*** ^ UNITED STATES ENVIRONMENTAL PROTECTION AGENCV
>
WASHINGTON D.C. 204SO
April 18, 1986
QFF.ce o*--
THE ADM I NlIJTP* * TQ
Honorable Lea M. Thcmas
Administrator
U. S. Environmental Protection Agency
401 M Street, S. W.
Washington, D. C. 20460
Dear Mr. Thomas:
The Water Quality Subcommittee of the Environmental Effects, Transport
and Fate Committee of the Science Mvisory Board has cortpleted its review
of the Agency's Ambient Water Quality Criteria for Dissolved Oxygen — Fresn
Water Aquatic Life, The Subcommittee addressed the issue of whether the
docLcnent comprises a scientifically adequate discussion and evaluation of
the scientific literature concerning dissolved oxygen in fresh water aquatic
systems.
The Subcommittee assessed six major scientific issues including; the
invertebrate problem? laboratory-field implications; additive stresses ana
enemies! interactions? growth rate reductions; oxygen criteria levels -, and
dissolved oxygen monitoring conditions. The Subcommittee reccrnmended tnat EPA
staff make various vnodif ications to the treatment of these and otne-r issues,
In general, however, the Subcommittee concludes that the docuuent is well-
organized and research whose logic and conclusions are scientifically
defensible.
We appreciate the opportunity to review this document and request an
Agency response to the attached report.
Sincerely,
John Neuhold, Chairman
1 later Quality Si±>cotrmittee
tnvironnsental Kffects, Transport
and Fate Committee
Norton Nelson, Chairman
Executive Committee
-------
SAB-EETFC-86-02Q
SCIENCE AD/ISQRY BOARD REVIEW OF THE
HATE? QUALITY CRITERIA DOCUMENT FOR DISSOLVED CKYGSN
'.later Oualicy Subcommittee
Environnentai fvftscts, Transport aod Fate Ccnmittee
Science Aa^isorv Boara
April 1986
-------
EPA NOTICE
This report has been written as a part of the activities of the Science
Advisory Boara, a public advisory group providing extramural scientific
information and advice to the Mrrdnistrator and other officials of the
Environmental Protection Agency. The Board is structured to provide a
balanced expert assessment of scientific natters related to problems
facing the Agency. This report has not been reviewed for approval by
the Agency, and hence the contents of this .report do not necessarily
represent the views and policies of the Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or teeanrendation for use.
