternatives. In many of these situations, altering or
controlling future sources of pollution might be the
only feasible solution. This might involve controls
on the amount of growth that could occur in an area,
or controlling where new facilities could locate. Or
the solution might involve affecting the way in which
land is used, e.g., the design of residential subdi-
visions.
In establishing the levels of controls for dis-
chargers in water quality segments where higher
levels than BAT are required, consideration of the
social and economic costs and benefits of alternative
control strategies is required. The control strategies
could range from technological solutions to land use
solutions. If there are no feasible solutions with costs
equal to the benefits then the goal of swimmable
waters and a balanced ecosystem may need to be re-
laxed.
146
-------
IV. CONCLUSION
Dr. Izrael pointed out that there are three methods
of setting the MPEL:
• one that is ecologically desirable;
• one that reflects public health;
• one that we can technically achieve.
In the United States we are currently following a
mixture of these approaches with an emphasis on
doing what is technologically feasible, but not be-
cause a mixed strategy is necessarily better. Rather,
our philosophy is that if it can be done, then do it,
even if technology over-protects or under-protects
public health and ecological objectives. As a scientific
community we must find out what information the
administrative process needs in order to fully imple-
ment a mixed strategy that balances all three of the
limiting conditions for establishing the MPEL. To
achieve this balance we need to improve our knowl-
edge concerning concentrations of pollutants and
their health and/or ecological effects. We also need
to further our understanding of the frontiers of tech-
nology.
We also need to improve our ability to make better
decisions on specific control requirements for indi-
vidual dischargers. We need to improve our ability
to make waste load allocations. It does not help the
regulatory program if we have complete information
on what the water quality criteria should be, but we
cannot monitor or determine the control measures
that are to be levied against point and non-point
sources to achieve these criteria. A possible effort of
future exchange could be improving the monitoring
and analytical techniques used to relate water quality
criteria to required control programs within a water
body.
147
-------
The Effect of Polluting a Biogeocoenosis With Strontium 90
on a Population of Mammals and the Zoocoenosis
V. Ye. Sokolov, A. I. Il'yenko, D. A. Krivolutskiy *
Industrial pollution of the biosphere significantly
affects the ecology of the organisms that populate
the biogeocoenosis. This effect may be expressed by
disruption of the biological processes taking place in
the population, as well as by change in the relation-
ships between organisms, and between organisms and
the physicochemical factors of the habitat. Chemical
pollution of the biogeocoenosis accumulated by its
components can be considered to be a new abiotic
factor for animals that acts on the population, com-
munity, and on the biogeocoenosis as a whole, along
with the usual environmental factors of the habitat
in which the organisms live.
The experimental introduction into the biogecoe-
nosis of radioactive elements such as strontium 90,
the chemical analog of Ca, for example, which is
distinguished by the simplicity of determining the
quantitative distribution of the isotope in a popula-
tion, and among the components of the biogeocoe-
nosis by radiometric and dosimetric methods, pro-
vides a way to establish the most vulnerable links
in the animal ecology in the face of chemical pollu-
tion. The disruptions discovered in the ecology of
populations and zoocoenosis in a biogeocoenosis
polluted by strontium 90 can, possibly, serve as a
model and be valid for many types of chemical
pollution.
The question of the combined complex effect on
animals of certain factors man has introduced into
natural biogeocoenoses is being discussed widely to-
day. We will devote special attention in our report to
the effect that factors of the habitat which are normal
for the natural populations have against a back-
ground of radioactive pollution of the biogeocoenosis.
The work was performed in two open-air cages,
each with an area of one hectare. One of the cages
was polluted with strontium 90 at a pollution density
of from 1.8 to 3.4 millicuries/m-. The other cage,
which we shall conventionally call the "clean cage,"
had a background pollution consisting of artificial
radioisotopes on a global level.
The cages were populated by two species of ro-
dents: field (Microtus agrestis) and northern red-
backed (Clethrionomys rutilus) voles. The beasts
were trapped in the cages once a month, and were
marked by amputating the toes in such a way that
each zooid had its own individual number. Blood for
study was drawn simultaneously, and part of the tail
was severed for radiometric analysis.
Systematic observation of the tagged population
yielded much data, and these are included in the
report.
The effect of polluting a biogeocoenosis with
strontium 90 on the age structure of a popu-
lation and the course of the biological cycles
The populations of the two species of voles kept
under the conditions that prevailed in the open-air
cage where the experimental section had pollution
levels of 1.8-3.4 millicuries/m2, showed differences
in the degree of accumulation of strontium 90
-yttrium 90 in the skeleton (Il'yenko, 1967; Soko-
lov, Il'yenko, 1969). The concentration of isotopes
in the boney tissue of the field (Microtus agrestis)
and northern redbacked (Clethrionomys rutilus)
voles was 0.11 ± 0.001, and 0.5 ± 0.001|ucuries/g,
respectively, and the dose rate from the isotopes de-
posited in the skeleton was 0.77, and 0.35 rad/day
for the bone marrow. The survival of the overwin-
tered populations of these species of voles, that is
for those who lived in the cage for more than a year,
was not identical under these conditions (Figure 1).
Evidently because of differences in the dose of ir-
radiation from the isotopes deposited in the skeleton,
the mortality was higher, and the overwintered popu-
lation died off more rapidly, among the population of
field voles, than was the case in the clean area. This
pattern was less pronounced among the population
of northern redbacked voles. The mortality among
the population of field voles during the winter in the
•A. N. Severtsov Institute of Evolutionary Morphology and Animal Ecology of the Academy of Sciences of the USSR, Moscow
149
-------
/o
100
80
60
40
20
- *" °, FIELD
. \\
• v
• \\
\ Xo
VOLES
• V
Experimental — ^\
•
•
i i i i i
\— - Control
\
N^ \
>v \
\. \
X. N
\Xo
1 1 ^«
NORTHERN REDBACKEO VOLES
Experimental
4 5 678 9 10 11
MONTHS
FIGURE 1.
SURVIVAL OF OVERWINTERING
POPULATION OF VOLES IN AN
AREA POLLUTED BY STRONTIUM
90 AND A "CLEAN" AREA.
polluted and clean cages was 95.2% and 50%, re-
spectively. It is possible that differences in the radio-
sensitivities of these species also affected the mor-
tality rate. Thus, the LD5o/3o for the northern red-
backed voles is 960 rads, and is somewhere around
600 rads for field voles. The life span of field voles
that survived the winter in the populations under
study differed little, and those differences found
were statistically unreliable. The mean life span for
beasts in the polluted section, and in the control,
was 11.6 ± 0.8, and 12.5 ± .08 months, respec-
tively.
The populations of small mammals receive dif-
ferent doses of ionizing irradiation at different sea-
sons of the year. The result is that the survival rate
of the groups of beasts in the population in the pol-
luted area is unequal (Figure 2) among the field
voles born during different reproduction periods.
Most of the beasts born at the beginning of the sum-
mer do not live until the spring of the next year. The
100
80
60
40
20
Control Group
Polluted Territory
100
80
60
40
20
iy
Animals surviving to
second half of winter
Animals born at the beginning of summer
I I Born at the end of summer
FIGURE 2.
RATIO OF AGE GROUPS IN A POPULA-
TION OF FIELD VOLES IN WINTER AND
SPRING, AND THEIR SURVIVAL RATE
IN THE POLLUTED TERRITORY AND
IN THE CONTROL.
number of animals born at the end of a reproduction
period was 98% of the spring population. Only
half of the animals born at the end of the summer
survive the winter and live until the spring of the
next year. Few deaths are recorded among the con-
trol population, however, The structure of the rodent
population in the polluted territory changes substan-
tially as a result.
150
-------
Disruption in the reproduction rate was established
during the study of reproduction patterns in the
rodent population in the experimental sections pol-
luted with strontium 90 at levels of from 140 to
1200 millicuries/m2. The ovum mortality rate before
implantation increased with increase in the strontium
90 - yttrium 90 concentration in the skeletons of
long-tailed mice (Apodemus sylvaticus) (Figure 3).
The sharp increase in the ovum mortality rate,
greater than that in the control by a factor of 3, was
observed when the concentration of strontium 90 —
yttrium 90 in the animals' bone marrow was 0.003
/icurie/g. The number of embryos in the females
among the population of northern redbacked voles
decreased with an increase in the concentration of
isotopes in the skeleton (Figure 4). The study of the
biological patterns found among isolated populations
under open-cage confinement revealed that reproduc-
tion among the field vole population ended a month
earlier than among the control population of this
species confined under analogous conditions, but on
unpolluted territory.
The Effect of Climatic Factors
The action of ionizing radiation on a population
is particularly strong during the extremal period of
the existence of small mammals, when such action
intensifies the action of all the unfavorable factors.
This is why the mortality rate for the rodent popula-
tion is particularly high in the winter time in an area
polluted by strontium 90.
The rodent population does not renew itself in
the winter time, something that is characteristic of
other seasons of the year, and is the result of these
animals' reproduction. The population dies off rap-
idly during this period, the result of the effects of a
number of unfavorable climatic factors. The most
critical times in the lives of the rodents are the peri-
ods of establishment of the permanent snow cover,
and of the melting of the snow in the spring. Winter
weather conditions are a lesser cause of death among
small mammals and act differently on different
species.
10
o
a.
C
O
u.
ill
00
V)
UJ
o
LL.
O
LU
APODEMUS SYLVATICUS
Back- 5x10~4 1 x10~3 3x 10~3 9x10~32x10~* 6x10~2
ground
CONCENTRATION OF RADIOISOTOPES IN THE SKELTON,
H CURIES/g
FIGURE 3. MORTALITY OF OVULES BEFORE IMPLANTATION
IN A POPULATION OF LONG-TAILED MICE
151
-------
I
.fl
i e
10-1
10
-2
10-
control
Concentration in skeleton, wcuries/g
FIGURE 4. NUMBER OF EMBRYOS IN A
POPULATION OF NORTHERN
REDBACKED VOLES AS A
FUNCTION OF THE CONCEN-
TRATION OF STRONTIUM 90 -
YTTRIUM 90 IN THE SKELE-
TONS OF FEMALES
The data in Figure 5 were obtained from constant
observation of individually tagged beasts kept in
open-air cages. As may be seen, the mortality rate
among the population of field voles on the territory
polluted by strontium 90 was caused primarily by
ionizing radiation during all seasons of the year, the
effect of which is intensified by climatic factors. The
autumn and spring mortality rates among the popu-
lation of northern redbacked voles, a more radio-
resistant species already mentioned above that re-
ceived smaller doses of ionizing irradiation, were
determined by the effects of weather conditions un-
favorable to the animals. Only in the wintertime was
the more severe destructive effect of ionizing irradia-
tion clear cut.
Thus, the effect of unfavorable habitat climatic
factors on a population is intensified against a back-
ground of ionizing irradiation of the animals' bodies.
The Effect of Biotic Factors
The effect of the usual habitat ecological factors
is intensified considerably when the population con-
3 40 .
60
£ 40
o
20
NORTHERN
REDBACKED VOLES
Autumn Winter Spring
•i POLLUTED CAGES
CD CLEAN OPEN-AIR CAGES
Total for
autumn-spring
period
FIGURE 5. MORTALITY IN POPULATIONS
OF VOLES IN A SECTION POL-
LUTED WITH STRONTIUM 90
IN WINTER.
sists of herbivores, weakened by irradiation by ioniz-
ing radiation.
Study of the patterns of the use of a territory by a
rodent population, using individual tagging of the
animals and conducted by many researchers, has
shown that the sections over which these animals
forage will change in size, depending on the effect
on the population of many of the usual habitat
factors. Table 1 lists our data on the dimensions of
the individual sections for adult male and female
field voles during the most intensive reproduction
period, midsummer. The population density was the
same in the polluted and clean open-air cages.
The combined action on the females of ionizing
radiation and the physiological load caused by preg-
nancy and feeding the young was expressed by a
reduction in the area of the territory they used to a
much greater extent than was the case for males af-
fected just by ionizing radiation.
152
-------
Table 1
Change in the Area of Individual Sections in Popu-
lations of Field Voles Kept in Open-Air Cages in a
Territory Polluted by Strontium 90, M2 (Mature
Animals)
Polluted Section
Clean Area
Differences
in area of
sectors from
Number of Mean Number of Mean mean dimen-
Sexual individual area of individual area of sions in
groups sections section sections section control, %
Males 93
Females 134
828±52
473*15
50
41
860±81
696±43
3.7
32,0
Ionizing radiation causes a reduction in the nat-
ural resistance of the organisms to bacterial exo-
and endotoxins and infections and parasitic diseases.
The increase in the sensitivity of irradiated animals
probably is caused by a reduction in their resistance
to infections, and not by an increase in the virulence
of the microorganisms in their bodies. A number of
blood parasites were detected in the blood of the
field voles we studied. Most were leukocytogregarines
(Leucocytogregarinae microti~). The number of
beasts in whose blood endoparasites were detected
was considerably greater in the population subjected
to the constant action of pollution by strontium 90
during all seasons of the year than was the case
among the control. Their summertime number ex-
ceeded the number of infested voles in the control
cage by a factor of more than 6. None of the control
animals were found to have blood parasites in the
wintertime, generally speaking, whereas the number
of infested animals was 16.5% in the experimental
cage. Consequently, the effect on the rodent popula-
tion of ionizing radiation from the strontium 90 de-
posited in the skeletons was expressed by a substan-
tial increase in the number of beasts infested with
blood parasites. Moreover, there was an increase in
the number of blood parasites in the voles thus in-
fected within the territory polluted by this isotope.
Thus, 1 to 8 leukocytogregarines per 100 leukocytes
were found in the blood of the control animals,
whereas 40-60 were found in the blood of voles in
the polluted area, and in excess of 80-100 speci-
mens in individual cases.
The beta-radiation effect on the bone marrow from
the radioisotopes deposited in the skeleton caused an
increase in the number of leukocytes in the peripheral
blood of the voles. The number of leukocytes was
even higher among the beasts infected with blood
parasites in the population of field voles living on
the polluted territory. It seems obvious that the effect
of blood parasites on the bodies of the voles is in-
tensified considerably against a background of ioniz-
ing irradiation (Figure 6).
The degree to which the various species of small
mammals can become infested with parasitic mites
•o
c
(Q
VI
3
o
£.
-------
found that the number of mites on the infested ani-
mals, and the population infestation rate, increase
with increase in the concentration of strontium 90 in
the boney tissue. These patterns were found to be
characteristic for populations of all four species of
rodents.
Figure 7 shows the differences in the relative abun-
dance of gamasid mites on the four species of mature
rodents, which differences are associated with change
in the dose-rate of ionizing radiation from the stron-
tium 90 — yttrium 90 accumulated in the skeleton.
There is no doubt that irradiation of the bodies of
small mammals by beta-radiation from the radioiso-
topes deposited in the skeleton weakens the animals,
leads to inhibition of defense reflexes in protecting
against ectoparasites, and causes an additional, great-
er effect by the latter on animals in the population.
Changeability in Criteria in Animal
Populations
As a rule, the effect of pollution of a locality by
radioactive fission products on certain parameters of
an animal population is studied by comparing the
mean values of the criteria, or properties, of the
populations being compared. At the same time, a
number of researchers have shown that an equally
important example of the biological effect of en-
vironmental factors on an animal population is the
changeability, or the differing qualitative condition,
of the population composition, reflecting the adapt-
ability interrelationship between the organisms and
living conditions and changes in these conditions.
An evaluation of the extent of the effect of pollu-
tion on the population, expressed as the coefficient
UJ
u
I
to
<
u.
O
X
UJ
0
Long-tailed Mouse
i—i 2
0.12 control
0.0025
(\
on
I
nei
n i
I
-let
! KXXXXXI
1CK6
3 V
1.4 „„„ 0.003
0.12 control
*
•
-
^
^
/
/
/
/
Field Mouse
7-
y
vXXXXXi
^
1.7
012 control
0.003
»
KXXXXXXXM
NXXXXN
rui
idn
I
a Vole
n
21 0.034
0.36 control
DOSE RATE, rads/day
FIGURE 7. CHANGE IN THE INDEX OF ABUNDANCE OF MITES IN
MATURE RODENTS AS A FUNCTION OF THE DOSE RATE
OF BETA-RADIATION IN THE SKELTON
154
-------
of variation, was found to be a more discriminating
and reliable index than the mean value of the cri-
terion (Il'yenko, 1968).
We studied the changeability in a number of pop-
ulation and morphological criteria in the populations
of certain soil invertebrates and small mammals over
an experimental territory polluted by strontium 90.
An increase in the degree of changeability in cri-
teria in the populations in the polluted locality was
a general pattern for all the species studied.
Organisms incapable of active migrations for long
distances were selected as the object for studying
changeability in soil animals. It can be assumed that
the populations in the polluted section consisted only
of animals that had experienced the effect of ioniz-
ing radiation over the course of several generations,
and which, moreover, had not been increased by
immigrants from the surrounding territory. The ob-
servations were of two species of soil myriapods
among the geophilids. A reliable increase in change-
ability in the number of pairs of legs was noted only
for Pahymerium ferrugineum males. A tendency
toward an increase in changeability was noted for
the remaining groups, which makes it important
to take into consideration the very rigid control
conditions (Krivolutskiy, Smurov, Snetkov, 1972).
