&EPA
United States
Environmental Protection
Agency
Environmental Research
Laboratory
Athens GA 30613
EPA/600/9-88/001
May 1988
Research and Development
Fate and Effects of
Pollutants on Aquatic
Organisms and
Ecosystems:
Proceedings of USA-
USSR Symposium,
Athens, Georgia,
October 19-21, 1987
-------
-------
EPA/600/9-88/001
May 1988
FATE AND EFFECTS OF POLLUTANTS
ON AQUATIC ORGANISMS AND ECOSYSTEMS:
Proceedings of USA-USSR Symposium,
Athens, Georgia, October 19-21, 1987
Edited by
Robert C. Ryans
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
ATHENS, GEORGIA 30613
-------
DISCLAIMER
The information in this document has been funded in part by the United
States Environmental Protection Agency. Papers describing EPA-sponsored re-
search have been subject to the Agency's peer and administrative review, and
have been approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use by the
U.S. Environmental Protection Agency.
ii
-------
FOREWORD
For a decade and a half, cooperation and exchange of scientific informa-
tion under the USA-USSR Agreement on Cooperation in the Field of Environmental
Protection has helped both countries in their efforts to control environmental
pollution. These efforts are pursued in recognition of the international
nature of the problem: pollution knows no boundaries.
Three projects are carried out under the joint USA-USSR Agreement's
Working Group on Cooperation in the Area of Water Pollution Prevention.
These are: Project 02.02-11 "River Basin Water Quality and Management,"
Project 02.02-12 "Protection and Management of Water Quality in Lakes and
Estuaries," and Project 02.02-13 "Effect of Pollutants on Aquatic Organisms
and Ecosystems: Development of Water Quality Criteria."
Over the years, scientific delegations and individual scientists have
traveled to each other's countries to visit scientific institutions, perform
joint research, and exchange technical information. This Proceedings presents
associated with the most recent formal symposium, which was held in Athens,
Georgia USA, in 1987.
Rosemarie C. Russo
Director
Environmental Research Laboratory
Athens, Georgia
iii
-------
IN MEMORIAM
This document is dedicated to the memory of Dr. Nikolai V. Butorin.
Dr. Butorin was among the first USSR scientists involved in the USA-USSR
Agreement on Cooperation in the Field of Environmental Protection. He was
intimately involved with the program over 15 years. His scientific expertise
and enthusiasm for cooperative work between the two countries were invaluable
contributions to the success of the Working Group on Cooperation in the Area
of Water Pollution Prevention.
IV
-------
ABSTRACT
The 14 papers in this proceedings present research by scientists and
engineers of the USA and USSR in three projects under the Working Group on
Cooperation in the Area of Water Pollution Prevention. The Working Group is
a component of the USA-USSR Agreement on Cooperation in the Field of Environ-
mental Protection. Included among the papers are reports of modeling of
runoff of substances from agricultural watersheds, of modeling lacustrine
systems, of modeling toxic pollutant risk to aquatic organisms, and of
modeling tributyltin exposure. Social and economic aspects of water quality
management are examined and an integrated system for controlling water use
and conservation is discussed. Effects of ammonium ions on mineral exchange
in fish and ammonia distribution and excretion by fish are examined. The
first use of a uniform toxicity test in both countries is described and a
system for remote monitoring of ecosystem conditions is presented. Pesticide
exposure is examined through aquatic community studies and in studies of
resistance mechanisms in carp and perch. Buffer capacities of freshwater
ecosystems for metals is examined and the relationship of trace metal body
burdens and gill damage in fish to surface water acidification from atmospheric
deposition is explored.