-------
Table of Contents
Roster of Water Quality Subcommittee ,. i
x,
Roster of Environmental Effects, Transport and
Fate Committee ,,.*.,,...,,, i............. i i
r
I. Execut ive Sunraary ..,..,,.,.....,,.......,.... .\ 1
\
II. Introduction ................................................ 3
A. Historical Perspective 3
B. Charge to the Subcommittee ...,,..*. 4
III. General Cdn^ients on the Dissolved Oxygen
Criteria EDcunent 5
IV. Philosophical Issues for tne Criteria Docunssnt ..,..,,.,,,... 6
V. Review of Scientific Issues in the Criteria DDCL
A. The Invertebrate PrODlem
3. labcratcry-Fielci L^j.icatiC'1'? ,..,...........*....* , 9
C. "Ccn-ti'B 5tL"^5?!;--5 5ix Cu.t^ij.cal Interactions .,„,,...,,.., •.„
D, Growth Pane Reduction , -,, 12
E. Oxygen Criteria Levels 13
F. Dissolved Oxygen Cnt3ria ana >!onitoring Conditions ..... 1-
1. Continuous Monitoring 16
2, Grab Sainpling ..., ,..,..».,. ^"*
3. Fourier Analysis . > .,...,.,... 18
G. References 20
-------
U. £. ENVIRONMENTAL PROTECTION
SCIENCE ADVISORY
Environmental Effects, Transport and Pate Committee
Water Quality Criteria Subcommittee
Or* John M. Neuhold, Chairatan
•College of Natural Resources
Utah State University
Logan, Utah 84322
Or. Terry F. Yesie, Director
Science Advisory Board
United States Environmental
Protection Agency
401 M Street, S.W.
Washington, 'D.C. 204tsO
Dr. Melbourne R. Carriker
Professor of Marine Science
College of Marine Studies
University of Dslaware
Lewes, Delaware 19953
Dr. ton Coble
Director
Wisconsin Co Operative t'lshery
Research Unit
University of Wisconsin
Stevens Point, Wisconsin 54481
Or. Rolf B. tertuny
Professor* Environmental
and Industrial Health
School of Public Health
University of Michigan
Ann Arbor, Michigan 48109
Dr. Kenneth Jenkins
Prot*assor of Biology
California State University
at Long Beach
Long Beach, California 90804
Er. Charles Norwood
4958 Escobedo Drive
Woodland Hill, California 91364
Dr. Bernard Patten
Professor
Department Of Zoology
University of Georgia
Athens, Georgia 30602
-------
11
SCIENCE AWISQKY BOARD
MS PATE COMMITTEE
Dr. Rolf B, Rartung, Chain-ran
Professor, Environmental
and Industrial Health
School of Public Healtn
University of Michigan
Art»r» Michigan 48109
Dr. Terry F. Yo$ie, Director
Science Advisory Board
United States Environmental
Protection Agency
Science Advisory Board
Washington, D, C. 20460
IlL-lBEPS
Er. Martin ^lexancer
Professor
0=rert-*ient of Agronomy
Cornell dm, varsity
Ithaca, ^ew York 3,4853
Dr. Wilford R, Gardner
Head, Deparatroent of Soils,
water, and Engineering
University of Arizona
Tucson, Arizona 85721
Or. Robert Huggett
College of William and Mary
ttairrnsnr Department of
Chemical Oceanography-
Virginia Institute of Marine
Science
Gloucester Point, VA 23696
Cr. Kenneth Jenkins
Professor of Biology
California Stats University
at Beach
Long Beach, CA 90804
Dr. John Laseter
Envire-Health
Suite 800
990 North Bowser Rose
Richardson, Texas 75031
Or. John Neuhold
Department of Wildlife science
College of Natural Resources
Utah State University
Eogan, Utah 84322
-------
I. Executive Surnary
A water quality criteria document assesses and articulates the
scientific basis for establishing levels of a substance at which specific
organism end ecosystem effects will result. Selecting the level of that
substance permitted in water systems represents a policy decision
based upon decision maker judgments of acceptable risks. In view of
the fragmentary nature. of the evidence concerning this risk, the Subcom-
mittee Tecawenos EPA maintain a conservative approach by upwardly revising
criteria levels for oxygen,
The levels for dissolved oxygen stated In the criteria document derive
primarily from laboratory toxicity studies of fish species, while field
studies of fish and invertebrates, which provide the food base for these
fish, receive raininal discussion. The Subcotnmitttae recommends that EPA
give greater eirpnasis to available published field studies. If EPA judges
the current field'study as too limited, it should carry cut
anfl/or sponsor research to obtain the needed information.
Chemical interactions and additive stresses are critical factors in
establishing criteria levels for oxygen and other substances. The Sub-
committee recommends that EPA provide a more thorough discussion of how
these factors were considered and integrated with other information in
preparation of the oxygen criteria document.
Inconsistencies appear in the documentation relating to growth rate
reduction by oxygen deprivation and what is termed to be "slight,
or severe" growth i
-------
- 2 -
The Subcommittee recommends that EPA upwardly revise the criteria levels
presented in Table 6 of the criteria document. This recommendation is
based upon a recognition of the limitations caused by the lack of data for
organisms other than fish, the possibility of deleterious effects of lower
dissolved oxygen levels (besides those determined for fish), the fact that
the proposed dissolved oxygen levels that would iittpair fish production and
the likelihood of deleterious effects on fish resulting fron the interaction
oetween aissovled oxygen levels and other stressors-
Effective monitoring is essential if the states are to successfully
irplerrent water quality criteria* Ihe sampling and analysis plans for
"Tonitoring data should, therefore, conform to the biological criteria. In
the case of a criteria that is expressed as a periodic function, the sampling
plan should require; sarnies to be equally spaced in tirse, sarpiing rates
that are Mgh enough to 3'"cid aliasing, and total sanple sizes large enougn
so that the lowest nrequency o: unpettance can be 'observec.