All cases of deviation from the norm in the number
of segments in the antennae (14 in this case) also
were taken into consideration for all myriapods.
The reliability of the difference in the sampled per-
centages of myriapods with these deviations was eval-
uated by the 0-Fisher method, and leaves no doubt
as to the statistically reliable increase in. the fre-
quency with which deformities are found in the ex-
perimental population.
Figure 8 shows our data (Il'yenko, 1968; Sokolov,
Il'yenko, 1969) on changeability in certain of the
morphological criteria in populations of northern
redbacked voles on an experimental territory pol-
luted by strontium 90 at levels of from 0.6 to 3.4
millicuries/m2. The coefficient of variation increases
in the majority of the morphological criteria under
these conditions of existence. This points to an in-
crease in the degree of changeability in animals in a
population inhabiting a biogeocoenosis polluted by
strontium 90. Statistical processing of the data
showed the high degree of reliability of their dif-
ferences. A graphic representation of the distribu-
tion of criteria in the population on the polluted
territory, and in the control, showed that the excess
in the criteria distribution curve changes under the
effect of ionizing radiation, and the amplitude of
their changeability in the population subjected to
chronic irradiation increases because of the appear-
ance of a known number of animals with the criteria,
the value of which exceeds the limits of change-
ability in the same indices for animals in the control
population. Asymmetry of the criteria distribution
curves will be seen in some cases, and this can be
Changeability of
control population
rr/7/r i
' r/ // ' '
nzzzn
n/ / / n
ll/// Al
I I// //I I
i i// //i 1
1 K///I
L
v r r n
/\
Body weight
Body length
Ear length
Tail length
Rel. weight of spleen
Rel. weight of liver
No. leukocytes
No. agranulocytes
Length of cranium
Height of cranium
Width of cranium
No. embryos in females
Life span
I I // /\ I
' r/ //I I
Changeability in
population in
polluted section
FIGURE 8. CHANGEABILITY IN MORPHO-
LOGICAL CRITERIA AND POP-
ULATION INDICES IN A
POPULATION OF NORTHERN
REDBACKED VOLES IN A
SECTION POLLUTED BY
STRONTIUM 90.
explained by the different effect of ionizing radiation
on different animals in the population. The reliability
of the differences in the distribution curves of the
indices for the populations being compared was con-
firmed by checking in terms of the chi-square dif-
ference criterion. The numerical values of the co-
efficients of variation for different criteria in the pop-
ulation increase with increase in the concentration
of strontium 90 in the skeleton of the northern red-
backed voles. Consequently, the coefficient of vari-
ation can be a reliable index of the relationships
between a radioactive environment and animal pop-
ulations when studying the effect of radioactive pol-
lution of a territory considered a factor in the
organisms' habitat.
The increase in the degree of changeability in
interior (the relative weight of the spleen and liver,
formed elements) and population indices (fertility
indices, life span, and the like) serves as criteria for
the degree to which pollution affects animals' bod-
ies, and the population. An increase in the degree of
changeability in exterior and craniological criteria,
155
-------
which are systematic criteria, may be indices of evo-
lutionary transformations taking place in the poplua-
tion.
An increase in the changeability of morphological
criteria in an animal population on a polluted terri-
tory shows that a change in the direction of natural
selection can take place, because certain genotypes
resistant to ionizing radiation will begin to predom-
inate numerically in the population. Natural selec-
tion under these conditions inevitably leads, in time,
to the occurrence of morphophysiological changes
in the population, on the way to adaptation of popu-
lations to radioactive pollution of a locality.
Effect on Zoocoenosis
Soil animals are one of the most convenient ob-
jects to use to study the effect of pollution on the
animal population in a locality, because they are
closely linked with the soil and vegetation cover,
are relatively immobile, and comprise 70-80% of
the total zoomass of land communities. Then too,
study of the effect of pollution on soil animals is of
definite interest. Soil animals deposit a goodly part
of the active salts in their bodies and participate in
soil-formation processes. The unfavorable effect of
pollution on soil animals can be shown to be unfavor-
able for the entire biogeocoenosis.
The numbers of mesofauna were reduced by a
factor of more than 2 in wooded sectors polluted by
strontium 90 in a density of 1.8-3.4 millicuries/m2.
The differences were not the same throughout the
warm season of the year. They reached a peak at
the end of summer, the reproductive period for most
of the groups of soil animals in the region under
study (Gilyarov, Krivolutskiy, 1971). Not all groups
of major invertebrates had their numbers reduced
equally in the polluted section, something particu-
larly characteristic of diplopods and earthworms.
Saprophages, the consumers of the debris layer, the
main component of the biogeocoenosis containing
strontium 90, sometimes were virtually absent in the
polluted sections. The number of predators de-
creased slightly on the whole, but sluggish groups,
the centipedes in particular, completely disappeared
from the population. The number of predatory bee-
tles remained almost unchanged. This probably can
be explained by the appearance of an additional
source of food for the predatory beetles in the form
of leaf-eating insects, a part of the life cycle, mov-
ing through the soil or plant litter. This additional
source of food probably compensates somewhat for
the direct action of radiation. Moreover, constant
increases in the beetle populations, at the expense
of surrounding sections, cannot be precluded. Noted
in the section polluted by strontium 90 is extensive
damage to birch leaves by leaf-eating insects, by
gypsy moth, caterpillars, sawfly larvae, and others,
in particular. In June 1969, 93.5% of the leaves
here were damaged, as opposed to 2.5% in the con-
trol. Radioactive pollution of the environment also
had an unfavorable effect on the development of the
entire soil profile by soil animals. The vertical dis-
tribution of animals on a mean annual basis showed
a general reduction in the soil population, particu-
larly that in the deep levels, which cannot help but
lead to weakening the soil formation rate.
Only the shelled mites of the microfauna in these
same sections were studied in detail. Their numbers
in the soils studied were 15-20 thousand specimens
per square meter, and could not be differentiated
reliably between the experimental and control sec-
tions. However, the variety of species of mites in
the polluted territory was definitely decreased, par-
ticularly in the case of the species inhabiting the soil
surface and the litter. Thus, what had appeared here
was a tendency toward simplification of the coenosis
in the polluted territory, even with respect to those
animals whose overall abundance remained un-
changed. This was particularly characteristic of the
species inhabiting the litter and the upper soil levels
(Krivolutskiy, Turchaninova, 1969). Observations
of ants are of interest in studying the effects of pol-
lution on complexes of land animals, because a con-
siderable percentage of the biomass is made up of
these animals, and their activities to a considerable
extent regulate the number of leaf-eating insects. It
is well known that ants react keenly to all forms of
man's economic activities in the forest (improve-
ment cutting, pasturage, burning of the litter, proc-
essing timber with agricultural chemicals). In the
case of long-term radioactive pollution, we were un-
able to detect any diminution in the ant population
in a quite large section in which the control and
experimental sections were, in all other characteris-
tics, practically identical, and which, prior to the in-
troduction of the radioisotope, were a single section
of the forest. It should be added that changes in the
ant population caused by various forms of man's
economic activities in the forest could be readily re-
corded in the research area by using the same pro-
cedure (Krivolutskiy, Baranov, 1972).
It is interesting that radioactive pollution of the
environment has a strong, well-recorded effect, even
on soil protozoa (Korganova, 1973), one of the
most radioresistant organisms in nature. This effect
can be explained only by the high level of radiation
in the water films surrounding the soil particles at-
tributable to the strontium 90 adsorbed by the soil.
BIBLIOGRAPHY
Gilyarov, M. S.; Krivolutskiy, D. A., 1971, "Radioecologi-
cal Research in Soil Zoology," Zool. zh. [Zoolosical
Journal], 50, 3, 329-342. *
ll'yenko, A. I., 1967, "The Biological Effect on a Popula-
tion of Voles of Pollution of the Soil Surface By Stron-
tium 90," In collection: Ekologiya mlekopitayushchikh i
ptits [The Ecology of Mammals and Birds], "Nauka"
Press, Moscow, 122-126.
Il'yenko, A. I., 1968, "The Effect of Pollution of a Section
of a Locality by Strontium 90 on Changeability in a
Population of Small Mammals," Zool. zh. [Zooloaical
Journal], 47, 9, 90-97. * al
156
-------
Il'yenko, A. I., 1971, "The Degree of Infestation of Popula-
tions of Small Mammals By Gamasid Mites in Sections
Polluted by Strontium 90," Zool. zh. [Zoological Jour-
nal], 50, 2, 243-246.
Korganova, G. A., 1973, "The Effect of Experimental Pol-
lution of the Soil on Soil Protozoa," Zool. zh. [Zoological
Journal], 52, 6, 939-941.
Krivolutskiy, D. A.; Baranov, A. F., 1972, 'The Effect of
Radioactive Pollution of the Soil on an Ant Population,"
Zool. zh. [Zoological Journal], 51, 8, 1248-1251.
Krivolutskiy, D. A.; Turchaninova, A. A., 1969, "Some
Changes in the Structure of the Animal Population of
Forest Soils Under the Effect of Pollution by Strontium
90," In book: Problemy pochvennoy toologii [Problems
of Soil Zoology], "Nauka" Press, Moscow, 207-209.
Krivolutskiy, D. A.; Smurov, A. V.; Snetkov, M. A., 1972,
"The Effect of Radioactive Pollution of the Soil by
Strontium 90 on Changeability of Certain Organisms,"
Zhurn. obshchey biol. [Journal of General Biology], 33,
5,587-591 .
Sokolov, V. Ye.; Il'yenko, A. I., 1969, "Radioecology of
Land Vertebrates," Uspekhi sovrem. biol. [Successes of
Modern Biology], 67, 2, 235-255.
157
-------
Effects of Pollution on Species and Populations of Fish
and Birds
Howard E. Johnson *
I. INTRODUCTION
In very recent time man has attained the ability
to cause major changes in his environment and that
of other living organisms. As our technology and
populations continue to grow human society has
become increasingly aware of environmental limita-
tions and our dependence on stable ecological sys-
tems.
The biota of an ecological system are the result
of long evolutionary development where balance
and equilibrium has been established between or-
ganisms and between these organisms and their
environment. The different species within these bio-
logical communities have evolved together but each
has developed rather specific physiological prefer-
ences and tolerances and they differ in many aspects
of their life cycles from the reproductive stage to
death. Frequently the species of greatest value to
man are those least tolerant to environmental fluc-
tuations or those highly restricted in their distribu-
tion in the environment. Even minor environmental
change may be detrimental to their productivity or
their very existence.
The strategy of pollution control to protect eco-
logical systems differs from that where human health
is a prime consideration. Rather than effects on the
individual organism our primary concern is for pro-
tection of natural populations, functional biological
communities or whole ecosystems. Relative to man,
most animal populations have high reproductive rates
and short life spans which produce a biological
excess. Therefore, losses of individuals are not sig-
nificant unless the fractional loss reduces potential
harvest by man or is selective for a particular seg-
ment of the population. This generalization does not
apply to species of birds or fish that have long life
spans and low rates of population turnover.
Although the goal of pollution control is directed
toward maintaining viable populations and the in-
tegrity of biological communities, our knowledge of
factors regulating such complex systems is deficient.
We can define the lethal threshold concentration of
a pollutant for an individual species, but the pol-
lutant concentration which will cause a major detri-
mental effect on a community or ecosystem is diffi-
cult to predict. Thus our immediate and most prac-
tical goal is to understand the effects of pollutants
on individual species that are known to have major
ecological significance or are of greatest economic
benefit to man. If environmental requirements for
these selected species are determined and appropriate
controls implemented, we will have made a first
major step in protection of the environment. Protec-
tion of these key species will also protect many other
species that have similar or greater environmental
tolerances.
Nevertheless the selection of beneficial or key spe-
cies according to their economic or social value is
disturbing to many ecologists. Some species not
protected by the standards may be eventually elim-
inated from some environments with a resultant
change in the biological community structure that is
detrimental to even desirable species. As we reduce
the diversity of an ecological system we often de-
crease its stability and internal regulation and re-
quire increased costs and energy in management.
While this may be appropriate for some forms of
agriculture or highly developed areas it is not feas-
ible for all regions.
Thus our approach to ecological problems should
necessarily be conservative. We should identify some
major regions where human activity is highly re-
stricted and retain these areas for the continued study
of natural communities and to develop a reservoir
of species that may be of major benefit to man in
the future.
But social and economic demands for develop-
ment of new regions require that pollution control
standards are developed now according to our best
scientific knowledge. Within any major region we
can identify biological communities and species that
are of major ecological or economic importance. Our
research and control efforts should be developed to
identify and protect these species. This includes de-
termination of critical pollution concentrations as
•Associate Professor, Department of Fisheries and Wildlife, Michigan State University
159
-------
well as the unique features of the environment that
are required for their survival; for example, fish
must have suitable conditions for spawning, migra-
tion, and food organisms.
II. ESTIMATION OF MAXIMUM
PERMISSIBLE LEVELS
The effects of most pollutants on an organism or
a population can be described as a continuum ex-
tending from concentrations that are lethal after
short term exposure to a level that has no adverse
effects even after continuous exposure through an
entire life cycle. Intermediate to these extremes, ef-
fects occur at different levels of biological organiza-
tion that may or may not have ecological significance.
Some changes merely reflect temporary or adaptive
responses of individuals within their normal physi-
ological range. Ideally, however, knowledge of the
full range of dose-effect is important for decisions of
cost-benefit analysis and to determine maximum per-
missible concentrations in the environment.
Lethal Concentration of Pollutants
Because of their simplicity short term toxicity
tests have been the method used most frequently to
measure pollutant effects on fish and birds. In aquatic
studies, fish are generally exposed to different con-
centrations of a toxicant for 4 days and the results
expressed as the concentration that is lethal to 50%
of the fish (96 hr LC-50). These data indicate the
relative toxicity of compounds or the sensitivity of
various species to a single compound, but the LC-50
data do not represent acceptable pollutant concen-
trations. It has been common practice to estimate
maximum permissible levels of a particular pollutant
by multiplying the LC-50 value by an application
factor such as 0.1 or 0.05. The value of the applica-
tion factor is arbitrary but it is usually based upon
scientific judgment and field evidence (Sprague,
1971).
In the United Kingdom, investigations have
shown an empirical relationship between the esti-
mated short term toxicity (48 hr LC-50) of mix-
tures of common pollutants and the presence or
absence of fish in certain rivers (Alabaster et al.,
1972). In separate investigations with rainbow trout
(Salmo gairdneri), Lloyd (1961), Herbert (1962),
Herbert and Shurben (1964) and others have found
the toxicity of some common pollutants is approx-
imately additive. The proportional toxicity of each
toxicant in the mixture is obtained by dividing its
concentration in the water by the estimated 48 hr
¥ _ _- Concentration of Toxicant A in Solution
LC-50, e.g.
48 hr LC-50 of Toxicant A
= the proportion of the lethal concentration of toxi-
cant A and similarly for toxicants B, C and D in the
mixture. The fractions for all toxicants in the mix-
ture are summed to give the proportion of the 48 hr
LC-50 for the mixture.
Using this technique in field studies, Herbert et
al. (1965) found generally good agreement between
predicted and observed toxicity in a number of fish-
less rivers in Great Britain. Alabaster et al. (1972)
reported good fisheries generally existed in waters
where the proportion of the 48 hr LC-50 was ap-
proximately 0.28 and marginal or no fisheries existed
where the proportion was higher than 0.28. Brown
(1968) reviewed the application of this method ac-
knowledging the lack of scientific basis for additive
toxicity of toxicants with different lexicological prop-
erties but suggested the various pollutants may each
produce a non-specific but additive overall stress.
Sub-lethal and Chronic Toxicity of Pollutants
There is general agreement among biologists for
the need to consider the effects of pollutants at all
life stages of an organism, particularly the effects on
growth and reproduction. Mount and Stephan (1967)
described a chronic toxicity test specifically designed
to estimate acceptable toxicant concentrations for
fish. The test provides for exposing groups of a single
fish species to different concentrations of a toxicant
through an entire life cycle in order to find the high-
est toxicant concentration that has no adverse effects
on survival, growth and reproduction of the species.
They suggest an application factor for the toxicant
can be calculated by dividing the "no effect" concen-
tration by the 96 hr LC-50. The application factor
could then be multiplied times the 96 hr LC-50 for
other species of fish in another water body to estimate
acceptable concentrations.
The hypothesis for this approach is that the ratio
of the "no effect" concentration and the 96 hr LC-50
of a particular compound does not vary greatly
among different species even though each species
may vary greatly in their sensitivity to the compound.
Examples of acute toxicity values for several pesti-
cides and heavy metals and the calculated application
factors are presented in Table 1. In particular the
data for malathion illustrate the close agreement of
calculated application factors for different species
that vary widely in their sensitivity to the insecticide.