v
-------
-------
CONTENTS
Relation of Trace Metal Body Burdens and Gill Damage in Fish
to Surface Water Acidification from Atmospheric Deposition
T.A. Raines, C.H. Jagoe, F.J. Dwyer, and D.D. Buckler
Page
FOREWORD iii
IN MEMORIAM iv
ABSTRACT v
SESSION 1. RIVER BASIN QUALITY PLANNING AND MANAGEMENT
V. Saulys and A. Kuzin, presiding
A Screening Level Model for Tributyltin Criteria 1
L.E. Fink
Social and Economic Aspects of Water Quality Management 21
A.K. Kuzin, O.I. Kovaleva, and L.S. Garibova
Integrated Syatem of State Control for Water Use and Conservation
Based on the Example of the Azov-Black Sea Directorate .... 28
L.P. Yarmak
Mathematical Modeling of Runoff of Substances from Agricultural
Watersheds into Bodies of Water 35
L.M. Bondarenko, V.Z. Kolpar, and Yu.M. Plis
SESSION 2. EFFECTS OF POLLUTANTS ON AQUATIC ORGANISMS AND ECOSYSTEMS:
DEVELOPMENT OF WATER QUALITY CRITERIA
R. Schoettger and N. Butorin, presiding
Ammonia Distribution In and Excretion By Fishes 50
D.J. Randall, R.C. Russo, and R.V. Thurston
Effect of Ammonium Ions on Mineral Exchange in Freshwater Fish
and Crustaceans 58
G.A. Vinogradov
On-site Toxicity Testing: Applications in the United States
and Soviet Union 70
M.G. Henry, B.A. Flerov, V.T. Komov, and T.A. Heming
Various Resistance Mechanisms of Carp (Cyprinus Carpio L.)
and Perch (Perca Fluviatilis L.) to DDVF Organophosphorus
Compounds 78
G.M. Chuyko
90
vii
-------
SESSION 3. PROTECTION AND MANAGEMENT OF WATER QUALITY IN LAKES
AND ESTUARIES
R. Russo and A. Nikanorov, presiding
Buffer Capacity of Freshwater Ecosystems for Heavy Metals and
Hydrobiological Parameters Determining It 105
A.M. Nikanorov, A.V. Zhulidov, N.A. Dubova, V.F. Gekov, and
I.Y. Kamov
Community Response of Aquatic Organisms to Pesticide Stress . . . 122
S.J. Lozano
Theoretical and Methodological Aspects of Modeling Lacustrine
Ecosystems „ . . . 133
A.A. Matveyev, A.M. Nikanorov, Yu.A. Dombrovskiy, and
V.V. Selytin
Assessment of Risks of Toxic Pollutants to Aquatic Organisms
and Ecosystems Using a Sequential Modeling Approach ...... 153
R.A. Park, J.J. Anderson, G.L. Swartzman, R. Morrison, and
J.M. Emlen
Remote Monitoring of Ecological Condition of Aquatic Ecosystems . 166
A.A. Gittelson and A.M. Nikanorov
viii
-------
ACKNOWLEDGMENTS
Organizing and presenting a symposium and preparing a proceedings is
frequently a complex task, particularly when participants are from two
countries that are widely separated geographically and have different
languages. The dedication and hard work of Athens Environmental Research
Laboratory staff—particularly that of Ms. Joan 1. Price and Mr. J. Mac-
Arthur Long—in taking care of the myriad arrangements for the symposium are
greatly appreciated. The contributions of Ms. Martha M. Wilkes of Computer
Sciences Corporation who typed the final document are gratefully acknowledged.
The scientists, engineers, and environmental managers who participated
in the symposium, of course, are deserving of primary recognition. Co-
chairmen of the Working Group on Cooperation in the Area of Water Pollution
Prevention are Mr. Valdus V. Adamkus, Regional Administrator of USEPA Region V,
and Dr. Alexander K. Kuzin, Deputy Director of the All-Union Research Insti-
tute for Water Protection. The symposium was divided into three sessions.
Presiding over the session entitled "River Basin Water Quality Planning and
Management" were Mr. Vacys J. Saulys of the Great Lakes National Program
Office of USEPA Region V and Dr. Kuzin. Presiding over the session entitled
"Effects of Pollutants on Aquatic Organisms and Systems" were Dr. Richard A.
Schoettger, Director of the National Fisheries Contaminant Research Center,
Columbia MO, and Dr. Nikolai V. Butorin, Director of the Institute of Biology
of Inland Waters, Borok. Presiding over the session entitled "Protection and
Management of Water Quality in Lakes and Estuaries" were Dr. Rosemarie C. Russo,
Director of the Environmental Research Laboratory, Athens GA, and Dr. Anatoly
M. Nikanorov, Director of the Hydrochemical Institute, Rostov-on-Don. Finally,
the contributions of those who took time away from their busy research schedules
to prepare and present papers at the symposium are gratefully acknowledged.
ix
-------
-------
A SCREENING LEVEL MODEL FOR TRIBUTYLTIN CRITERIA
by
L.E. Fink1
ABSTRACT
Some portions of the Great Lakes ecosystem are exposed to significant
concentrations of tributyLtin, a compounds that is acutely toxic to aquatic
life. This paper reports on an environmental risk assessment to recommend
maximum release rates.