The monitoring guidance included in the dissolved oxygen criteria
document would be* significantly unproved by supplying guidance on saaplinc,
ambient dissolved oxygen. The suggested calculation procedure, while
different fron the classically used technique, is reasonsable if modified
to account for monitoring periods longer than one week, and if based on a
sampling frequency that can detect significart deviations from the function
that defines the no unacceptable effect level.
-------
- 3 -
II. Introduction
A, Historical Perspective
The Federal government promulgates water quality criteria for the
purpose of controlling pollutant tenacity in aquatic ecosystems. Hie
intent of these criteria is to control excess concentrations o£ a tox-
icant by establishing maximum permissible levels that will protect aquatic
life. Scientists and statt and Federal agencies do not define oxygen in
the aquatic system as a toxicant and, rather than being concerned with
having oxygen* they are concerned with not havir^ enough to protect
and/or sustain aquatic life*
Oxygen is an essential elenent for the maintenance of aerobic life
in any biotic system. Oxygen enters the aquatic environment through
diffusion frcm tne atnospnere ana is produced by aquatic plants in the
process of photosynthesis. Pollutants in the aquatic environment, par-
ticularly organic pollutants, provide the substrate for the production
of roicrobial ccnsnunities which use the oxygen in their metatJoltc activi-
ties carpets with fish, aquatic invertebrates and other aquatic orga-
nisms, Waste heat, when discharged into aquatic systems, also has the
direct effect of reducing the amount of oxygen that can dissolve in
water • column and thus affect its availability to organisms. It also re-
duces oxygen indirectly through creating a temperature environment amenable
to the proliferation of microbial organisms.
Since organic materials and heat cause the reduction of oxygen in
water columns, a logical step would be to set upper limits for them.
-------
- 4 -
however, the organic materials are many and varied, jnd are often not toxic.
though maximum heat tolerances can be defined for various species, tnese
levels are often greater than those that will materially affect oxygen
concentrations. Thus, oxygen becomes the substance £i_ which criteria
*•
\ ;
are established, and the criteria express the result of ccrapiex interactions
\
Of pny£icalr chemical and biological factors, In turn,- these oxygen levels
produce effects on organisms within the ecosystem and, consequently, on the
eccsyst™ itself.
The criteria development strategy used by the Environmental Protection
Agency sets limits for oxygen concentration in the water column to protect
aquatic organisms as well as the eeosystsn. This strategy was first fonsula-ec
in a criteria format witn the publication of Vvater Quality Criteria 1972
f\AS/.*v3C 1973) and implemented by EPA, four years later in Qual i ty Criteria
£cr water (EPA 1976). Section 304.(a)(l) of the Clean Water Act [33 U.S.C.
13i4(a)(l)] requires that E?.% publisn and periodically upcste s-oient
quality criteria reflecting the latest scientific information available on
the effects of pollutants on public health and welfare, aquatic life and
recreation. The draft document, Mbieot Water Quality Criteria for Dissolved
Qxyggm Fresh-Water Aquatic Life , critiqued in this review is EPA's latest
effort to fulfill the requirements of the Clean Water Act for tins particular
criteria*
B. Charge to the SuDcotimittee
On July 29, 1935 the Division ot Criteria and of the Otfice
ot Water requested the Science Advisory Board to review the draft document,
for the purpose of improving its scientific quality. The Executive Corroittee
-------
- 5 -
of the Science Advisory Board accepted this request and assigned the
of reviewing the document to the Water Quality Subconmittee of its Environ-
mental Effects, Transport and Fate Coramttee.
Members of the Subcotnwttee received copies ot the document in
September, 1935 and net in public session to review it at EPA's Environmental
Research Laboratory - Duluth on October 10-11, 1985. The Subcoreattee
discussed and evaluated the scientific adequacy of the document with its
principal author and other EPA staff and arrived at a concensus during the
neeting. Subsequently, individual Subcotteiittee members wrote individual
sections of this report and subjected their analyses for review and editing
by their colleagues. This final report articulates the scientific concensus
derived curing the public ineeting and follow-up discussions of successive
draft r^pcrzs,
If-ft report is organizes into tnree najor sections involving general
comments, and a aiscussion of philosophical issues and scientific issues,
III. General Comments on the Dissolved Oxygen Criteria Dscunsent
Ihe Subcommittee concludes that this is the best prepared criteria
document the Science Advisory Board has reviewed to date. The document
is well-organized, researched and referenced. Generally, its arguments,
logic and conclusions are scientifically defensible although its premises
w*re sometimes arguable. Hie overall quality of the thinking, research and
writing is cownendable.