This tends to support the hypothesis that chronic
toxicity data may have broad application to other
species.
The validity of predicted toxicity from chronic
toxicity tests in the laboratory requires verification
in field studies. Eaton (1973) reported that prelim-
inary results from studies of a natural stream inten-
tionally polluted with copper seem to demonstrate
the validity of predictions made from chronic toxicity
tests with copper in the laboratory.
160
-------
Table 1.
Acute and Chronic Toxicity Values and Calculated
Applications Factors for Some Heavy Metals and
Pesticides
Chemical
Fish Species
96-Hour
LC«o Application
(mg/U Factor*
Pesticides
Diazinon
Captan
2, 4-D
Butoxy-
etbanol
ester
Carbaryl
Methoxy-
chlor
Malathion
Lindane
Heavy Metals
Chromium +"
Copper
Cadmium
Methyl-
mercury
Fathead Minnow
Fathead Minnow
Fathead Minnow
Fathead Minnow
Fathead Minnow
Fathead Minnow
Bluegill
Brook Trout
Fathead Minnow
Brook Trout
Fathead Minnow
Brook Trout
Rainbow Trout
Fathead Minnow
Bluegill
Brook Trout
Fathead Minnow
Bluegill
Green Sunfish
Fathead Minnow
Brook Trout
6
0.065
5.6
9
0.0075
10.5
0.08
0.20
50
26
33
50
69
0.47
1.1
0.1
31
20
20
0.04
0.096
.0005
.10
.05
.023
.107
.02
.04
.02
.5
.38
.03
.01
.003
.03
.02
.09
.001
.0015
.0025
.006
.003
• The "application factor" is the ratio obtained by dividing experi-
mentally determined maximum concentrations of a pesticide that nave
no detectable effect during chronic exposures by the 96-hour LCm for
that species.
Short Term Indicators of Chronic Toxicity
A major criticism of long term tests to define
chronic toxicity or no effect levels is the long time
and expense required for the test. It is obvious that
such tests can be conducted with only a few of the
major pollutants and species that are important.
Although additional testing may show some data
have broad application, other tests of shorter dura-
tion are necessary to predict long term effects.
McKim et al. (1970) have recorded changes in
several characteristics of brook trout blood after
only six days exposure to copper concentrations that
eventually affected reproduction, growth, and sur-
vival of fish after 8 months exposure (Figure 1). In
subsequent tests, Drummond et al. (1973) recorded
changes in behavioral activity of brook trout (cough-
ing response, locomotor activity and feeding activity)
that appear to be useful as short term indicators of
copper concentrations that have no long term effects
on the species. Cairns et al. (1973) have similarly
reported that changes in locomotor activity and
breathing movements of fish can be monitored as
indicators of both lethal and sublethal concentra-
tions of zinc in water.
Although chronic toxicity tests may reveal pol-
lutant concentrations that affect reproduction, growth
and survival of fish the practical significance of dif-
ferent degrees of effect must be determined; i.e., how
important is 20% decrease in egg production or fry
survival. Jensen (1971) used a mathematical model
for yield of a brook trout population to examine the
predicted effect of decreased egg production or in-
creased mortality in "0" age groups, that occur due
to a toxic pollutant. His results indicate that even
a 5% increase in mortality of the "0" age group
results in a substantial decrease in yield (Figure 2).
More importantly his data indicate that an increased
mortality in the "0" age group would not be detected
as a significant change in the population for at least
two or three years. Thus in natural fish populations,
serious damage could occur due to effects on repro-
duction or survival in the early life stages long before
field monitoring could identify the change as dif-
ferent from natural fluctuation in population density.
These results emphasize again the importance of a
conservative approach to setting standards for pro-
tection of fish and wildlife populations. Although
threshold levels of effect can be identified, their
values are usually related to only a few species and
some allowance is probably necessary for protection
of other more sensitive species including food or-
ganisms.
III. EFFECTS ON BIOCIDES ON
BIRD POPULATIONS
Bird populations can be affected by the introduc-
tion of biocides and other toxic materials to the na-
tural environment. The species of birds and number
of individuals in a particular region usually remain
relatively steady year after year. A major objective
of wildlife biology is to predict changes in the popu-
lation as a result of observing changes in demo-
graphic parameters—fecundity and mortality—and
to determine the cause of these changes. These para-
meters are usually measured by the number of eggs
laid per clutch, frequency at which clutches are laid,
survivorship of eggs and young to the age of first
reproduction, and survival of adults throughout their
lifetime (Ricklefs, 1973). These statistics of birth
and death have been characterized for a hypothetical
population by Ricklefs (1973) to demonstrate the
relationship between recruitment, and adult mortality
as a function of population density (Figure 3). In
this example an annual mortality rate of 40 percent
suggests that the population density will increase
to a density of 125 individuals. As the annual mor-
tality rate increases the density will decrease and the
population will eventually become extinct if the an-
nual adult mortality is over 80 percent. Although
each bird species will have its own hypothetical
model, these types of data are necessary in order to
evaluate the effects of a toxic material that is released
into the natural environment.
161
-------
1 YEAR
6MONTH
1 MONTH
4 DAY
1 HOUR
DECREASE BLOOD ENZYME - TRANSAMINASE
DECREASE SURVIVAL-ADULTS
DECREASE VIABILITY AND
HATCHABILITY OF EGGS
100% MORTALITY FRY-OUVENILES
24% MORTALITY FRY^JUVENILES
GROWTH OF FRY RETARDED^
REVERSIBLE CHANGES
IN BLOOD FACTORS
REVERSIBLE CHANGES
IN BLOOD FACTORS
|96 HOURS. LC-60UUVENILES)|
I INCREASED COUGH FREQUENCY]
J'. ^
I REDUCED FEEDING ACTIVITY]
" ^X*
•T L INCREASED LOCOMOTOR ACTIVITY]
_ ^^^^^^^^^^^^^•^^•W"<^"*a*«^^^W^HVBBMIIMIWI^^^v^>^^M^^Hpv|
10
20
30
40
50
60
70
80
90
100
COPPER CONCENTRATION
(jig/ liter)
SOURCE: (DATA FROM McKIM et. al., 1970: McKIM AND BENOIT 1971,
AND DRUMMOND et. al., 1973).
FIGURE 1. SOME EFFECTS OF DIFFERENT COPPER CONCENTRATIONS
AND EXPOSURE TIMES ON BROOK TROUT (SALVELINUS
FONTINALIS).
162
-------
1000
0% INCREASE IN MORTALITY
01234
GENERATION
SOURCE: A.L. JENSEN, TRANS. AMER. FISH SOC. 100 (3): 466-469; 1971.
5
NUMBER
FIGURE 2. YEARLY YIELD FROM BROOK TROUT POPULATION WITH
ADDITIONAL MORTALITY IMPOSED ON THE "O" AGE GROUP.
ADULT MORTALITY RATE
20
40
120
60 80
ADULT POPULATION SIZE
SOURCE: R.E. RICKLEFS. NATIONAL ACADEMY OF SCIENCES, WASHINGTON, D.C.; 1973
FIGURE 3. THE RELATIONSHIP BETWEEN ADULT MORTALITY AND RECRUITMENT
AS A FUNCTION OF POPULATION DENSITY.
163
-------
Most biological factors influencing recruitment
and mortality in bird populations are related to the
density of individuals. As density increases, mortal-
ity rates tend to increase and the rate of recruitment
decreases. The reverse is true as density decreases.
However, the effect of pollutants on bird populations
is usually independent of density and mortality will
continue despite a decrease in numbers. Survival is
possible only if the population develops genetic re-
sistance to the pollutant. Although there are no
definite examples of genetic resistance to pollutants
in birds, Ford and Prince (1973) have demon-
strated a wide range of genetic variability among
individual mallards (Anas platyrhynchos) to the
effects of methylmercury chloride on the production
of ducklings. Analysis for genetic variability has not
been included in most evaluations of toxic materials,
but should probably be included in future tests to
improve predictions on the potential effects of pol-
lutants on birds.
Different bird species vary widely in their response
to specific pollutants, but usually species within a
taxonomic order show a more uniform response than
those between orders. Standard test species in the
United States are mallards and ring-necked phesants
(Phasianus colchicus) (Tucker and Crabtree, 1970).
These species are common, widespread and im-
portant economically in this country.
Most pollutants are present in combination in the
natural environment and present a potential for syn-
ergistic effects. Many of these effects will alter re-
cruitment rates at lower concentrations than adult
mortality rates, Thus data on reproduction is neces-
sary to evaluate the environmental impact of most
pollutants. A discussion of a specific pollutant, me-
thylmercury, can best illustrate these considerations.
The wide use of mercury and mercury compounds
in agriculture and industry has increased natural
levels of mercury in certain regions. In Sweden start-
ing about 1955, a decrease in the popoulation of
seed-eating birds was attributed to the use of methyl-
mercury compounds as seed-dressing agents. A
similar instance of widespread mercury contamina-
tion in the Canadian prairie and depletion in the
abundance of seed-eating birds, including pheasants,
and their avian predators has been reported by
Fimreite et al. (1970). A number of investigators
have pointed out that a reduction in the reproductive
potential could theoretically occur if pheasants con-
sumed seed grain treated with alkylmercury com-
pounds (Borg et al., 1969; Fimreite et al., 1970;
Fimreite, 1971; Adams and Prince, 1972; Spann et
al., 1972). Although only indirect evidence is avail-
able, this predicted decline occurred in Sweden, the
Canadian prairies, and it may have been responsible
for declines in pheasant populations in other parts of
the range in North America. Removal of the alkyl-
mercurials as seed dressing has apparently corrected
these problems. A 1966 ban in Sweden resulted in
an immediate drop in the merucury levels in the
tissues of seed-eating birds and a corresponding in-
crease in the number of birds (Borg et al., 1969;
Johnels and Westermark, 1969; Jernelov, 1972).
When other seed-dressing agents such as sub-
lethal daily doses of captan, dieldrin, diazinon, cap-
tan plus dieldrin, and captan plus diazinon are fed to
laying pheasants reproductive declines are also ob-
served (Stromborg and Prince, 1974). They found
that decreased egg production was the most signifi-
cant effect with dieldrin and captan acting ad-
ditively, while captain and diazison showed signifi-
cant synergism.
IV. PERSISTENT POLLUTANTS
Some pollutants are of particular concern because
of their wide dispersal in the environment and their
tendency to accumulate in the food chain with po-
tential adverse effects at high tropic levels including
man. Most chemicals of this nature are synthetic
organic compounds developed in the past few
decades. The chlorinated hydrocarbon insecticides
such as DDT, dieldrin, and heptachlor are examples
but a number of other compounds widely used in
industry also have properties of persistence, mobility
and bioaccumulation. The polyclorinated biphenyls
(PCB's) have been used in industry since about 1930
but their wide distribution and ecological hazard was
not recognized until the late 1960's (Peakall and
Lincer, 1970).
PCB's are used in a variety of products and manu-
facturing processes including synthetic resins, hy-
draulic fluids, lubricants and transformer fluids. Their
loss to the environment has probably occurred
through direct discharge of some fluids and oils
directly to water, loss from dumps and incomplete
burning in incinerators, and vaporization to the at-
mosphere. Through transport in the atmosphere and
in rivers, PCB's have become distributed throughout
the biosphere, and accumulated in various concen-
trations in invertebrates, fish and birds (Nisbet and
Sarofim, 1972).
PCB's like other chlorinated hydrocarbons have
a very low water solubility (less than 1 mg/liter).
In surface waters PCB's have been reported at con-
centrations of parts per billion in a few highly in-
dustrialized rivers but most measurements indicate
ambient levels of 10 to 120 parts per trillion.
Acquatic invertebrates and fish can accumulate PCB
to levels of 103 to 105 times those in the ambient
water and sediments. Because these compounds are
resistant to degradation they can be passed through
the food chain to higher trophic levels. Additional
accumulation occurs in birds and mammals associ-
ated with the aquatic environment to concentrations
164
-------
that are 107 to 10s times higher than ambient water
concentrations (Nisbet and Sarofim, 1972; Stalling
and Mayer, 1972).
The concentrations of PCB in fish and birds indi-
cate high levels of contamination in industrial
regions but low levels are widespread in other areas.
Fish in some highly contaminated waters of the U.S.
have concentrations ranging from 10-800 ppm but
average concentrations in fish from various bays and
estuaries or unpolluted rivers are 1 ppm or less
(Panel on Hazardous Trace Substances, 1972).
The PCB's are generally considered to be only
moderately toxic to animals with short term ex-
posure. However, fish appear to obtain their greatest
uptake directly from water and this appears to pre-
sent a greater hazard than dietary sources. Due to
their low solubility in water short-term tests of 96
hours or less do not adequately reflect their toxicity
to fish or other aquatic organisms. The toxicity of
the various PCB's varies between species but for 15
day exposure ranges between 50—30 /ug/1 PCB in
water. With longer exposure the LC-50 values are
decreased significantly (Stallings and Mayer, 1972).
A pollutant which can be concentrated in organisms
and is more toxic at longer exposure concentration
presents an increased hazard to reproduction and
growth. Some tests with fathead minnows (Pime-
phales promelas) indicate mortality of adults and
reproductive failure after continuous exposure to
concentrations of approximately 2 to 8 ppb PCB in
water.
In the Great Lakes, the eggs of salmon and trout
were reported to have concentrations as high as 12
to 17 ppm PCB (fresh weight) and may be partially
responsible for the poor reproductive success of these
species in Lake Michigan (Veith, 1969; Johnson and
Ball, 1972). Johannson et al. (1970) reported high
losses in prehatching salmon eggs that contain PCB
residue of 14 to 34 ppm (lipid weight).
Other indirect effects of PCB on fish and bird
populations have been reported. Hansen et al.
(1971) found a high incidence of fungus infection
in fish exposed to PCB in laboratory and Duke et
al. (1970) found similar lesions in wild fish from an
estuary grossly polluted with PCB. Ducks exposed
to sublethal concentrations of PCB exhibited in-
creased susceptibility to viral hepatitis (Friend and
Trainer, 1970).
The difficulty of assessing the effect of a chemical
pollutant on wild populations is compounded by the
simultaneous occurrence of other pollutants or a
change in some other environmental factor. High
concentrations of PCB have been found in some wild
bird populations that are predators on fish. Although
these populations have experienced reproductive
failure and population declines, the effect of PCB's
is not clear because of the presence of DDT, mercury
and other toxins. Where compounds affect reproduc-
tion, growth and behavior of organisms at concen-
trations below the acute lethal level, the effect may
be expressed as a slow change in the size or relative
abundance of a population. At present, levels of PCB
in many parts of the environment occur at concen-
trations that exceed levels known to have adverse
effects on organisms in the laboratory. Therefore,
it is likely that adverse effects are occurring in popu-
lations or communities but it is very difficult to show
this conclusively.
Many new chemicals which are potentially hazard-
ous substances are introduced into commerce an-
nually. Strategy for control of these pollutants must
be directed at identification and testing for harmful
environmental effects before they are used in large
quantities or widely distributed in the environment.
Many of our present techniques for testing the toxic-
ity of new compounds are inadequate for evaluation
of their environmental impact. Consideration must
be given to testing in more sophisticated model eco-
systems and biological communities where rates of
transport, degradation, bioaccumulation and trophic
transfer can be evaluated.
Models to describe sources, transport pathways
and effects within global and regional environmental
areas are needed to define monitoring schemes. An
example of such a descriptive model is shown in
Figure 4. Such models may have broad application
for predicting movements and effects of many com-
pounds that have similar properties and distribution
in the environment. Perhaps more than any other
form of pollution the problems associated with per-
sistent synthetic compounds emphasize the import-
ance of a conservative approach in pollution con-
trol.
V. MULTIPLE EFFECTS OF
POLLUTANTS
Although we often investigate the individual effects
of pollutants on an organism, it is evident that effects
in environmental systems are usually multiple and
highly complex. Attempts to set maximum permis-
sible levels of a pollutant must consider the full
range of environmental interactions that may occur.
Temperature is an example of an environmental
factor that has multiple effects on aquatic systems.
Heated water discharge for electrical power plants is
expected to increase in the next 30 years with sub-
stantial quantities of our surface waters used in the
cooling systems. Individual heated discharges may
have only minor effects on the environment but col-
lectively several discharges in close proximity may
create major environmental changes.
Individual aquatic species have specific thermal
requirements that differ according to their life stages
and influence their success in the environment.