INTRODUCTION
Tributyltin (or TBT) compounds are used, among other applications, in
antifoulant pesticides (USEPA 1987). Of particular environmental concern is
their incorporation into paints that are applied to submerged surfaces in
aquatic environments to prevent the attachment of algae, barnacles, and
other encrusting organisms. As a result of the use of TBT—containing paints
on recreational and commercial vessels, the compound has accumulated in water
and sediment of a number of Great Lakes harbors and connecting channels
(Maguire 1983) and in fish from one harbor (R.J. Maguire, Canada Center for
Inland Waters, personal communication). Thus, there is widespread exposure
to TBT in some portions of the Great Lakes aquatic ecosystem.
TBT is acutely toxic to aquatic lite, particularly molluscs. This is
not surprising, because one important target pest is the barnacle. For
example, the 15-day LC50 tor mussel larvae is about 0.1 ug/L. The 96-hour LC50
for the juvenile chinook salmon is 1.5 ppb, and the 48—hr EC50 for the water
flea, daphnia magna, is about 1.7 ppb.
Concentrations measured in the water column near the sediments in some
Great Lakes harbors are within an order of magnitude of levels determined to
be acutely toxic to embryo-larval and juvenile fish (Maguire and Tkacz 1985).
Thus, potentially toxic levels are accumulating in some portions of the Great
Lakes ecosystem. Because of its widespread distribution in the aquatic eco-
system and its extreme toxicity to many forms of aquatic life, TBT may pose a
•l-Great Lakes National Program Office, U.S. Environmental Protection Agency,
Region V, Chicago, IL, USA.
1
-------
substantial risk to sensitive aquatic populations when used in anti-foulant
paints.
The above considerations have prompted the U.S. Environmental Protection
Agency to undertake a 2-year review of the registration of TBT anti-foulant
paints. The regulatory policy question at issue was: Should the registration
of tributyltin-based anti-foulant paints be revoked or modified to assure pro-
tection of aquatic life in fresh and salt water aquatic ecosystems? Recently,
EPA completed its review and has recommended that:
i. The-registration of TBT anti-foulant paints on recreational vessels
less than 65 feet in length should be revoked.
2. The release rate of TBT from anti-foulant paints cannot exceed 168
ug/cm^-day during the 14-day initial release phase immediately after
freshly painted boats are put into the water. After the initial
phase, the release rate cannot exceed 4 ug/cm^-day.
3. TBT anti-foulant paints are to be applied by licensed pesticide
applicators only.
4. TBT paint wastes are to be disposed of in an environmentally sound
manner.
5. TBT concentrations in fresh waters should not exceed 20 ng/L at any
time.
The purpose of this screening level analysis is to evaluate the environ-
mental consequences of the regulatory proposals outlined above using an ana-
lytical methodology that EPA refers to as risk assessment.
At this juncture, it is necessary to briefly summarize what is meant
here by risk assessment. In evaluating the ecological risks posed by a man-
made chemical, two distinct analyses must be conducted: 1) hazard assessment
and 2) exposure assessment. In a hazard assessment, the toxicological effects
of the substance are studied, and no-effect levels or acceptable risk levels
are derived for the most sensitive organisms and the most sensitive toxicologi-
cal endpoints. In an exposure assessment, the rate at which the most sensitive
organism is dosed is calculated. The rate at which an organism is dosed in
its natural setting is determined by the concentration of biologically available
toxicant in each environmental medium with which it comes into contact, the
duration of contact, and the rate of uptake during that contact. Accumulation
ot the toxicant within the organism is determined by the rate of uptake in
competition with the rates of metabolism and depuration.
In an exposure assessment, one or both of two distinct approaches can be
taken:
1. The exposure assessment can be based on measured concentrations in
the various media with which the organism comes into contact, and
measured loading rates to each medium; or
-------
2. The exposure assessment can be based on estimated concentrations and
loadings using conservative assumptions and simple models.
The latter approach is otten referred to as a mass balance approach. The essence
of the approach is summarized in Figure 1. The former approach suffers from
two drawbacks:
1. It requires a great deal ot environmental and source data. For
diffuse sources, these data are often hard to obtain.