There are several areas in which the document can be further improved.
Ihese include! 1) the philosophical basis for a criteria docupient that
reflects, in part, how the material in such a document is presented, and
-------
_ 6 _
2) the validity of some of the premises upon which the scientific arguments
rest.
IV, Philosophical Issues for the Criteria Document
A criteria document is a scientific document in which scientific data
are presented ana arguments synthesized to provide evidence for the effect
of seme substance on organisms and systems of organising. This particular
document concerns the effects of lowered oxygen concentrations on freshwater
aquatic life. The levels that affect, partially affect or do not affect
aquatic life define criteria levels that are supported by evidence from
the scientific literature baseo upon laboratory experiments anc field
observations.
The choice of one of these criteria levels as the level at whicn it
is either pemissiDle or net perrassibie to allow an effect is a policy
decision, The Subcgranitt^ oelieves that the document blurs the distinction
between -scientific ana po.icy issues and choices. Specifically, on pages
24-25, criteria levels are aefined for proauction utpaxoient of organisms
during "embryo and larval stages" «nd "other life stages" for salronid
and non-salronicl waters* These levels are supported by documentation in
previous pages although it is questionable whether the ordinal lavels of
growth unpaitttient used (slight, moderate, severe) have any real meaning,
On Table 6 (page 26), however, the criteria levels selected fall into the
"impair.ent allowed" area. Table 6, therefore, presents a policy d
-------
_ 7 -
The Subcommittee believes that the water quality criteria for
dissolved oxygen should be designed to protect the entire aquatic ecosystem*
Fish are only one component of this ecosystem, but they receive virtually
exclusive attention in the criteria document. The justification for this
focus includes: 1) the assntption that the protection of fish will also
ensure the protection of other organisms or groups of organisms, and 2)
little information exists upon which to inclusion of biota in
considerations underlying the proposed criteria. The first of these
assumptions is not supported by documentation;; it is an article of faith
(but a strong one since roost scientists would accept the premise that
health of a *?ater body is reflected in the status of its ichthyofauna),
The second is largely true, but EPA should what literature exists
ana state the conclusicsns that could be dra^n from it. The relative lack
of information about microbxal, plant and invertebrate responses does not
justify completely ignoring them.
The ecosystems in question are complex, organized networks of biotic
and abiotic interactions. About such networks, including food webs and
feedback control loops, there is growing knowledge that indirect effects
propagated over time may be more important than instantaneously expeciencea
direct effects* The dissolved oxygen criteria document focuses on the
direct effects of oxygen upon fish. Such direct (but unknown) effects on
other groups nay propagate as the result of indirect effects to the fish
over a long period o£ time through the ecosystem's interconnecting network.
Ihus, after a period of years, fish populations could becone seriously
iimpaeted by slowly induced changes in other groups even though no standarcs
based on fish requirements were ever violated.
-------
- 8 -
Given the goal of protecting all biological foms ana the integrity
of their vital interactions witnin the aquatic ecosystem, the responsible
pragmatic response to the lack of relevant knowledge is to employ a
conservative approach. Accordingly, the low dissolved oxygen levels reeetn-
nended in Table 6, justified solely by ichthyofaunal considerations* shoulc
be adjusted upwards to achieve at least the protective level afforded by
the present Redback standards.
V. Review of Scientific Issues in the Criteria Document
A, ihe Invertebrate S>rQbls*n
The concensus position of trie cocument states that if all stages of
fish are protected, invetebrate ecrtnumties, although not necessarily
unchanged, should be adequately protected. The document does not state the
scientific support for this pcsiticr.. Rather, it asserts that some invertebrate
species are as sensitive as "rccerately susceptible" fisn.