165
-------
HUMAN
FISHING
DRINKING
WATER
IMMIGRATION
EMIGRATION
IMMIGRATION
EMIGRATION
FALLOUT
HIGHER CARNIVORES
PREDATION
FIRST CARNIVORES
PREDATION
PREDATION
HERBIVOROUS
FISH
PREDATION
IZOOPLANKTON
GRAZING
PLANTS
GRAZING
DIRECT
UPTAKE
LEACHING
SEWAGE
WATER
DREDGING
DUMPING
EXCRETION
FILTERING
BOTTOM ORGANISMS
UPTAKE
SEDIMENT
EXCRETION AND DECAY
SOURCE: ENV.RONMENTAL RESEARCH 5:249-362; 1972
CODISTILLATION-^
RIVER FLOW _
IRRIGATION
DECOMPOSITION
FIGURE 4. MODEL OF INPUT, TRANSPORT AND FATE OF PCB
IN THE FRESHWATER ENVIRONMENT
166
-------
Heated discharges can be managed to promote cer-
tain species if these requirements are known and
protected. The maximum and minimum tolerance of
desirable organisms may dictate the allowable dif-
ferential between heated discharge and ambient water
temperature. Thermal stress which is not lethal can
have detrimental effects by increasing an organism's
susceptibility to predation (Coutant, 1969). Each
species also has some physiological optimum tem-
perature for growth. However, as temperature in-
creases food requirements also increase (Brett et al.,
1969) and eventually a temperature is reached where
growth ceases because of the high metabolic demand.
Species that are forced to live near these marginal
temperature extremes may be competitively excluded
by species of greater thermal tolerance.
Fish in temperate climate zones require a seasonal
fluctuation in temperature or winter chill period for
successful development of the gametes and spawning.
A period of low temperature is apparently required
to allow storage of yolk material which cannot be
effected successfully at higher temperatures because
storage of food material is prevented at the higher
metabolic rate (Merriman and Schedl, 1941). If
fish are attracted to and remain in elevated water
temperatures during the winter months they may
suffer reproductive failure.
Other effects of temperature that may indirectly
affect populations of aquatic organisms result from
interaction with other pollutants. Macek (1969) re-
ported insecticide toxicity to fish was increased at
higher temperatures and Edsall and Yocum (1972)
reported lake trout (Salvelinus namaycush) doubled
their rate of methylmercury accumulation from water
when the water temperature was increased from 5 °C
to 10°C. Pretreatment or exposure to a toxicant may
reduce the tolerance to thermal stress in fish (Silber-
geld, 1973).
An increased incidence of various fish diseases
has been associated with elevated water tempera-
tures in various locations (Mihursky, 1969; Brett,
1956). Ordal and Pacha (1967) found increased
virulence and effect of several diseases on salmonids
held at increased water temperature.
There is also some indication that heated water
discharges may accelerate the process of eutrophica-
tion in lakes and create conditions favorable for
nuisance algae (Patrick, 1969).
Temperature of course is only one of the many
environmental variables that cause multiple or "syn-
ergistic" effects with other pollutants and different
organisms. Nearly any pollutant can have multiple
effects, many of which are not recognized until after
harmful effects have occurred. Wood, 1971 (cited by
Bahr, 1972) reports the interaction of stream and
lake enrichment by sewage effluent and the discharge
of inorganic mercury as an example. The large
bacterial populations created by sewage enrichment
act as an "environmental catalyst" in the formation
of the highly toxic methylmercury. Bacteria were
probably always present in natural waters and so
were small amounts of the element mercury. But in-
creased discharge of elemental mercury in the
presence of high bacterial populations allowed the
accelerated production of methylmercury. Fish, birds
and human populations have suffered from this un-
predicted combination of environmental events.
VI. POLLUTANT EFFECTS ON
BIOLOGICAL COMMUNITIES
The laboratory testing of pollutant effects with
important species has been recognized as a practical
and necessary method for development of immediate
pollution control standards. However, the limitations
of the data obtained for a single species under highly
artificial conditions is also recognized. An organism
exists in nature as an integral part of the biological
community, both exerting effects and being affected
by interactions within the system. Improved methods
for evaluating pollutant effects are needed to measure
population or community responses to toxicants or
other environmental perturbations. This is the next
logical step for understanding the significance of
effects observed at the species or individual organism
level.
Warren and Davis (1971) have described methods
of using laboratory stream ecosystems to measure
dynamics of primary and secondary production of
relatively simple communities of organisms. Even in
these simplified systems the interactions between
biotic and abiotic variables are complex, but much
greater control can be exercised than in truly natural
communities. The information obtained from these
simulated communities or ecosystems offer a new
level of sophistication for interpretation of environ-
mental changes.
Maki and Johnson (1974) have used a series of
model streams colonized by natural stream organisms
to assess the effect of a toxicant TFM (3-tri-fluoro-
methyl-4-nitrophenol) on community structure and
metabolism. TFM, which is used in streams to con-
trol populations of the parasitic sea lamprey (Petro-
myzon marinus), was introduced into streams under
simulated field conditions. Their data show the toxi-
cant caused immediate but reversible changes in the
composition and abundance of sensitive invertebrate
species and in the productivity of algal communities.
Using radiolabeled TFM in separate sections of the
stream they were able to measure the translocation
and accumulation of residues in various organisms
and the sediment. Although each of their measure-
ments could have been conducted in isolated systems
or with individual species, only by using a model
community were they able to measure the rate of
change and interaction of the different species.
167
-------
Model aquatic ecosystems have also been em-
ployed to study the uptake and translocation of
chemicals between several trophic levels. Isensee et
al. (1973) studied the distribution of radiolabeled
alkyl arsenicals in algae, snails, Daphnia, and fish in
a simulated food chain. Their data provide informa-
tion on the potential for bioaccumulation by the
various species and transfer to higher trophic levels.
Each of the above examples is an attempt to im-
prove on methods for assessing the effects of pol-
lutants in the environment. Undoubtedly new, more
precise methods will be developed. Although these
methods will be more expensive initially the results
should greatly improve our ability to predict and
control environmental problems.
VII. SUMMARY
Human society in its present form is highly de-
pendent upon stable biological communities that
form the basic structure of our life support systems.
Our failure to recognize the integrated nature of
ecological systems and man's dependence upon them
has led to many of our present environmental prob-
lems.
Natural biological communities in a region consist
of populations of organisms that have many common
attributes. These similarities form a common de-
nominator necessary for effective function and regu-
lation of ecological systems. An immediate and prac-
tical need for pollution control is to recognize and
protect the environmental requirements of individual
populations of species within these regions. However,
this must be considered a limited first step and much
greater effort should be given to understanding the
interrelationships between populations and communi-
ties.
Regional resource management plans offer an op-
portunity to view entire biological communities or
ecosystems as functional units. Even at this level
of organization we can err in management practices
if we consider such regions as independent units.
Ecological systems do not have distinct boundaries
and change in one region will often effect changes in
another.
Fish and wildlife populations are only one com-
ponent of biological communities that we wish to
maintain. However, in addition to their social and
economic value, species of fish and birds often serve
as important indicators of environmental quality.
Their role and general productivity within a region
are indicative of change that may be either to the
benefit or detriment of human society.
ACKNOWLEDGEMENTS
I extend my appreciation to the following people
for their assistance in preparation of the manuscript:
Dr. Harold Prince, Dr. Walter Conley, Mr. Alan
Tipton, Ms. Judy Boger and Ms. Jackie Church.
LITERATURE CITED
Alabaster, J. S., J. H. N. Garland, I. C. Hart and J. F. DC
L. G. Solbe. 1972. An approach to the problem of pol-
lution and fisheries. Symp. Zool. Soc. Lend. No. 29: 87-
114.
Adams, W. J. and H. H. Prince. 1972. Survival and repro-
duction of ring-necked pheasants consuming two mer-
curial fungicides. P. 306-317 in Environmental Mercury
Contamination, R. Hartung and B. D. Dinman, Editors.
Ann Arbor Science Publishers, Inc., Ann Arbor, Michi-
gan, U.S.A. 349 pp.
Bahr, T. G.( R. A. Cole and H. K. Stevens. 1972. Recycling
and ecosystem response to water manipulation. Tech.
Rept. No. 37: 124 pp. Institute Water Research, Mich.
State Univ., East Lansing, Mich. 48824.
Borg, K., H. Wanntorp, K. Erne, and E. Hanko. 1969.
Alkyl mercury poisoning in terrestrial Swedish wildlife
Viltrevy 6(4): 301-379.
Brett, J. R. 1956. Some principles in the thermal require-
ments of fishes. Quart. Rev. Biol. 31(2): 75-87.
Brett, J. R., J. E. Sbelbourn, and C. T. Shoop. 1969. Growth
rate and body composition of fingerling sockeye salmon,
Oncorhynchus nerka in relation to temperature and ration
size. J. Fish. Res. Bd. Canada 26(9): 2363-2394.
Brown, V. M. 1968. The calculation of the acute toxicity
of mixtures of poisons to rainbow trout. Water Research
2: 723-733.
Cairns, J., Jr., R. E. Sparks, W. T. Waller. 1973. The rela-
tionship between continuous biological monitoring and
water quality standards for chronic exposure P. 383-402
in Bioassay Techniques and Environmental Chemistry,
G. E. Glass, Ed. Ann Arbor Science Publishers, Inc.,
Ann Arbor, Michigan, U.S.A.
Coutant, C. C. 1969. Temperature, reproduction and be-
havior. Chesapeake Science 10(3&4): 261-274.
Drummond, R A., W A. Spoor, and G. F. Olson. 1973.
Some short term indicators of sublethal effects of copper
on brook trout, Salvelinus fontinalis. J. Fish. Res. Bd
Canada 30: 698-701.
Duke, T. W., J. I. Lowe, and A. J. Wilson, Jr. (1970).
A polychlorinated biphenyl (Aroclor 1254) in the water,
sediment, and biota of Escambia Bay, Florida. Bull. En-
viron. Contam. Toxicol. 5: 171-180.
Eaton, J. G. 1973. Recent developments in the use of lab-
oratory bioassays to determine "safe" levels of toxicants
for fish. PP. 107-115, in Bioassay Techniques and En-
vironmental Chemistry, G. E. Glass, Ed. Ann Arbor
Science Publishers, Inc., Ann Arbor, Michigan, U.S.A
Edsall, T. A. and T. G. Yocum. 1972. Review of recent
technical information concerning the adverse effects of
once-through cooling on Lake Michigan. Lake Michigan
Enforcement Conference, September 19-21, 1972, Chi-
cago, Illinois. 86 pp.
Fimreite, N. 1971. Effects of dietary methylmercury on
ring-necked pheasants. Canadian Wildlife Service Occa-
sional Paper No. 9. 39 pp.
Fimreite, N., R. W. Fyfe, and J. A. Keith. 1970. Mercury
contamination of Canadian Prairie seed eaters and their
avian predators. Canadian Field Naturalist 84(3): 269-
276.
Ford, J. R. and H. H. Prince. 1972. Effects of methylmer-
cury chloride on mallard reproduction. Presented at 1972
Annual Wildlife Disease Conference, June 13-15, Uni-
versity of Michigan, Ann Arbor, Mich.
Ford, J. R. and H. H. Prince. 1973. Distribution and excre-
tion of methylmercury chloride in mallards. Presented at
35th Midwest Wildlife Conference, December 2-5, 1973,
St. Louis, Missouri.
Friend, M. and D. O. Trainer. 1970. Polychlorinated bi-
phenyl: Interaction with duck hepatitis virus. Science
170: 1314-1316.
Hansen, D. J., P. R. Parrish, J. I. Lowe, and P. D. Wilson.
1971. Chronic toxicity, uptake, and retention of Aroclor
168
-------
1254 in two estuarine fishes. Bull. Environ. Contain, and
Toxicol. 6(2): 113-119.
Herbert, D. W. M. 1962. The toxicity to rainbow trout of
spent still liquors from the distillation of coal. Ann.
Appl. Biol. 50:755-777.
Herbert, D. W. M. and D. S. Shurben. 1964. The toxicity
to fish of mixtures of poisons. I. Salts of ammonia and
zinc. Ann. Appl. Biol. 53: 33-41.
Herbert, D. W. M., D. H. M. Jordan and R. Lloyd. 1965.
A study of some fishless rivers in the industrial midlands.
J. Proc. Inst. Sew. Purif., pp. 569-582.
Isensee, A. R., P. C. Kearney, E. A. Woolson, G. E. Jones,
and V. P. Williams. 1973. Distribution of alkyl arsenicals
in model ecosystem. Environ. Sci. and Tech. 7(9): 841-
845.
Jensen, A. L. 1971. The effect of increased mortality on the
young in a population of brook trout, a theoretical anal-
ysis. Trans. Amer. Fish. Soc. 100(3): 456-459.
Jemelov, A. 1972. Mercury and food chains. P. 174-177 in
Environmental Mercury Contamination, R. Hartung and
B. D. Dinman, Editors. Ann Arbor Science Publishers,
Inc., Ann Arbor, Michigan, U.S.A. 349 pp.
Johnson, H. E. and R. C. Ball. 1972. Organic pesticide pol-
lution in an aquatic environment, pp. 1-10 in Fate of
Organic Pesticides in the Aquatic Environment, Ad-
vances in Chemistry Series, No. 111.
Johnels, A. G. and T. Westermark. 1969. Mercury contam-
ination of the environment in Sweden. P. 221-241 in
Chemical Fallout, M. W. Miller and G. G. Berg, Editors.
Charles C. Thomas, Springfield, Illinois. 531 pp.
Lloyd, R. 1961. The toxicity of mixtures of copper sulfate
to rainbow trout (Salmo gairdneri Richardson). Ann.
Appl. Biol. 49: 535-538.
Maki, A. W. and H. E. Johnson. 1974. Effects and fate of
TFM (3-trifluoromethyl-4-nitrophenol) lampricide in
model streams. Presented at 24th Annual Meeting, Mid-
west Benthological Society, Cincinnati, Ohio, March 27-
29, Cincinnati, Ohio.
McKim, J. M. and D. A. Beneir. 1971. Effects of long-
term exposure to copper on survival, growth and repro-
duction of brook trout (Satvelinus fontinatis). J. Fish.
Res. Bd. Canada 28(5): 655-662.
McKim, J. M., G. M. Christensen, and E. P. Hunt. 1970.
Changes in the blood of brook trout (Salvelinus fonti-
nalis) after short term and long term exposure to cop-
per. J. Fish. Res. Bd. Canada 27: 1883-1889.
Merriman, D. and H. P. Schedl. 1941. The effects of light
and temperature on gametogenesis in the four-spined
stickleback, Apeltes quadracus (Mitchell). J. Exp. Zool.
88(3): 413-449.
Mihursky, J. A., A. J. McErlean, and V. S. Kennedy.
1970. Thermal pollution in aquaculture and pathobiology
in aquatic systems. J. Wildl. Dis. 6: 347-355.
Mount, D. I. and C. E. Stephan. 1967. A method for estab-
lishing acceptable toxicant limits for nsh-mamathion
and the butoxyethanol ester of 2. 4-D. Trans. Amer
Fish. Soc. 96: 185-193.
Nisbet, I. C. T. and A. F. Sarofim. 1972. Rates and routes
of transport of PCB's in the environment. Environ. Health
Perspectives 1:21-38.
Ordal, E. J. and R E. Pacha. 1967. The effects of tempera-
ture on disease in fish. In Water Temperatures: influ-
ence, effects and control. Proc. 12th Pac. Northwest
Symp. Water Pollut. Res., Portland, Oregon. Fed. Wat.
Pollut. Contr. Admin.
Panel on Hazardous Trace Substances Report 1972.
PCB's—Environmental Impact. Environmental Research
5:249-362.
Patrick, R. 1969. Some effects of temperature on fresh-
water algae, pp. 161-185. In Biological Aspects of Pol-
lution, P. A. Krenkel and F. L. Parker, Editors, Vander-
bilt University Press. 407 pp.
Peakall, D. B. and J. L. Lincer. 1970 Polychlorinated bi-
phenyls: Another long-life widespread chemical in the
environment. BioScience 20: 958.
Ricklefs, R. E. 1973. Fecundity, mortality and ayian dem-
ography, p. 366-435. In Breeding Biology of Birds, D. S.
Farner, Ed. National Academy of Sciences, Washington,
D.C. 515 pp.
Silbergeld, E. K. 1973. Dieldrin. Effects of chronic sublethal
exposure on adaptation to thermal stress in freshwater
fish. Environ. Sci. and Tech. 7(9): 846-849.
Spann, J. W., R. G. Heath, J. F. Kreitzer, and L. N. Locke.
1972. Ethylmercury p-tolune sulfonanilide: lethal and
reproductive effects on pheasants. Science 175 (4019):
328-331.
Sprague, J. B. 1971. Measurements of pollutant toxicity to
fish-Ill. Sublethal effects and safe concentrations. Water
Research 5: 245-266.
Stalling, D. L. and F. L. Mayer, Jr. 1972. Toxicities of
PCB's to fish and environmental residues. Environ. Health
Perspectives 1: 159-164.
Stromborg, K. D. and H. H. Prince. 1974 Pesticide inter-
actions: Alteration of reproductive performance in pheas-
ants fed combinations of seed treatments. (In press).
Tucker, R. K. and D. G. Crabtree. 1970. Handbook of
toxicity of pesticides to wildlife. U.S. Dept. of Interior,
Bureau of Sport Fisheries and Wildlife Resource Publica-
tion No. 84. 131 pp.
U.S. Environmental Protection Agency. 1973. Effects of
pesticides in water-a report to the States. USEPA, Wash-
ington, D.C. 20460.145 pp.
Veith, G. D. 1972. Recent fluctuations of chlorobiphenyls
(PCB's) in the Green Bay, Wisconsin, region. Environ-
mental Health Perspectives 1: 51-54.
Warren, C. E. and G. E. Davis. 1971. Laboratory stream
research: objectives, possibilities and constraints. Ann.
Rev. Ecol. Syst. 1: 111-144.
169
-------
The Influence of Environmental Factors on
Population Dynamics (Mathematical Model)
M. Ya. Antonovskiy *
1. We shall, in this report, discuss the following
thesis. A population is a complex system containing
a great many different heterogeneities that do, and
do not, appear directly. At the same time, the ob-
served, relatively stable, regime in which a population
functions is the resultant of the effects of many
factors, a number of which can cause instability.
The effect of the environment on a population can
remove it from its stable regime and can bring into
play internal mechanisms embedded in the popula-
tion which can lead to the fundamental restructuring
of the population, even when the magnitude of the
external effect is small (more precisely, when there
is a small change in the level of the effect of the en-
vironment on the population).
We shall, in this report, consider simple mathe-
matical models of the different processes that can
take place in "theoretical" populations that will, in
our opinion, quite graphically illustrate the different
aspects of the thesis we have set forth.
If we are to apply the results we obtain to real
populations, it goes without saying that we must con-
struct a model that will take into consideration the
specificity of the interaction between the different
factors in these populations. The appearance of this
specificity can be expressed in the form of the corres-
ponding quantitative dependencies constructed on
the basis of a statistical analysis of the observed data.
However, the difficulties that can arise when at-
tempting to determine the precise form of the quan-
titative dependencies in biological systems are well
known. Consequently, the approach to their study,
which A. N. Kolmogorov [1] has defined as "obtain-
ing quality results from quality prerequisites," is an
extremely valuable one.
2. First of all, let us consider the phenomenon of
the influence of the environment on the nature of the
appearance of the internal mechanisms that depend
on density, and which regulate the dynamics of popu-
lation number and structure.
Study of the influence of density on population
dynamics, experimentally, as well as using mathe-
matical models, began a very long time ago.
•Central Institute of Economic Mathematics, Moscow
The logistic equation
dN/dt = N(a-bN)
(1)
was considered for the first time in connection with
study of population dynamics by Fehrhulst in 1844,
and led to the conclusion that there was a stationary
level for the number to which the number in a pop-
ulation would tend monotonically. This fact is a
classical example of negative feedback.
This conclusion retains its force, even when one
must solve the more general equation
dN/dt = Nf(N)
(2)
where
f (N) is a monotonically decreasing function,
change in the sign of which has no
effect on the conclusion reached as to
the monotonic tending to some sta-
tionary level.
In fact, there are four types of number dynamics for
populations functioning in a steady-state (in terms
of time) environment that can occur (and they can
be observed). These are:
(1) there is a steady stable state to which the
system tends monotonically;
(2) there is a steady stable state to which the
system tends in the form of attenuating oscillations;
(3) there is a stable, periodic, regime. At the
same time, the magnitude of the period can be dis-
associated from the values for certain of the biologi-
cal characteristics of the population (sexual matura-
tion period, life span, and the like);
(4) the number undergoes irregular oscillations.
A discussion of this subject, as well as a great deal
of experimental data, is contained in the basic mono-
graph written by Andrewartha and Birch [2]. The
book by Makfed'yen [3], while shorter, is full of in-
formation on this subject.
Population heterogeneity (and age differentiation
in particular) is the most obvious candidate for the
role of the cause of the effects listed.
However, at this point let us turn our attention to
other possible causes. It may well be that the same
171
-------
population can experience all the types of dynamics
listed, depending on the intensity of the effect of
some external factor, such as a chemical or a phy-
sical agent, for example, that could have an effect
on the fertility of females, or on the survival rate
of specimens. Another example of an external factor
is the degree of industrial exploitation of the popula-
tion.
This can be demonstrated on a model that does
not take population heterogeneities into considera-
tion.
Let us illustrate what we have said by a well
known example. Let a population consist of speci-
mens that have reached sexual maturity at year 1,
arid let the natural mortality and the fertility of the
females be independent of age. Further, let the num-
ber of additions to the population per mature speci-
men be expressed by the monotonically decreasing
function v(N,/t). This function depends on the num-
ber of mature specimens, N, as well as on the para-
meter p., characterizing the influence of the environ-
mental factor.
In this case the population dynamics equation has
the form
A<>— Ae
1/x
N(t + 1) = ( 1 - \)N(t) + N(t)
(3)
Here X is the mortality factor.
We shall not, here, dwell on the formulation of
the general theorems that can be proven naturally
with models of this type, but instead will consider the
concrete, often encountered, example where
where
(4)
c is some constant characterizing the eco-
logical capacity of the environment.
0 N
FIGURE 1. MODEL POPULATION FUNCTION
Eq. (4) makes it clear that the number of addi-
tions increases with increase in p..
Investigation of this model permits us to draw the
following conclusions:
Extinction
Monotonic growth Case 1
to stable steady
state
Tending to steady Case 2
state with attenu-
ating oscillations
Periodic or non- Cases 3
periodic regime and 4
We should point out that when /i>Ae2/x, the
steady state will become unstable and the type of
dynamics will depend on /x, as well as on the original
number in the population (as distinguished from
cases 1 through 3). The type of dynamics will, for
example, depend on the state in which the population
was found when the parameter /x took a new value.
There is a stable periodic regime for certain values
of p in which the population will be found for the
"majority" of original conditions. Although the set of
such original conditions for which the population
will not be found in this regime is a small one, it has
a measure of zero, the role of such conditions does
manifest itself in the fact that in the case of other
original conditions the reaching of the stable periodic
regime is preceded by a more or less lengthy period
(one that depends on the original conditions) of ir-
regular oscillations "simulating" these unstable re-
gimes. The length of the period will change in a most
complex fashion with increase in /x. There are certain
11 values at which a phenomenon corresponding to
case 4 will occur, and this we can call the ergodicity.
The period of the stable cycle will become very large
when the xi values are close to this, and periodicity
will in fact disappear.
Let us consider in detail what happens when the
parameter /*. passes through the critical value Ae2/x.
A trajectory with a period of 2(N,, N2, Nu N5, . . .),
which has the following properties, will appear in
the neighborhood of the steady state N= 1/c In LI/A:
(a) it is locally stable; that is, for any original
value N(O) sufficiently close to Na, N(2t)-»N,,
N(2t+l)->N2, when t-*°o;
(b) at the same time there are those original
values N(O), the numbers for which N(t), do not
tend to this stable trajectory; for example, N(O)
= l/c In /x/A;
(c) let us designate by A the set of those original
values N(O) for which t exists such that
N(t) = l/cln tiA.
Then A is an infinite denumerable set. The situation
described in (b), above, will occur for any value
N(O) of the closure A of set A. The trajectory will
tend to the original trajectory of period 2 for any
other_ original values. It is curious to note that the
set A is a "perfect Cantor set" of measure zero.
Similar effects are retained in more complicated
models, in those closer to reality, in which the age
172
-------
structure of the population, change in fertility with
age, and so on, are taken into consideration.
These models show that predetermined levels of
effect of a hypothetical external factor are critical
so far as the population is concerned; that is, the
population will change its type of behavior when such
value is reached.
There can be a situation in which the presence
in the environment of an important ingredient, one
however that is difficult to measure by using direct
methods, can be established from the behavior of
"indicator" populations that had been well-studied in
laboratories, definite species of micro-organisms, or
plants, for example.
We should point out that in recent years many of
the published papers have considered population dy-
namics models based on type 4 relationships, or
similar models with divisions into a small number
of age classes. We can, by way of an example, point
out reference [4], which deals with humpback salmon
population, and in which a critical value was found
at which the "natural" regime loses stability and in
which computer data are cited for the case corre-
sponding to the last row in our table.
3. Let us now go to the influence of the environ-
ment on a population that is the result of the effect
on the genetic mechanism. The first phenomenon we
shall analyze is the following. A genetically hetero-
geneous population in a steady (in some particular
sense) state can be subjected to the effect of environ-
mental factors that act on specimens of different
genotypes in a differentiated manner. In this case the
environmental factors act as a selection mechanism,
and the rate at which this selection is made can prove
to be extremely high. However, if, at the same time
the influence of density on population dynamics still
is taken into consideration, it can be shown that
even the presence of a small advantage, obtained by
a definite part of the population as a result of the
effect of an environmental factor, can, with the
passage of time, lead to a fundamental rearrange-
ment of the genetic structure of the population (dis-
appearance of certain genotypes, and the like).
Let us consider the following simple model in
order to explain this thesis.
Let us assume the population to be genetically
heterogeneous and consisting of two species (pheno-
types). These phenotypes can be distinguished by a
character that can be determined by one gene, found
in two variants, a dominant A and a recessive a.
Crossing of them takes place in accordance with the
Hardy-Weinberg law, and each of the species ex-
periences pressure from the number in the entire
population, as described in paragraph 2, for ex-
ample. We can assume, however, that the value of
the parameter /i characterizing the effect of the en-
vironment is different for the two phenotypes,
although both values are in the zone ensuring the
existence of a stable steady state. Let us further sup-
pose that A = 1; that is, that each generation produces
offspring once. This assumption is not critical for
conclusions that can be reached using a model, but
simplifies the calculations. If we designate the num-
ber of the total population at time t by N(t), and
the share of gene A in the genofund of the offspring
produced by the generation living at time t as x(t),
then
x(t+ 1) =
x(t)
- x
(t)]2
N(t +
where
= juN(t)e-cN (t) [l - a[ 1 - x(t)]2]
(5)
(6)
/*! = fj. is the value of the environmental factor
for the phenotype corresponding to
genotypes A A and Aa;
/x2 = /i(l-a) is the value of this factor for the 0<
< 1 phenotype corresponding to geno-
type aa.
In both these numbers lie between 1 and 3, the
net populations containing only genotype AA, or
only genotype aa, have stable steady numbers, and
the difference between them is small if a is small.
It may readily be seen that the mixed population
too has a stable steady state:
x= 1, 1/c In /*,
during which the recessive species disappears. But
it is more interesting to estimate the rate at which
this species is displaced.
Eq. (5) will provide this estimate quite readily.
For example, if the original percentage of the
species with genotype AA is equal to p0, and is equal
to P! after time n, we have
-ln[l Xa(l Xp,,)2)] -In [1-«(1 -Pi)2)]
The second example of the influence of the en-
vironment on the population through the genetic
mechanism that we shall analyze illustrating the
theorem is that the genetic species can prove to be
necessary in order to maintain the number of the
population under variable (periodically changing,
for example) environmental conditions.
It has long been established (references [6, 7, 8]
for example) that there are clearly expressed differ-
ences between generations in populations of some
small rodents, and other animal "ephemerals" that
live for a few months and multiply several times in
the summertime. The situation can be described,
albeit very roughly, as follows.
Specimens with relatively low fertility predominate
in a generation that overwinters and has offspring at
the beginning of the season. However, a generation,
173
-------
for example, that is born in the first half of the sum-
mer ("descendants" of the previous generation) is
predominated by specimens with a very short sexual
maturation period, is very fertile, and has a short
life span. Since the population is substantially re-
duced during the wintertime, the population is re-
stored to its former figure. It is considered estab-
lished, at least for some populations, that the dif-
ference between the species we have described
("living through the winter" and "summer") is in
fact genetic in nature.
We have been considering a very simple mathe-
matical model that suggests one possible approach
to an explanation for the need for periodic rear-
rangement of the genetic structure of a population
in order for that population to survive.
Just as was the case with the preceding models,
we must bear in mind the fact that our purpose
simply is to present a graphic, mathematically cor-
rect demonstration of a definite phenomenon by the
example of a simple "theoretical population." We
should note that a number of the simplified assump-
tions can be discarded without creating any difficul-
ties. There are others that are more significant in
content, however, and which we shall discuss in what
follows, that must be taken into consideration, even
in the more complex models, those in which the
effect of the influence of polymorphism on support
of the number in the population also can arise.
Let the population under consideration consist (as
in the preceding example) of two phenotypes de-
termined by one gene with dominant (A) and re-
cessive (a) alleles.
Let us also assume that two reproduction cycles
take place during a season, and let us, for purposes
of simplicity, assume that each generation reproduces
but once. The generation born during the second re-
production cycle will not reproduce in this same
year, and the part of this generation that overwinters
is the first of the reproducing group the following
year.
Let «! be the mean number of descendants among
the specimens with genotype AA and Aa, and o2
= /fai (/?>!) be the mean number of descendants of
the specimens with genotype aa.
Let us make the further assumption that there will
be no deaths among those specimens in the popula-
tion that do not reproduce during the summertime,
and that during the fall and winter A, percent of the
number of specimens with genotypes AA and Aa,
and A2 percent of the number of specimens with
genotype aa, will die; O^A^A^l.
Accordingly, the specimens with genotypes AA
and Aa (let us call them the "winter" specimens)
have a relatively low fertifility rate, but tolerate the
fall and winter better. Those specimens with the
"summer" genotype aa are more fertile, but at the
same time are more sensitive to environmental
changes.
We shall here, as in the preceding model, take it
that panmixia takes place in the population.
Now let X, Y, and Z be the numbers of specimens
with genotypes AA, Aa, and aa, respectively. N = X
+ Y + Z is the total number of the population, and
x = X/N, y = Y/N, z = Z/N
are the percentages of the corresponding genotypes.
The vector (X, Y, Z) determines the structure of
the population at each moment in time. If the entire
generation lives until the next propagation period the
structure of the succeeding generation can be de-
termined by the operator T
T(X, Y,Z) = (X',Y',Z')
where
v,_ai(X+l/2Y)2
Z' =
N
(ou/2 Y+a2Z)(Z+l/2Y)
N J
g1(X+l/2Y)(Z+l/2Y)
N
+ (ai/2 Y+a2Z)(X+l/2 Y)
N '
Let us further assume that
S(X, Y, Z) = [(1-\1)X, (1-A,)Y, (l-x2)Z].
Then the population with the structure (X, Y, Z)
at the beginning of the summer will have the struc-
ture
(ST2) (X, Y, Z) = (X,, Yj, Z,)
at the beginning of the next summer.
It can safely be pointed out that the corresponding
transformation of the variables (x, y, z) (for which
x + y + z = l) is completely determined by their values
(that is, it does not depend on N). The analytical
form of the representation of the triangle used in
this case is much more complex than in the conven-
tional iterations of quadratic representations in
mathematical genetics (see reference [8]). Formally,
the reason for this fact reduces to this. One is forced
to make a certain normalization, as distinct from a
purely genetic situation, in which the sum of the
coordinates can be preserved automatically (even if
panmixia is lacking) in order to retain the condition
that x + y + z= 1 because of the differences in fertility
and mortality among the different genotypes.
Obviously then, a pure population consisting only
of specimens with genotype AA will survive if
a12(l-X,)>l. (7)
Analogously, a pure "summer" population will
survive when
a22(l-X2)>l.
(8)
We shall demonstrate the fact that a mixed popula-
tion, in which "winter" and "summer" specimens
are present, will survive in the case of certain values
174
-------
of the parameters «i, <*2, Xi, X2, that satisfy neither
of the inequalities at (7) and (8).
We should point out that in the model under con-
sideration the dependence of fertility on density has
been omitted for reasons of simplicity. However, if
this dependence is introduced, as was done in the
preceding model, similar effects will occur here as
well.
We have made yet another simplification, one
that allows us to avoid making cumbersome calcu-
lations. To that end, let us assume that X2 = l (the
"summer" specimens all die during the winter). This
of course is not compatible with the inequality at
(8), so in this case it is convenient to use the vari-
ables N andij=y/2.
Let us put ai = a, X! = A, a2 = )3a. Direct calcula-
tions will lead to the following relationship, one
determining population dynamics during the year
where
= (l-X)Na2(l
(9)
(10)
(20-2)K2
The value ?/=0 is a fixed point on the represen-
tation of the segment [0, Vi] itself, found by using
Eq. (10). This point is stable when /3<3, with the
result that in a mixed population of heterozygous
specimens (and, consequently, the percentage of
summer specimens that appear during the season)
it vanishes. In this case the conditions for the sur-
vival of the mixed population are given by the in-
equality at (7), just as was the case for the pure
population.
However, when /? > 3, the point ^=0 will be-
come unstable, and another stable, fixed point i?
= (/8-3)/2(/3-l), will appear (see Figure 2).
What this means is that each spring the popula-
tion will contain a definite percentage of heterozygous
specimens, so that summer specimens will appear in
the course of the reproduction season, and will re-
produce. These specimens represent a not inconsid-
erable part of the population. The condition for the
survival of a mixed population, readily obtained from
the inequity Na(N, ijT^N, has the following form
when /8>3
(ID
8(3-1)
Thus, if for given a and 0
I +
(0 - 3V
1-X L 8(0-lV
the pure "winter" population will die out, but the
mixed population will not! (Let us recall that the
pure "summer" population too will die out, because
A.-1.)