Because biotic camr%initi-ss consist of many species at various tropnic
levels, and since the role of these species in the metabolism of the ccnnumty
is not fully kno*rn with reference to the production of food resources far
fin fish, it is important to recognise the potentially significant role of
at least "dominant" invertebrate species and their tolerance to low levels
of dissolved ojcygen.
At a minumim, the document should refer to the literature or. require-
ments for dissolved oxygen by frose invertebrate sjfecies that have been
studied. Groups of invctebrates fat should he checked ana that are xpportant
directly as prey species for fish with ccranercial or recreational value, or
indirectly as important processors of energy in the acosysten.
-------
— Q «
unionid clams, fingernail elsas, Amphipods, crayfish and insects of
orders of Plecoptera and Ephemeroptera. Hie document reports that few
appropriate data are available that address the effects of reduced dissolved
oxygen on freshwater invertebrates, fhese data should, nevertheless, be
reviewed and briefly summarized.
B. Laboratory-Field Implications
It is not at all clear that the methods used to achieve oxygen
levels in experutental units in order to determine oxygen requirements for
fish and invertebrates did not produce sparious results evoking other
biological responses. A statement clarifying the metnoaology would be
usetul.
Laboratory tolerance tests for oxygen are most often designed to
hold all extraneous variables constant to ascertain the main effect of
oxygen. Even if efforts are rade to determine the interactive effects of
other physical, chemical and biological factors with Oxygen levelsf
scientists se-ldcm succeed in duplicating the natural environment experi-
mentally, Thus, the comparability between laboratory and field results
remains unknown unless field observations are designed to verify laboratory
results. A discussion of the advantages and disadvantages of laboratory
tests and field observations in the document's "Introduction", while
eliminating a discussion of disadvantages after each of the sections,
would considerably enhance the scientific presentation of the document.
Coupled with this addition to the Introduction should be removal or
raodificafcion of the negative statements in the text, The available data
are much better than the? document suggests. Numerous laboratory studies
-------
-10-
of fish in which experimental conditions (e.g., sp cies, temperature,
duration) varied have produced conclusions for the -.effects of dissolved
oxygen upon fish for a number of variables. These results provide a stronger
s
basis upon which to develop confidence in the data ir.an if all the experirtents
had uniform designs. Added to this factor are the results of the field studies
which support, to a large extent, the results of the laboratory studies,
The combination of laboratory hypothesis testing and field verification is
a very powerful one*
Statements in the document complaining about the variability of testing
conditions, suggesting that scientists investigating dissolved oxygen had
unusual */hins and preferences, stating that the data base is fraught with
inconsistency, and denigrating metabolic and physiologic studies because
tney require extrapolation and assumptions, are unjustified anc counter-
productive. Similarly, statements that belittle laboratory studies at the
beginning of the presentation Of laboratory results and field studies after
the presentation of field results do nothing but reduce confidence in the
document and should either be eliminated or rephrased to emphasize the
power of similar conclusions from varying laboratory and field approaches.
The negative tone referred to above also gives the impression that
violation of the criteria is permissable. Specifically, pages 28-29 include
such statements as "Scene deviation below acceptable concentrations would
probably not cause significant harm"; discussion of significance {iitportance)
of conditions that fail to .meet reccnmended criteria"; and "excursions
below minimum recommended values are likely to be appreciable". Such
statements cast doubt upon how seriously EPA considers the criteria and, in
the opinion of the Subcommittee, should be stated more carefully.
-------
- 11 -
C. Additive Stresses and Chemical interactions
The dissolved oxygen criteria document breaks irnportant ground
for EPA in that it attenpts to deal with the effects of both physical and
chemical stresses on the sensitivity of organisms to low dissolved oxygen
concentrations. Particular emphasis is placed on the adverse effects of
high temperatures and tolerance to low dissolved oxygen. The document
states that "Concern for this temperature effect was a consideration in
establishing these criteria, especially in the establishing of those
criteria intended to prevent shore-term lethal effects" (page 19).
Section VI elaborates on this statement by noting that "The dissolves
oxygen concentrations in the criteria are intended to be protective .at
typically hign seasonal environmental tenperatures" even though these
temperatures "are often higher than those used in the research from which
the criteria were generated" (pages 25-26).