K
FIGURE 2. POPULATION DYNAMICS
175
-------
The coefficient l + 08-3)2/8(/3-l) measures
the advantage in viability imparted by polymorphism.
Let us use a numerical example that will show
quite graphically the population dynamics in this
case to illustrate our conclusions.
Let a = 2, 0 = 5 (that is, a2 = 10), A = 7/9. In this
case /t0 = Vi. Let us set X = Y = 100.
Times
X
Y
Z
N
t
100
100
0
200
t+1
225
150
25
400
t + 2
450
450
100
1000
t+3
100
100
0
200
1 year
For purposes of comparison with a pure
>pulation
tt+1t+2
populatii
"winter"
t+3
X=N
200
400
800
178
We have selected limit X values for which the in-
equality at (11) will be transformed into an equal-
ity, despite the fact that from the standpoint of our
model this is an exceptional case, in that a small
change in A will lead to dying out, or to exponential
growth of N and to increasing oscillations in the
magnitudes X, Y, Z. However, this behavior will be
typical of a model in which fertility is a function of
density.
We shall conclude our discussion of this model
with a few remarks. Along with the simplifications
already mentioned, we have assumed that one gene
has a simultaneous influence on fertility and sur-
vivability under severe environmental conditions.
This assumption would not appear to be very natu-
ral at first glance. However, it can be supposed that
the "winter" and "summer" species can be dis-
tinguished by metabolic intensity, the consequence
of which is a longer period of sexual maturation and
a longer life span among the "winter" specimens.
Here consideration must be given to the fact that
the greater fertility of the "summer" specimens is
the result of their success in reproducing a greater
number of times.
A scheme such as this can be realized in a model
if, needless to say, the number of reproduction cycles
per season is increased (for the "summer" specimens
at the very least). Another extremely interesting di-
rection is associated with study of the oscillations
that take place in the genetic structure in a popula-
tion of long-lived organisms (see reference [9]).
4. The final subject we shall touch upon in this
report is the effect man has on a population as it is
exploited; that is, by the withdrawal of some of the
specimens from the population. The practical im-
portance of this subject is obvious, and its study
covers a considerable part of mathematical ecology
(see the well-known book by K. Uatt [10], for ex-
ample, in which the author concentrates on the
question of the effect exploitation has on a popula-
tion).
176
We should like to turn our attention to the con-
nection that exists between the subject of optimum
exploitation of a population and the principal theo-
rems in mathematical economics.
Let us consider the following model. Let a popula-
tion consist of N+1 mature groups. Let us designate
the number in the ith group by m, i = 0, 1, N.
Fi is the mean number of offspring per unit time for
the specimens in the itb group, pi is the probability
that specimens aged i will live to age i+1 under
natural conditions (no exploitation).
Let us, in addition, introduce the index Wi, the
exploitational worth of the specimen.
Let us give our external effect on the population
(or control) by the vector (uc, u,, ..... UN), where ui
is the percentage of the ilh group meaningfully with-
drawn from the population (for the time segment
under consideration).
We must, in order to establish population and
harvest dynamics, stipulate the moment in time when
the harvest must be removed (this need not be done
if a continuous time model is considered).
If the harvest is removed at the beginning of the
period we shall call the corresponding model model
A. It has the form
N
i=0
(12)
The harvest is equal to
N
w
V ui
= ) ,— v
I— I l-U. 1
587
1 = 0
(Wi can obviously be equated to the beginning of the
period as well.)
If we remove the harvest at the end of the period
(and after taking the effect of natural mortality into
consideration), we shall call the corresponding model
model B.
N
(13)
j\\ =(1 -ui.1)pMn..1,i
The harvest is equal to
N-l
ui-iPi-lni-iWi-j - 2, uiPi
i=0
-------
(Wi in this case can be equated to the end of the
period.)
We shall, for purposes of definiteness, henceforth
consider the type A model.
The problem, in the most general of terms, is that
of constructing that sequence u(1), u(2), ..... of equa-
tions that will maximize the harvest.
We have to know the factors upon which the dy-
namics of the number in a population depends in
order to state the problem more precisely (density,
nutrition, and the like).
It can be taken, for example, that the parameters
Fi, pi, Wi depend on n in such a way that an increase
in n will reduce their values.
Another approach is that there is some set of
reserves and that a limitation of the following type
can be satisfied
V
(15)
where
on is the consumption of reserve j by one
specimen in the ith group;
Fi, pi, Wi are constants that still satisfy the
limitation at (15).
Then we must clearly distinguish what it is that
we take as the criteria. If expenditures required to
remove the harvest are ignored, we can, for ex-
ample, maximum the summed harvest over a long
period of time.
We should point out that we must, naturally, limit
those controls that depend solely on the present state
of the system.
Employing the considerations that can be used for
the proofs of the so-called theorems of main lines in
consumption (see reference [11], Chapter IV), we
can prove the following. There exists that temporally
independent steady-state equation u = u0, ui . . . ,
UN_]( and the vector n, invariant with respect to this
equation, that is, satisfying the condition set forth
at (16)
i=0
(16)
vn>
and any optimum trajectory of length T always will
reduce to a small neighborhood of state n, except for
the end segments at the beginning and at the end,
where the controls used at this time are close to con-
trol u. We will not, here, discuss in detail the occur-
rence of the analogy between the economic and the
ecological magnitudes. We shall simply note that the
limitations of (15) play the same role that the exoge-
nous reserves play in an economic system; that is,
those reserves the system itself does not condition
(manpower, for example).
The considerations cited, which are new only
in part, have been advanced as a result of discussions
between the participants in the group engaged in the
mathematical modeling of ecological systems in the
Laboratory of Functional Analysis and Models of
Economic Dynamics at the TsEMI [expansion un-
known] of the Academy of Sciences of the USSR.
The group includes M. Ya. Antonovskiy, V. I. Dani-
lov, V. A. Iskovskikh, and A. B. Katok.
Academician S.S. Shvarts raised for our group
questions concerned with the interaction of ecologi-
cal and genetic factors. The phenomena he pointed
out are at the basis of the mathematical models con-
sidered in paragraph 3.
Discussions of the questions related to the influ-
ence of density on population dynamics with our
coworkers at the Institute of the Ecology of Plants
and Animals of the Academy of Sciences of the
USSR, and with V. S. Smirnov, and V. G. Ishchenko
in particular, were extremely useful.
BIBLIOGRAPHY
1. A. N. Kolmogorov. "Qualitative Investigation of Math-
ematical Models of Population Dynamics." Problemy
kibernetiki, No. 25, 1972. pp. 100-106.
2. H. G. Andrewartha, L. C. Birch. The Distribution and
Abundance of Animals. The University of Chicago
Press, Chicago, 1954.
3. E. Makfed'yen. Ekologiya zhibotnykh [The Ecology of
Animals]. "Mir" Publishing House, Moscow, 1965.
4. V. L. Andreyev, A. A. Nagorskiy, A. P. Shapiro. "Mod-
eling a Fish Population With a Two-Year Life Cycle
and a Single Spawning." Problemy kibernetiki, No. 25,
1972. pp. 167-176.
5, H. D. Ford, E. B. Ford. "Fluctuation in Numbers and
Its Influence on Variation in Melitea Aurinia." Jor.
Trans, Royal Ent. Soc., London, Vol. 78, 1930.
6. S. M. Gershenzon. "Seasonal Changes in the Frequency
of the Occurrence of Black Hamsters." DAN SSSR, Vol.
43, No. 9, 1945.
7. S. S. Shvarts. Osnovy evolyutsionnoy ckologii zhivot-
nykh [Fundamentals of the Evolutionary Ecology of
Animals]. Sverdlovsk, 1969.
8. F. Frank. "The Causality of Microtime Cycles in Ger-
many.'';. Wildlife Mgt., 21, 113-121 (1957).
9. S. S. Shvarts, V. G. Ishchenko. "Dynamics of the Ge-
netic Composition of a Population of Rana terrtstris."
Byulleten' Moskovskogo obshchestva ispytateley prirody.
Old. biologii., Vol. 83, No. 3, 1968. pp. 127-133.
10. K. Uatt. Ekologiya i upravleniye prirodnymi resursami
[Ecology and Control of Natural Resources], "Mir"
Publishing House, Moscow, 1971.
11. Kb. Nikaydo. Vypuklyye struktury i matematicheskaya
ekonomika [Convex Structures and Mathematical Eco-
nomics], "Mir" Publishing House, Moscow, 1972.
177
-------
Global Climate and Human Activity
M. I. Budyko,* I. L. Karol *
1. Models of Changes in Global Climate
Previous investigations (Budyko, 1969, 1971,
1972; SMIC 1971, and others) have shown that
modern economic activity already has had a marked
influence on global climatic conditions, and this in-
fluence is increasing rapidly.
This fact has given rise to the important task of
estimating possible climatic changes in the future
for different conditions of further economic develop-
ment.
The only way to accomplish this task is to use
numerical models of climate theory to determine
changes in meteorological conditions.
Models of a climate investigated for purposes of
studying its possible future changes should satisfy
certain requirements, but these latter are less im-
portant for models used in investigations of the pres-
ent day climate.
The model ought not include empirical data on
the distribution of individual climatic elements, es-
pecially those which change significantly with
changes in climate.
The model should take into consideration all types
of heat flows with a marked effect on the tempera-
ture field, and the law of the conservation of energy
should be satisfied.
Climate is formed by:
1. the distribution of radiation in the atmosphere;
2. the transfer of heat and moisture by the move-
ments of the atmosphere and ocean;
3. the hydrology of the underlying surface (it has
a greater influence on the local climate of the lower
troposphere, and is not considered here).
The heat balance equation
Q(1-A)-I = F, (1)
links the flow of heat from the sun's shortwave radia-
tion, Q (A is the albedo of the earth-atmosphere
system), the infrared (IR) radiation of heat, I, for
this system, and the heat transfer, F, by the move-
ments of the atmosphere and ocean, as basic factors.
These factors depend significantly on the following
parameters (Budyko, 1971): A, the albedo of the
earth-atmosphere system, increases sharply over (a)
a snow and ice surface, (b) clouds (especially dense
•Main Oeophyiical Laboratory, Leningrad
clouds), (c) without clouds somewhat more over
land bare of vegetation than over the sea.
The following also have a great effect on the dis-
tribution of radiation in the atmosphere: (a) clouds
and water vapor absorbing and scattering short- and
longwave radiation; (b) dust, doing the same to the
sun's shortwave radiation; (c) carbon dioxide, CO2,
which absorbs IR radiation; (d) ozone, O3, which
heavly absorbs the sun's ultraviolet (UV) radiation.
Global movements of the atmosphere are deter-
mined by the planetary distributions of its tempera-
ture and humidity and, in turn, form them. Their
intensity depends primarily on the difference in tem-
peratures at the equator and at the poles, and be-
tween the continents and oceans. The relationship
between the nature and intensity of ocean currents
and external conditions is more complex. It is only
partially linked with atmospheric circulation.
These climatic factors are linked by the following
positive and negative feedback rings.
1. Polar ice, almost doubling A, reduces the air
temperature, T, and thus promotes its own distribu-
tion. This is a positive feedback ring, which exerts
a great influence on the distribution of T (Budyko,
1971).
2. Clouds, generally speaking, reduce the radia-
tion passing to the earth, and reduce T, and evapora-
tion, E, which leads to a reduction in cloud cover.
This is a negative feedback ring. But they do inhibit
IR radiation and block and radiation reduction in T.
This is a positive feedback ring.
3. Increase in absolute humidity with increase in
T leads to the absorption of longwave radiation by
the atmosphere and to its heating. This is a positive
feedback ring.
4. Increase in T increases the intensity of the at-
mosphere's IR radiation. This is a negative feedback
ring.
5. Increase in the CO, in the atmosphere reduces
the intensity of its absorption by the ocean, given a
rise in temperature and in the acidity of the waters
of the latter. This leads to an even greater accumula-
tion of CO2 in the atmosphere, and is a positive feed-
back ring (Machta, 1973).
179
-------
6. The intensity of heat and moisture transfer de-
creases with warming of the polar atmosphere, and
so does the transfer of heat from the equator to the
poles. This is a negative feedback ring.
Negative circulation rings increase climatic sta-
bility; positive ones decrease it.
The parameters of virtually all of these feedback
rings have not yet been adequately studied, and it is
possible that other feedback rings exist, but if they
do they are not significant.
Man can affect climate in the following ways.
(1) A change in the components of the heat bal-
ance of the underlying surface, of a city, a reservoir,
irrigation of fields and of agricultural land, is the
most noticeable, and comparatively well-studied
change in climate, but is only local, and in the lower
troposphere.
(2) The emission of heat into the atmosphere
even now is changing local climate in cities ("heat
islands" in the center of a city). When the current
rate of growth in the world's energy requirements
(about 6% a year) is extrapolated, the magnitude of
this emission in the 21st century will become com-
parable to the magnitudes of the components of the
heat balance for the underlying surface, particularly
in urban and industrial agglomerations (Budyko,
1972).
(3) The increase in the CO2 in the atmosphere
attributable to the combustion of fossil fuel in recent
decades is 0.2% per year (Machta, 1973). Accord-
ing to the estimates contained in this reference, the
mean concentration of CO2 in the atmosphere in
1860, was approximately 10% lower, and in the
year 2000, will be 15-20% higher, than its current
value.
(4) The increase in the content of atmospheric
aerosols. Stratospheric aerosols on the whole reduce
solar radiation incoming to the lower atmosphere
and the underlying surface. Heating of the strato-
sphere by radiation absorbed by these aerosols has
almost no influence on the temperature of the tropo-
sphere, and thus as a practical matter reduces the
meteorological solar constant (Kondrat'yev, et al.,
1973).
The effect of tropospheric aerosols is more com-
plex. They scatter and absorb radiation, and this
respectively increases and decreases the albedo (A)
of the "earth-atmosphere" system. This change in A
depends on the effective albedo of the aerosols, A*,
and on the albedo of the underlying surface, A.. If
A. is small, and if AsA» (over snow and clouds), then
an increase in C reduces A, and increases T. A. de-
pends on the aerosol's optical properties and can
change greatly in different regions, and in different
seasons, but as a mean, Aa is close to the mean for
A» in the absence of clouds, snow, and ice (Kon-
drat'yet, et al., 1973). Aerosols under a cloud layer
have little influence on the radiation flux near the
earth.
Aerosols of anthropogenic origin now comprise
10-20% of the total mass of atmospheric aerosols.
There are predictions that suggest a doubling of this
percentage by the year 2000, if special measures are
not taken to reduce the emission of aerosols into the
atmosphere by man (SMIC, 1971).
(5) Mass flights of aircraft at a level of about 20
km can result in the accumulation in the stratosphere
of the combustion products exhausted from their
engines, in a decrease in the ozone content by inter-
action with the nitrogen oxides they emit, and in an
increase in stratospheric aerosols (Grobecker, 1973).
The possible effect of these aerosols on climate was
indicated above, and a reduction in the ozone layer
will lead to an increase in the biological danger from
the sun's UV radiation. A change in the ozone con-
tent can change the climate in the stratosphere (es-
pecially its circulation), because the absorption of
solar radiation by ozone to a considerable degree de-
termines the heating and temperature of the strato-
sphere.
Various assumptions can be made with respect to
each of these factors, on how each will change in the
future as further economic development takes place.
The task of climatological investigations is to esti-
mate climatic conditions for possible variants of
change in these factors. The question of the reliability
of estimates of these changes is worthy of serious
attention because of the great practical importance
of global climatic changes.
Two different approaches can be used in consid-
ering this question. The first involves the verification
of the numerical models used for calculating the cli-
mate of the future using empirical data on climatic
changes in the past. This approach was used in pre-
vious work by one of the authors (Budyko, 1971,
1972, and others).
The second approach is based on a comparison
of the results of calculations of climatic conditions
of the future, using different, independently devel-
oped, models of climatic theory. The feasibility of
using this approach appeared recently as a result of
the making of a series of investigations of anthropo-
genic climatic changes.
Let us consider the results of calculations of the
influence of increasing the production of energy on
climate, made using various models of climatic
theory.
Washington (1971) used a three-dimensional
model of climatic theory to calculate the change in
the air temperature field in the troposphere as the
result of an additional heat inflow of 50 cal/cm2/day,
equally distributed over the surface of the continents.
180
-------
The mean air temperature in this case rose from 1 °
to 2° C in the tropics, and from 8° to 10° C in the
higher latitudes of the northern hemisphere.
A calculation using a semiempirical model of cli-
matic theory (Budyko, 1969) yields approximately
the same increase in the mean global temperature of
3° C for these conditions (note that in neither case
was the dependence of the albedo of the earth-
atmosphere system on temperature changes taken
into consideration).
Estimates of change in the carbon dioxide con-
centration with thermal conditions in the atmosphere
were made in our paper (Budyko, 1973) and in that
by Manabe (SMIC, 1971; Smagorinsky, 1974).
Let us note that the feedback between changes in
thermal conditions in the atmosphere and the posi-
tion of the snow and ice cap were taken into con-
sideration in all these studies.
A comparison of the results of the calculation of
the effect of doubling the concentration of carbon
dioxide on the mean latitudinal air temperature,
made using our semiempirical climatic theory, and
the general climatic theory used by Manabe, shows
good agreement between the estimates of the in-
fluence of change in carbon dioxide concentration on
climate obtained using the different models. Hence,
it follows that in such case the mean air temperature
increases by 2°-3° C in the low latitudes, and by up
to 10° C in the high latitudes.
The question of the influence of an aerosol on
thermal conditions in the atmosphere has been dis-
cussed in numerous studies. It is determined by the
radiation flux in the atmosphere as a function of the
aerosol concentration, and by the air temperature as
a function of changes in shortwave radiation flux.
Leaving the first part of this link aside for the
moment, let us consider the results of the calculations
of air temperature as a function of the inflow of solar
radiation. These calculations have been made re-
peatedly for different models of climatic theory with-
out taking into consideration the feedback between
thermal conditions in the atmosphere and polar ice.
This feedback was taken into consideration for
the first time within the framework of a semiempiri-
cal model of climatic theory (Budyko, 1969). This
study established that the mean air temperature at
the earth's surface changes sharply with small fluc-
tuations in the solar constant, and that a reduction
of approximately 2% in the solar constant will re-
sult in complete glaciation of the earth. A similar
conclusion was obtained as a result of the develop-
ment of a series of different semiempirical models of
thermal conditions in the atmosphere, including con-
sideration of the feedback indicated above (Sellers,
1969; Schneider and Gal-Chen, 1973).
The conclusion drawn with respect to the high
degree of the sensitivity position of the polar ice to
small changes in the inflow of solar energy resulted
in obtaining a quantitative explanation of the process
of the development of quaternary glaciation (Budyko
and Vasishcheva, 1971; Berger, 1973).
The question of air temperature as a function of
the solar constant was studied in the last papers
written by Manabe (Smagorinsky, 1974) who
used general climatic theory and included consid-
eration of the feedback between thermal conditions
and the snow and ice cap. Here, it was found that a
2% increase in the solar constant resulted in a 1.2°
C mean change in the air temperature for the hemi-
sphere, according to the Manabe and Vezerold local
model of thermal conditions when changes in abso-
lute air humidity were disregarded. The change was
2.4° when absolute air humidity as a function of
temperature was included and the same model was
used, but was 3.1° C when Manabe's model of gen-
eral climatic theory was used. The latter magnitude
differs comparatively little from the result of the
analogous calculation made using the above-men-
tioned semiempirical model of the thermal conditions
in the atmosphere, from which it follows that the
indicated increase in temperature is 4.5° C.
In estimating the magnitude of the relative reduc-
tion in the solar constant sufficient to completely
glaciate the planet, Manabe, as well as the authors
of the semiempirical studies, found it comparatively
small, and approximately equal to 4%.
Thus, a great many studies based on the use of
totally different models of climatic theory produce
what are comparatively close estimates of the influ-
ence of a change in climate-forming factors on cli-
mate. This permits us to use these models with a
known degree of confidence in making calculations
of climate changes that could result from the devel-
opment of economic activity.
2. Changes in Climate by the End of the 20th
Century
Let us examine the climatic changes that could
take place by the end of our century as a result of
anthropogenic factors. Calculations suggest that in-
crease in energy production (by approximately 6%
a year) will have no significant influence on climate
before the end of the 20th century, although this in-
fluence could be very substantial over the next cen-
tury (Budyko, 1972).
Calculations of the influence of aerosols on ther-
mal conditions in the atmosphere are difficult to
make because of the lack of data on the mean values
of the parameters of the models used, particularly
as the data pertain to the coefficients of radiation
absorption by aerosol particles. Here it is that anal-
ysis of empirical data is of considerable importance
in estimating the influence of an aerosol on radiation
and thermal conditions.
The data from observations made by a group of
actinometric stations in Europe, Asia, and America
181
-------
were processed by Z. I. Pivovarova and the results
were used to chart the secular course of direct solar
radiation in a cloudless sky over the past decade for
purposes of estimating the influence of an anthropo-
genic aerosol on climate (Budyko and Vinnikov,
1973). The conclusion reached as a result of this
investigation was that anthropogenic aerosols have
in recent years reduced direct radiation in a cloudless
sky by approximately 6%, and the summed radia-
tion for average conditions by approximately 0.5%.
The use of the Yamamoto and Tanaka (1972)
model, in which the influence of aerosol on back-
scattering and radiation absorption is taken into
consideration, resulted in the finding that this reduc-
tion in radiation corresponds to a reduction in the
mean air temperature at the earth's surface of ap-
proximately 0.5° C.
Although, as noted in this study, this estimate is
very approximate in nature, it can be assumed that
further rise in the aerosol concentration can lead to
a significant change in climate. However, it is quite
difficult to predict the future quantity of atmospheric
aerosols.
We can assume that failure to combat atmospheric
pollution will result in the ratio of increase in the
anthropogenic aerosol mass to its current value being
equal to the analogous value for the anthropogenic
carbon dioxide mass in the atmosphere. Given this
assumption, the quantity of anthropogenic aerosol
can increase by a factor of 2.5 to 3 by the year 2000.
But because many countries are taking steps to re-
duce atmospheric pollution, the increase in the quan-
tity of anthropogenic aerosols will come to a halt,
and their quantity by the end of the century will re-
main at the current level, or may even decrease.
Given these conditions, increase in the carbon dioxide
concentration, leading to a warming trend, will exert
the major influence on change in climate.
A variety of models of climatic theory can be used
to estimate this warming trend. What follows from
the above-cited Manabe and Vezerold local model
of climatic theory is that a 20% increase in carbon
dioxide concentration will increase the mean global
air temperature 0.5°. Manabe's calculations, based
on the use of a three-dimensional model of general
climatic theory, showed a 0.7° C increase in the
mean global temperature for this case, and approxi-
mately twice that for the high latitudes.
A similar calculation, using our semiempirical
model of thermal conditions in the atmosphere and
the relationship between air temperature and car-
bon dioxide concentration according to Rakipova-
Vishnyakova, and given the assumption of the quasi-
stationary state of polar ice (see Budyko, 1972),
showed a 0.6° C increase in temperature, and, given
the assumption of stationary ice conditions, 0.9° C.
As we see, there is little difference between these
estimates.
Calculations using a semiempirical model of ther-
mal conditions in the atmosphere show that the area
of sea ice in the northern hemisphere can be reduced
by 10 to 20 percent by an increase of 0.5° to 1 ° C in
the mean global temperature. This will be accom-
panied by a several degree increase in temperature
during the winter in the high latitudes, something
that can greatly influence Arctic navigation condi-
tions, as well as many branches of the national econ-
omy in the northern regions.
Fluctuations in the amount of precipitation that
falls, associated with change in thermal conditions,
can be of even greater practical importance. The
overall warming trend can be accompanied by a
reduction in the mean meridional temperature gra-
dient, because the air temperature in the high lati-
tudes rises much more than it does in the low lati-
tudes with increase in the carbon dioxide concentra-
tion. As is known, changes in this gradient have
very definite influence on the intensity of zonal
circulation.
A series of empirical studies (Budyko and Vinni-
kov, 1973, and others) has established the fact that
there is a signicant relationship between the precip-
itation falling on continents and relatively small
changes in the mean meridional temperature gradient.
Reductions in this gradient that took place, for ex-
ample, during the warming trend of the 1920's and
1930's, were accompanied by a considerable reduc-
tion in precipitation in many intracontinental areas.
This led to a substantial increase in the number of
droughts covering large areas in the USSR, USA, and
other countries, to a drop in the level of many in-
ternal water expanses, including the Caspian Sea,
and to a number of other consequences.
Change in climatic conditions that can occur by
the end of the 20th century as a result of increase in
the carbon dioxide concentration will greatly exceed
the warming trend of the 1920's and 1930's, and
could have a great influence on agricultural produc-
tivity. The need for a more detailed study of anthro-
pogenic changes in climate, using models of the dy-
namics of the atmosphere and oceans, and created
just to study possible changes in climate, therefore is
obvious. These models should be nonstationary,
quite detailed, yet not too cumbersome. Calculations
of many variants of change in global climate, decade
by decade, then could be made by using them and
modern computers. These calculations could take
into consideration the influence of the oceans and
polar ice, as well as various assumptions as to man's
future effect on specific climatic factors. There should
be several such models, all incorporating the various
assumptions, and coordination of estimates of
changes in climate, calculated using them, would en-
able us to establish the reliability of these estimates
and models. Coordination of international scientific
efforts in this field is very important.
182
-------
Calculations using these models will enable us to
find the levels, and growth rates, of the anthropogenic
effect on global climate that will not exceed the nat-
ural variation in the scale of the warming trend of
the 1930-1940 period by the end of the 20th century.
There can be no doubt of the great importance of
those data for planning the optimum interaction be-
tween society and the environment,
3. Controlled Effect on Global Climate
One approach this effect can take in order to pre-
vent undesirable changes in climate is based on an
artificial change in the concentration of aerosol par-
ticles in the lower stratosphere (Budyko, 1974).
The nature of an aerosol in the lower layers of the
stratosphere was studied by Junge (1963), Cadle
(1972), and others, who established that there is a
layer of particles, the mean size of which is measured
in tenths of a micron, in the lower stratosphere. The
heaviest concentration of particles usually will be
observed in a layer about 2 km thick at a height of
18-20 km.
The particles in the stratosphere's aerosol layer
are mostly sulfur compounds. Junge hypothesized
that these particles are formed from sulfur dioxide
(SO2) which, oxidizing during photochemical re-
actions, is converted into sulfur trioxide (SO3), and
with water vapor forms dorplets of surfuric acid
(H2SO4), which comprises the major part of the
aerosol in the lower stratosphere (SMIC, 1971).
Aerosol particles are retained in the stratosphere
for a long time with the average from 1 to 3 years.
The mechanism involved in their precipitation, ap-
parently, is connected with their coagulation and
increase in size. The very slow movement of the
particles in the vertical direction contrasts with their
extremely rapid transfer along the horizontal. Ob-
servation of the distribution of particles in the strato-
sphere after volcanic eruptions shows that when an
eruption occurs in the extra-tropical latitudes, the
particles are spread over the entire hemisphere
within a few months by stratospheric winds. The
particles are distributed over both hemispheres with-
in this same time frame if the eruption occurs in the
equatorial zone.
According to observational data, aerosol particle
concentration can change over broad limits. Accord-
ing to Junge's observations in 1961, for example,
the concentration was of the order of magnitude of
0.01 jug/m3, and had reached 0.20-0.30 /xg/m3 in
1968-69. This marked increase in the concentration
is explained by the influence of the eruptions of the
Agung volcano, and of some others. In 1970, the
concentration of aerosol particles decreased to sev-
eral hundredths of a ^g/m3.
It is interesting to compare aerosol concentra-
tions for different years with the magnitudes of
direct radiation in a cloudless sky for the correspond-
ing periods of time. Z. I. Pivovarova used data from
observations made by a group of actinometric sta-
tions in the USSR and USA to calculate mean direct
radiation magnitudes. These were found to be
equal to 98% of the norm in 1961, to 93-94% in
1968-69, and to 97% in 1970. The definite con-
cordance between changes in aerosol concentration
and fluctuations in the direct radiation magnitudes is
worthy of attention. This makes it possible to take
the comparatively few changes in aerosol concen-
tration that are quite characteristic for the whole of
the northern hemisphere into consideration.
The direct proportionality between changes in
aerosol concentration and changes in radiation flux
follows from the theoretical considerations. The data
cited in the foregoing then show that a 0.06 /*g/m3
increase in the aerosol concentration reduces direct
radiation by 1%.
This significant dependence of the shortwave
radiation flux on the aerosol concentration in the
stratosphere explains the marked influence of explo-
sive volcanic eruptions on climatic conditions.
Actinometric observations have shown a marked
reduction in direct radiation all over the world, or
within the limits of one hemisphere, after explosive
eruptions. Direct radiation decreased by 6% after
the eruption of Agung volcano in 1963, and there
was an even greater reduction in radiation after the
eruption of Mount Katmai in 1912, when the reduc-
tion in direct radiation was 22% (Budyko, 1971).
Fluctuations in direct radiation after volcanic erup-
tions are accompanied by changes in summed radia-
tion that make up a comparatively small part of the
magnitude of changes in direct radiation. It can be
assumed that the influence of a reduction in summed
radiation on the temperature of the lower air layer
depends on how long the reduction in radiation lasts.
Meteorological observation data show no marked
change in the temperature near the earth's surface
after the eruption of Agung volcano, but there was
a reduction of 0.5 °C, on the average, for the north-
ern hemisphere after the eruption of Mount Katmai.
A reduction such as this in temperature apparently
is a small part of the change that would occur in
the event of a prolonged reduction in radiation.
These data therefore suggest that the summed radia-
tion decreased by approximately 1% after the first
eruption, and by 3% after the second.
The thermal conditions involved in the "warming
of the Arctic" were much closer to the steady state.
Direct solar radiation increased by approximately
2% in these years, and this increase continued for
almost twenty years. An increase such as this in
direct radiation corresponds to a 0.3% increase in
summed radiation. This should lead to an approxi-
mate 0.45°C rise in planetary temperature at the
earth's surface under steady-state conditions. This
183
-------
latter value is only slightly greater than the change
in temperature established from observational data.
Thus, the change in temperature in the lower air
layer when there are fluctuations in stratospheric
transmittance depends on two factors: the magnitude
of the change in the concentration of aerosol par-
ticles in the stratosphere; and the duration of this
change.
Let us examine the possibility of a change in cli-
mate on the scales observed during the "warming of
the Arctic" period, but in the opposite direction; that
is, providing for a reduction in the mean tempera-
ture at the earth's surface. This would mean a 2%
reduction in the mean magnitude of the direct radia-
tion, and this corresponds to a reduction of 0.3%
in the summed radiation.
We have already cited an empirical estimate of
change in radiation as a function of aerosol con-
centration in the lower stratosphere. Other calcula-
tions of this relationship, based on empirical data
and on a theoretical analysis, yield a somewhat
greater magnitude of the aerosol mass changing
shortwave radiation by the specified magnitude. The
mean of all these calculations of the magnitude of
the aerosol mass changing summed radiation by 1 %,
and the direct radiation by 6%, is close to 0.8 x 10-"
g/cm2.
A close look at this estimate will show that for
the purpose indicated above the aerosol mass in the
northern hemisphere should be increased by 600,000
tons.
All that has to be done to increase the mass of
the particles measured in the lower stratosphere to
this magnitude is to deliver 200,000 tons of sulfur
to the aerosol layer level, and this, after combustion,
will form 400,000 tons of sulfur dioxide. If we con-
sider that the mean life span of aerosol particles in
the stratosphere is 2 years, 100,000 tons of sulfur
annually would have to be delivered to the lower
stratosphere in the northern hemisphere in order
to create the necessary concentration of particles
consisting of pure sulfuric acid.
The combustion in the stratosphere of much
smaller quantities of sulfur than this probably would
actually be sufficient to increase the mass of aerosol
particles to 600,000 tons, because the hygroscopic
droplets of sulfuric acid will absorb water vapor
from the air.
The planning of such effects requires accurate es-
timates of the possible influence on stratospheric
aerosols of mass flights by supersonic transport air-
craft in the 1980's, because their influence can be
considerable (Grobecker, 1973). A study of the
causes, and of the nature, of natural fluctuations in
the content of these aerosols also is needed for
their prediction and consideration when carrying out
the effects.
Bringing about the effects on climate examined
above will prevent, or weaken, the changes in cli-
184
mate which could take place in 20-30 years as a
result of the influence of economic activity. But this
effect is not the limit within reach of modern tech-
nology. It evidently will be technically possible in
the near future to affect the climate, creating changes
in it that will be greater by an order of magnitude
than those that occurred in the climate in the 1920's-
1930's; that is, ensuring a few degree reduction in
global temperature. An effect on this scale could be
necessary in the 21st century, when the temperature
of the lower layers of the atmosphere could rise
considerably as a result of a considerable growth in
energy production.
The question of the desirability of using effects
on climate in the near future, when the influence of
economic activity on climate will be comparatively
slight, deserves study. Effects that will reduce the
mean temperature several tenths of a degree will be
economically desirable. Their use to change thermal
conditions will have little influence on economic
activity, but an increase in the amount of precipita-
tion can be of great practical importance in many
areas of insufficient moisture.
Let us note, in conclusion, that elucidation of the
possibility, in principle, of a marked change in
global climate by comparatively simple means raises
a number of questions.
Study of these questions means that attention
must be given to the international aspect of the
problem of effect on climate. Ye. K. Fedorov
(1958, and others) indicated repeatedly that the
implementation of any major project affecting cli-
mate is possible only on the basis of international
collaboration.