In spite of this assertion, the document does not discuss the inethocs
used to factor the temperature effects into the actual criteria numbers.
Instead, the reader is told that the criteria derive from the production
unpaiment estimates (pages 24-25), "which are in turn based primarily
upon growth data and information on temperature, disease and pollution
stresses." It is appropriate and important to incorporate a discussion
on additive stesses in a criteria document, but these data should not be
factored into the final criteria numbers without Clearly defining the
underlying rationale. Without that rationale, it is not possible to
critically evaluate the inclusion ot these data. The alternative is to
accept at face value that the criteria numbers adequately address the
problems of temperature, disease and pollution stresses.
-------
In general, the discussion'of additive chemical stresses on sensitivity
to lew concentrations of dissolved oxygen is well presented. It includes
a review of data on metals (2n, Pb, CuS* xenobiotics (roonchyaric phenols),
cyanide and amnonia. More importantly, though the document the
inportance of chemical stresses in section IV Br it is mentioned only in
passing in discussion of the National Criteria in Section VI, and is not
at all in the discussion of the relationship between criteria
and jnomtoring and design conaitions. These deficiences should be remedied*
This document successfully articulates the complex issues of eh«nical
interactions and aaaitive stresses, issues that should receive consiceration
in all of the criteria aocuraents developed by EPA. More tnorough discussion
of hew these data were appliec in deriving the criteria numbers and how
tney should be used in monitoring ana experimental aesign, however, would
help to integrate and strengthen the document.
D. Growth Pate Recuetien
The document Should acecrepany aata entries in Tables 1-5 vvith a
fuller discussion of the variation surrounding the values because, as
currently presented on page 6, the arguments are not well substantiated.
One can conclude from Table I that a "slight" growth impairment ocears at
an oxygen concentration of ? rag/1 and that growth unpainBent ot 4%-9% at
6 rng/1 might better be temed "noderate" for the following reasons: 1)
the absolute weight of fish {and probably invertebrate) flesh lost in
natural waters can be significant at such levels, particularly considering
the ee-onojnic loss of coraiws re tally exploited populations such as the
salmon produced in the Columbia River drainage basin? and 2) the likelihood
-------
- 13 -
of temperatures in natural waters exceeding those of the experiments
(Table I,) during critical periods of low dissolved oxygen. At dissolved
oxygen levels below 6 rag/I, growth of salmoruds was reduced by 10% or
more. One should conclude from this table that values of dissolved
oxygen concentration listed at the top of page 25 for slight, moderate
and severe production impairment should be 7, 6 and 5 rag/1, respectively.
E. Oxygen Criteria levels
The Suoccnmittes believes that the dissolved oxygen criteria in
Table 6 are too low because the criteria listed for the 7-day mean minimum
oeriva fron studies wnich have shown subletnal deleterious effects on
fisn* According to the statement on page 24, nearly all data on the
effects of low dissolved oxygen on organisms relate to continuous
exposures for periods cf hours to weeks, A 7-day mean minimum is a
^arioc of hours to s, wee** Therefore, the deleterious effects observsd
in die various stuaias should be expected to occur in a 7-day period. In
fact, according to the conclusions stated on page 25, the 7-day mean
minimm values recommended in Table 6 would cause moderate impairment of
production. Such values can hardly be protective of our nation's aquatic
resources.
Moreover, the recomnsrtded dissolved oxygen criteria derive mainly
frcm laboratory studies in which stresses that occur as the rule rather
than the exception in nature were eliminated and other stresses peculiar
to the laboratory setting *ere substituted, thus creating substantial
-------
- 14 -
uncertainties. Low dissolved oxygen exacerbates effects of such earsonly
occurring stresses on fish as temperature fluctuation, presence of ubiquitous
pathogenic organisms, competitive interactions, and presence of toxic
chemicals (pp. 19-21). Based upon studies which eliminated the
dissolved oxygen criteria cannot be to be protective for the
interactions of low or marginal dissolved oxygen levels and environmental
stresses that occur commonly in nature.