The method of affecting climate that we have
considered here will lead to change in climatic con-
ditions in the territories of many countries, but the
nature of these changes will be different in different
regions.
This is what obviously makes it desirable to con-
clude an international agreement resolving the ac-
complishment of effect on climate only on the basis
of projects that have been examined and approved
by responsible international agencies, with the agree-
ment of all interested countries. Such an agreement
should encompass measures for guided effect on
climate, as well as those types of economic activity
similar to mass flights in the lower stratosphere, that
could lead to unintentional changes in climatic con-
ditions.
4. Global Monitoring of Climatic Factors
A global system of continuous observations and
periodic examination of the factors forming global
climate, and its anthropogenic changes, is attracting
more and more attention from researchers as a com-
ponent part of global monitoring of the environment.
The magnitudes subject to global monitoring, the
-------
basic approaches to its accomplishment, approximate
accuracies and frequency of measurements, are
contained in the well-known SMIC handbook,
1971, and refined in the recently published Global
Monitoring Plan project (Munn, 1973). The global
monitoring of factors that have an effect on climate
is the primary goal of the recently organized net-
work of World Meteorological Organization stations
for measuring background atmospheric pollution
(Munn, 1973).
Continuous measurement and statistical process-
ing of primary data from natural climate-forming
factors are needed for:
1. direct and summed solar radiation, measured
on a standard basis by a network of some 1,000
actinometric stations, located in all geographical
regions, but unevenly (Budyko, 1971). Most of
these stations did not begin operations until the
second half of the 20th century, and their data are
as yet of little value in determining climatic changes;
2. the albedo, the mean cloud cover, and the
mean limit of the polar snow and ice cap. They are
most conveniently measured on a global scale with
standard weather satellite equipment (television
images) (SMIC, 1971). Long series of comparable
data must be collected in order to distinguish per-
manent changes in magnitudes, because these meas-
urements have only been made systematically since
the end of the 1960's;
3. the temperature of ocean surfaces, as mini-
mally characterizing thermal inertia and heat trans-
fer in the oceans. SMIC (1971) has pointed out the
inadequacy of the study thus far made of the energy
exchange between ocean and atmosphere, as well as
the need for just a rough working model of that ex-
change to define more precisely the magnitudes sub-
ject to global monitoring. The need for data on the
characteristics of the heat exchange that takes place
between the surface and deep layers of the ocean
and its surface currents is obvious, but it still is im-
possible to obtain continuous data of this type on a
global scale. The temperature of ocean surfaces can
be measured from satellites, given certain cloud con-
ditions (SMIC, 1971).
Certain data on factors describing the state of the
climate, and on how man has transformed the under-
lying surface and thus has had an effect on climate,
must be obtained periodically, and in definite regions.
Included, for example, are glacier mass, ocean level,
biomass of forests, area of cities and artificial reser-
voirs, and the like. Table 7.7 (SMIC, 1971), lists
these factors of the "climatic cadastre," the peri-
odicity, and regions to be examined.
The most important of the anthropogenic factors
affecting climate are continuous measurements of the
carbon dioxide and aerosols in the atmosphere. The
CO2 concentration undergoes little latitudinal and
seasonal changes, and its global distribution recently
was studied in some detail (SMIC, 1971). These
changes are greater in the northern hemisphere,
where the main seasonally changing sources and
flows of atmospheric CO2 into the continental
biosphere are located. Therefore, the global in-
crease in the mean CO2 concentration is easy to
detect by taking measurements on small islands dis-
tant from the continents, or in the southern hemi-
sphere. The main sources of anthropogenic CO2 are
in the temperate northern latitudes, while its oce-
anic flows and sources are in the polar and tropical
latitudes, respectively. Measurements of the meri-
dional distribution of the surface concentration of
CO2 therefore are necessary by a network of points
from the north to the south polar area. The network
of WMO base stations for measuring background pol-
lutants in the atmosphere is distributed with this
special requirement for station location taken into
consideration.
Attempts to estimate the summed annual anthro-
pogenic emission of CO2 into the atmosphere, and
its increase in recent decades, were published re-
cently (Keeling, 1973). A comparison between these
estimates and the measured increase in the content
of CO2 in the atmosphere possibly will permit re-
fining the rate at which absorption of CO2 by the
ocean takes place, as well as how it has changed over
the years. The importance of monitoring this value
is obvious, as is how it changes and depends on the
state of the ocean surface. We must, therefore, make
systematic measurements of the CO2 content in the
surface waters of the oceans in different latitudes
simultaneously with its content in the air above the
water, in order to estimate the rate of CO2 exchange
between ocean and atmosphere, and to study its
change. The necessary frequency of these measure-
ments in time and space still remains to be deter-
mined. The establishment of the physicochemical
causes of its change will be the primary goal of these
measurements and studies of the rate of CO2 ex-
change.
Global monitoring of atmospheric aerosols is more
complicated. The measurements of atmospheric tur-
bidity recommended and made by the network of
WMO stations are insufficient for distinguishing the
content of stratospheric and tropospheric aerosols,
the effect of which on climate, as indicated above,
differs somewhat. Individual, systematic (at least
once a season) measurements of the content of aero-
sols in the stratosphere up to the 30 km level, using
aircraft and balloons, with determination of the
chemical composition of the aerosols, of the type of
programs now conducted in the USA therefore are
needed (Cadle, 1972). These measurements are
particularly important after major volcanic eruptions.
They also will be urgently needed to establish the
possibility of global regulation of the climate by the
method proposed in paragraph 3.
185
-------
It is very difficult to establish permanent changes
in mean values on a global scale because of the great
changeability and diversity of the aerosol concen-
trations in the troposphere, particularly over areas
of intensive economic activity. Time-consuming proc-
essing of a great volume of data from a dense net-
work of stations is required in order to obtain re-
liable statistical characteristics. Measurements of
atmospheric turbidity, and analyses of filtered aero-
sol samples obtained by the network of WMO sta-
tions in terms of background atmospheric pollution
certainly will be inadequate for this purpose.
Ground measurements of mean global levels of
the content of aerosols in the troposphere can be
made most conveniently in the polar region, for from
the main sources of anthropogenic and natural aero-
sols. Attempts are being made to estimate the global
content of aerosols in the atmosphere from measure-
ments in Antarctica and Greenland, as well as from
those made by research ships on the high seas.
Data on the intensity and special features of global
transfer in different parts of the atmosphere are
needed for these estimates, and can be obtained by
using tracers; radioactive, or those like the inert gas
freon-11 (a halogenated hydrocarbon, CC13F) of
anthropogenic origin, recently measured by a re-
search ship in the oceans of the southern hemisphere
(Lovelock, 1973).
Study and monitoring of the global, more or less
homogeneous, layer of aerosols in the middle, and
particularly in the upper, troposphere, and which are
primarily of continental origin, also can be done by
aircraft sounding in the meridional plane. The lower
level of seasonal aircraft measurements of the con-
tent of stratospheric aerosols indicated above can
be omitted for this purpose.
Finally (and this is very important), detailed
climatological processing and analysis of standard
meteorological observations (temperature and pre-
cipitation in particular) is necessary to distinguish
climatic fluctuations now, and in the near future, as
these investigations are continued for the end of the
last century, and the first half of this one (Rubin-
shteyn, Ye. S., 1973).
All magnitudes measured should be associated as
much as possible with the quantitative models of
global climate discussed above. These magnitudes,
and the frequency and accuracy with which they are
measured, should meet the requirements for Initial
data, and should be capable of being compared with
the resultant calculated climatic parameters of the
models.
BIBLIOGRAPHY
Budyko, Mi. I., 1969, Izmeneniya klimata [Climatic
Changes], Gidrometizdat, Leningrad, 35 p.
Budyko, M. I., 1971, Klimat i zhizri [Climate and Life],
Gidrometeoizdat, Leningrad, 470 p.
Budyko, M. I., 1972, Vliyaniye cheloyeka na klimat [Man's
Influence on Climate], Gidrometeoizdat, Leningrad, 47 p.
Budyko, M. I., 1973, Atmosfernaya uglekislota i klimat [At-
mospheric Carbon Dioxide and Climate], Gidrometeoiz-
dat, Leningrad, 32 p.
Budyko, M. I., 1974, "An Approach to Affecting Global
Climate," Meteorologiya i gidrologiya, No. 2.
Budyko, M. I., Vasishcheva, M. A., 1971, "The Influence of
Astronomical Factors on Quarternary Glaciation," Me-
teorologiya i gidrologiya, No. 6, pp. 37-47.
Budyko, M. I., Vinnikov, K. Ya., 1973, "Modern Climatic
Changes," Meteorologiya i gidrologiya, No. 9, pp. 3-13.
Kondrat'yey, K. Ya., et al., 1973, Vliyaniye aerozolya na
perenos izlucheniya: vozmozhnyye klimaticheskiye voz-
deystviya [The Influence of Aerosols on Radiation
Transfer: Possible Climatic Effects] Izd. LGU, 266 p.
Rubinshteyn, Ye. S., 1973, Struktura kolebaniy temperatury
na severnom polusharii [The Structure of Temperature
Fluctuations in the Northern Hemisphere], Gidrometeoiz-
dat, Leningrad, 34 p.
Federov, Ye. K., 1958, "The Effect of Man on Meteorologi-
cal Processes," Voprosy filosofii, No. 4.
Berger, A. L. 1973. Theorie astronomique des paleoclimats,
v. I, v.II, Louvain.
Cadle, R. D., 1972. Composition of the stratospheric "sul-
phate layer," EOS. Trans. Amer. Geoph. Union, v. 55,
No. 9, 812.
Grobecker, A. J., 1973. United States Department of Trans-
portation. Research program for assessment of strato-
spheric pollution. DOT TST-2.1, Wash., D.C. 22 p.
Junge, C. E., 1963. Air chemistry and radioactivity. Acad.
Press.
Keeling, C. H., 1973. Industrial production of carbon diox-
ide from fossil fuels and limestone. Tellus. v. 25, No. 2,
p. 174-198.
Krey, P. et al., 1973, Project Airstream. Fallout program.
Quart, summary report. HASL-276, p. II-6—11-70.
Lovelock, J. E. et al., 1973. Halogenated hydrocarbons in
and over the Atlantic. Nature, v. 241, No. 5386, p. 194-
196.
Machta, L., 1973. Man's influence on the climate. A status
report. Lecture at the VI Session of Commiss. Atm. Sci.
WMO.
Munn, R. E., 1973, Global environmental monitoring sys-
tem. Action Plan for Phase I. SCOPE Rep. 3, Toronto.
Reiter, E. R., 1971. Atmospheric transport processes. P. 2.
Chemical tracers. U.S. A.E.C. TID 25314, Oak Ridge,
382 p.
SMIC., 1971. Inadvertent climate modification. MIT Press,
Cambr., Mass, and Lond. 306 p.
Schneider, S. H. & Gal-Chen, T., 1973. Numerical experi-
ments in climate stability. Journ. Geoph. Res., v. 78, No.
27, p. 6182-94.
Sellers, W. D., 1969. A global climate model based on the
energy balance of the earth-atmosphere system. I. Appl.
Met., v. 8, No. 3, p. 392-400.
Smagorinsky, J., 1974. Global atmospheric modeling and
the numerical simulation of climate. In Weather Modifi-
cation. Ed. W. W. Hess, Wiley & S, N.Y.
Washington, W. M., 1971. On the possible use of global
atmospheric models for the study of air and thermal
pollution. In Man's Impact on the Climate. Cambridge
MIT Press p. 265-276.
Yamamoto, G. & Tanaka, M., 1972. Increase of global al-
bedo due to air pollution. I. Atm. Sci., v. 29, No. 8, p.
1405-12.
186
-------
Concluding Remarks for the U.S.-U.S.S.R. Symposium on
Comprehensive Analysis of the Environment
Leland D. Attaway *
Comprehensive Analysis of the Environment in-
volves five major problem areas:
• Pollutant Characterization: the identification of
problem pollutants and the description of their
sources, emission characteristics, and transfor-
mation processes.
• Route and Fate: description of the relationship
between pollution emissions and ambient en-
vironmental quality (including intermediate and
ultimate biological locations).
• Effects: the relationship between pollution con-
centrations at the receptor (victim) and its ef-
fect upon that receptor, both health and eco-
logical. Indirect effects upon economic and so-
cial systems must be considered.
• Monitoring and Pollution Reduction Methods:
instrumentation for assessing emissions and
ambient quality, and technological and opera-
tional means for reducing emissions.
• Policy: the setting of emission or ambient pol-
lutant standards, land use prescriptions, etc.,
that is, the goals of comprehensive environ-
mental analysis.
These problems run the range of detail from micro
(e.g., human cell processes) through macro (e.g.,
land use allocations in a large region). We are es-
sentially at the beginning of developing knowledge,
tools and data in all of them. Therefore, one conclu-
sion we can safely draw is: we need advances,.and
international cooperation, at all levels in all these
problem categories.
Dr. Izrael has introduced in his paper on Com-
plex Analysis of the Environment a comprehensive
mathematical statement of the environmental anal-
ysis problem; a mathematical formula which encom-
passes all the above components, at all levels of ag-
gregation from micro to macro. It has two important
potential uses:
(1) Organizational: to identify all essential com-
ponents of the problem; to indicate their interrela-
tionships, so that research products on individual
components are designed to fit one another overall;
to determine the research priorities, by identifying
the strongest and weakest links in the knowledge
chain; and to coordinate the overall research pro-
gram.
(2) Analytic: to actually perform special analy-
ses. However, in its present form the mathematical
statement is more representative than analytical; the
form taken for applications will depend upon the
specific objectives and the state of knowledge at the
time.
We must use some such statement for both pur-
poses—but we must be making applications im-
mediately, and so cannot wait until our knowledge
permits us to use such a comprehensive expression.
Indeed, it is clear that the equation as it stands will
never be used in its analytic role—it simultaneously
has too much scope and too much detail. One or the
other has to be reduced, depending upon the spe-
cific application.
The regional analysis of Lake Baykal, discussed
in the paper System Analysis and Imitation Mathe-
matical Modelling, by Dr. Izrael et al., represents
what can be done today in Comprehensive Environ-
mental Analysis for a region: a group of sector
models (for example: air and water transport; land
use; pollution reduction methods) which can be ap-
plied independently of one another, but which feed
one another inputs. They include all the components
I listed earlier: pollutant characterization; pollutant
route and fate; effects; monitoring and pollution re-
duction methods; and policy, such as standard set-
ting and land use allocations.
The product of applying these tools for a specific
region must be a comprehensive plan for that region
covering the next ten or fifteen years. The environ-
ment is but one aspect of that plan. It should also
include guidance on:
(1) Environmental goals (expressed as ambient
air quality standards, water standards, land-use guide-
lines, etc.).
(2) Effluent standards, such as flue gas scrubber
control guidelines.
(3) Economic growth goals.
(4) Land use.
•Deputy Assistant Administrator for Program Integration, U.S. Environmental Protection Agency
187
-------
(5) Transportation.
(6) Utilities (water supply, power, waste collec-
tion and treatment).
(7) Population growth and distribution.
All these other (than environmental) aspects" must
be such as to meet the environmental goals over the
ten to fifteen year period. However, if the cost-ben-
efit balance point so indicates, then the environ-
mental goals may have to be relaxed; or the popula-
tion goals; or the economic goals. This plan must be
flexible enough to change annually, and it must be
a pragmatic blend of proceeding "today" with the
tools and knowledge available. It will be a blend of
Best Practical Technology application and ambient
proscriptions.
It is not feasible to set pollutant standards inde-
pendently of cost-benefit considerations. All pollu-
tion cannot be terminated, so clearly some pollution
is worth the attendant social costs. The lowest level
of visible organism response will often be too costly
a level for a standard, given the undamaging nature
of the effect. The standards sometimes must be set
at more relaxed levels closer to, but still safely below,
harmful effect levels—at least as an interim measure.
We thus must consider "standards" which change
over calendar time, in order to move closer and
closer to our ultimate goals but at a pace which will
not unacceptably damage the economy. But we must
converge to those ultimate goals.
Within the context of my earlier (just above) re-
marks, we would suggest the following candidates as
high priority joint projects. In all cases, maximum
use should be made of the product from Working
Groups already established within the rest of the
Environmental Agreement. Indeed, the principal role
of this Comprehensive Environmental Analysis Proj-
ect should be stimulation of integrated working group
efforts. The candidates are:
(1) Pathways, fates and effects of multi-media
pollutants, and related multi-media standards.
(2) Requirements for multiple standards regulat-
ing long and short term doses.
(3) Responses of aquatic and terrestrial ecosys-
tems to pollutants.
(4) Analysis techniques and measurement meth-
ods for biological and genetic effects.
(5) Techniques for evaluating socio-economic
costs and benefits.
(6) Evaluation of worth and feasibility of a joint
system for assessing pollution concentrations in the
biosphere.
•ft U.S. OOVEINMENT MINTING OFFICE i 1976 O—587-798
188
------- |