Under these circumstances the Subcommittee concludes that current
scientific data derronstrates that the proposed dissolved oxygen criteria
are unacceptably low ana recamsnas that higher (i.e, mare protective)
values supported by the infonstion on 24-25 be substituted as in
the taole below:
Quality^...Criteria tor Ambient jassolved Oxygen Concentration
30 aay mean
7 day mean
7 day mean
minimum
Cold Water
Earlv Life
NA
11.0 (8.0)
NA
1 day minimum 9,0 (6.0)
Criteria
Other Life
8.0
NA
7.0
3.0
Warm Water
Early Life
NA
6.5
NA
5,5
Criteria
Other Life
6»0
NA
5,0
3,0
F. Dissolved Oxygen Criteria and Morutorxng Condition's
One of the best features of this criteria document is its discussion
ot the biological effects of cyclic and other variations in dissolved
oxygen. The docunent points out that dissolved oxygen varies continuously
in a oail/ c^ele, and it clear that and annual cycles exist.
-------
- 15 -
The ambient value cf dissolved oxygen consists of the sum of these natural
cycles* plus the effects of any human activities. the criteria document
also notes that the sensitivity of some species to low dissolved oxygen
levels varies during the course of their life cycle.
Since dissolved oxygen levels, and the organism's sensitivity to
these levels, follows cycles, the natural way to describe the dissolved
oxygen'criteria is as a periodic function that does not produce unacceptable
effects. Assuming that the acceptable dissolved oxygen level represents a
sinusoidal function of time, the criteria would be expressed as an average
value, an amplitude, and a period. Ihese three numbers would determine
the mininformally acceptable dissolved oxygen value, the average acceptable
value, and tne tuna between minima.
The dissolved oxygen criteria rest on the assumption tnat the acceptable
effect level defines a function whose dominant coopcnent is a sinusoid
with a perltd of one day. For example, to protect the early life stages
of cold water species, tins assumption can be e>£pressed as:
D0(t) * 6.5 + l.S oos C(2V24)t] -«- £ (I)
where D0(t) is the acceptable dissolved oxygen level in lag/l, t is tire
in hours measured from a maximum, and 5 > -2.
Section IV of toe criteria document addresses the mpjrtant problem
of calculating with ironitoring data to determine whether anbient dissolved
oxygen is below the acceptable levels described by an equation like (1).
In the following paragraphs the Subcommittee discusses monitoring issues
raised by the calculation approach developed in Section IV.
The calculation approach suggested in the criteria document assumes
that dissolved oxygen is measured twice daily, once at the maximum and
-------
- 16 -
once at the rtaninum value. The example calculations (Table 7)
assume the availability of only seven pairs of observations, and fron
\
these pairs are dtfined; '
(1) daily maximum = roinuraizn of dissolved oxygen and
air saturation concentration?
(2) daily mean = average of that day's mmroum and maximum?
(3) 1-day minimum = minimum of the minima,
(4) 7-day re an minimum = average of 7 minima? and
(S) 7-day mean = average ot 7 daily, means.
The criteria are violated if any of the quantities 2 through 5 fall ssiew
values judged to be acceptable.
The guidance on monitoring coes not specify the sampling plan fir
acquiring the monitoring data. In the following paragraphs t**o saraling
plans will be considered. The first monitors dissolved oxygen continuously,
with the required maxima and minima found from the observed time history,
The Second possibility assumes that dissolvea oxygen is measured by grsts
saiftples taken at widely spaced time intervals.
1, Continuous Monitoring
With continuous retutoring, the suggested calculation procedure is
reasonable if the concentration of ambient dissolved oxygen is nearly
sinusoidal with a period of twenty-four hours. However, the calculation.
procedure does not address hew to deal with more than seven pairs of
points, as will be the case if -monitoring lasts more than a week. Mor
does it address cases where the observed time function is either not very
sinusoidal, or has a frequency greater than one cycle per day.
-------
- 17 -
If the concentration of amoient dissolved oxygen is periodic, but
not sinasoidal, it nay happen that the daily minimum value persists for
unacceptably long periods. If ambient dissolved oxygen has components
the periods of whit* are less than one day, it is possible the daily
minimum will be achieved more than once.
Based on the data presented in the criteria document, both of these
situations could lead to unacceptable effects, while not necessarily
violating the criteria. However, when ambient dissolved oxygen is eontin-
uously monitored, sufficient information exists to determine, on a site
specific basis, what modifications (such as limiting the tune at the
minimum) might be required to avoid unacceptable effects.
2. Grab Sampling
When anbient dissolved oxygen is nonitored by grab sampling at widely
spacea intervals, the calculation procedure may produce misleading results.
By saiiiplIrG; out of phase with the minima and maxima, it is possible to
consistently obtain a pair of daily neasutements that correctly estimates
the average concentration but give essentially no information about the
true anplitude of fluctuations around the mean. If as few as two samples
per day are taken, events with frequencies higher than one cycle per day
will actually appear as components with frequencies lower than one cycle
per day. This could lead to a misinterpretation of the detection of adverse
effects.
Another problem with estimating the criteria's parameters with widely
spaced sazrples results from the paucity ot information about the length
of tune over ^hich ambient dissolved oxygen remains near critical concen-
trations, since many physically possible continuous functions will pass
-------
- 18 -
through the measured points.
Frcm the discussion in the previous paragraphs? the Subccrmittee
concludes that the suggested calculations (assuraing monitoring lasts only
one week) are reasonable when the daily naxima and minima are sailed rapidly
enough to detect deviations that have unacceptable biological effects,
However, EPA could significantly unprove the monitoring guidance by
explicitly addressing the sampling rate tnat is required to determine the
acceptability of ambient dissolved oxygen levels. In the next section,
the Subcommittee discusses a classical approach to determining this rate*
3. Fourier Analysis
The purpose of monitoring dissolved oxygen levels is to verify that
aiibiant levels satisfy the bounds expressed by a biologically determined
sinusoidal Sanction like equation (1). To do this requires estimation of
the r>ean, amplitude, ana frequency cf the atibient concentration, and
comparison of these values witn chose judged acceptable. The classical
approach to tins problem is called Fourier analysis*
Fourier analysis expresses a function, X^* sampled at equally spaced
time intervals, as the sum o£ periodic conponents;
\ » BQ +! EJLjCQsCw-jt) 4- RjSintijt)] (2)
wtere 0 < 3 < n/2,
0 < t < n - 1,
itf -* 2ir j/n,
, and
AO is the mean ot the Xt, and R, is the amplitude o€ the jth ccmponent.
Thu analysis solves for the n coefficients of (2} using ordinary least
squares.
-------
- 19 -
By using Fourier analysis, it is possible to icentify the amplitudes
of selected frequencies with periods between two tines and n tunes the
sainpling rate. To determine a sampling plan for the Fourier analysis of
data, it is necessary to select a sampling interval.,A , so that no
frequencies higher than Q.5A are present. One also needs to select n,
the number of samples, so that n times A is equal to the fundamental period
of Xt.
Because the sum of the R, is proportional to the mean square of the
X(-, under appropriate statistical assumptions, it is possible to use
regression techniques together with the magnitudes of the R 2 to determine
rfhich of the w^ contributes most to the total variability of the Xt.
One of the advantages of using Fourier analysis to examne a tiTO
series for periocic components is the ability to know the consequences of
nisspecifying the sampling interval and the fundamental period, fcfoen the
sampling interval A is too large, frequencies higaer than 0.5 A are
aliased down to frequencies less than 0.5 A . For eacn frequency greater
than Q.5 A , the alias can be explicitly identified. Wien the period of
the sampled data is incorrect, there are frequency components that are
not equal to one of the w -j- This effect is called leakage and, as with
aliasing, its effect can also be precisely calculated.
Depending on the frequency band of interest, one should process
several thousand data toints in a Fourier analysis. Historically, this
posed a practical limitation on the use of tins technique. However, many
current software packages can perform the required calculations, accounting
for both aliasing and leakage.
-------
- 20 -
References
Bloomfield, Peter, Fourier Analysis ofTinte Seriest An Introduction,
New York; J» Wiley and Sons, 1976.
National Academy of Sciences-National Research Council, Water Quality
Criteria 197_2, Washington, D, C.: National Academy Press, 1973.
United States Environmental Protection Agency, Quality Criteria for
'Water, Washington, D* C.t Government Printing Office/ 1976.
------- |