LUAU IN
LOAD LOST
LOAD OUT - LOAD STORED - LOAD LOST
. VOLATILIZATION
. SEDIMENTATION (BURIAL)
. CHEMICAL REACTION
.. HYDROLYSIS
.. PHOTOLYSIS
.. REDOX
.. COMPLEXATION
. BIOCHEMICAL REACTION
.. BIOOEGRAOATIUN
LOAD STORED
. WATEK COLUMN
.. DISSOLVED
.. SUSPENDED SOLIDS
. SEDIMENT
.. DISSOLVED
.. SUSPENDED SOLIDS
. BIOTA
Figure 1. The mass balance approach.
3
-------
2. It has no predictive capability beyond simple projection of existing
trends.
Although the latter approach is also very data intensive, it incorporates
a predictive capability. With this predictive capability, it is possible to
evaluate the response of the aquatic ecosystem to a variety of source control
and cleanup options. Within the framework of this predictive approach, the
gathering of environmental data serves two purposes:
1. To identify emerging problems.
2. To allow an evaluation of the accuracy of model predictions (model
validation).
For many chemicals, the data needed for a detailed analysis of the sources,
transport, transformation, environmental distribution, and effects are not
available.
Fortunately in the case .of TBT, studies conducted by scientists at the
Canada Centre for Inland Waters in Burlington, Ontario, provide much of the
data needed to evaluate TBT transport, transformation and distribution in the
Great Lakes aquatic ecosystem. The data on the distribution of TBT in the
Great Lakes ecosystem also permits a comparison between model predictions and
actual environmental conditions.
To perform a regulatory analysis of the proposed changes to the TBT pesti-
cide registration, it is necessary to answer five questions:
1. What are acceptable levels of TBT in water, sediment, and biota?
2. What are the sources of TBT?
3. What are the loading rates from each source category?
4. What is the loading rate-concentration relationship?
5. Which source category(ies) need to be reduced by how much so as not
to exceed acceptable levels?
To help crystallize the issues, attention is focused on the St. Ciair
River-Lake St. Clair - Detroit River System, one of the most densely industri-
alized waterways in the United States and a heavily used corridor for commer-
cial shipping. But perhaps most important, there are enough boat slips on the
Michigan side of Lake St. Clair to accommodate 14,334 vessels of various sizes,
more than for any of the other Michigan Great Lakes.
Returning to the first of the five questions, EPA has proposed that the
concentration of TBT not exceed 20 ng/L in the water column at any time. (The
method of derivation of the fresh water quality criterion for aquatic life or
its adequacy will not be discussed here.)
-------
It is now necessary to quantify the loading rates to the system from rec-
reational and commercial vessels to see whether these sources can account for
the observed concentrations in the Detroit River. The data used to quantify
recreational and commercial sources are summarized in Table I. The results
of that analysis are indicated in Table II.
Next it is necessary to evaluate the fate of TBT in Lake St. Clair, which
acts like a shallow retention pond for the St. Clair River and has a retention
time of about 8 days. The limnological data used to calibrate the model were
obtained from Dr. Thomas Fontaine III (National Oceanic and Atmospheric Admin-
istration, Ann Arbor, MI, (personal communication). Important physiochemical
and fate rate data for TBT are summarized in Table A-l of the Appendix.
Before getting to the focus of the modeling analysis, it is important to
appreciate the significance of the coefficient describing the distribution of
the compound between particles and water at thermochemical equilibrium, the so-
called Rp value. When this value is divided by the fraction of organic matter
present in the sediments, we have the so-called Koc value. If the total
organic carbon (or TOC) value of any particular sediment is known, then its
Kp value can be estimated from the Koc value. The Koc value, in turn, can
be estimated from the n-octanol/water partition coefficient, which itself can
be estimated from quantitative structure-activity relationships (Leo et al.
1971). Because it is the truly dissolved fraction of the compound of concern
that drives volatilization, diffusion into the sediments and biological uptake,
and because the truly dissolved fraction is calculated from Kp and concentration
values for inorganic, organic and colloidal particles, the importance of an
accurate value for Koc cannot be overstated.
Partition coefficients were based on a KQC value of 9.2 X 10 reported
by Maguire and Tkacz (1985) and the assumption that algae were 35% TOC
(Connolly 1987), colloidal material was 35% TOC (Hassett, J., State University
of New York at Syracuse, personal communication) and that the concentration
of colloidal material was equal to that of the algae.
TBT TRANSPORT-FATE IN FRESHWATER AQUATIC ECOSYSTEMS
Based on the properties summarized in Table A-l of the Appendix, the fol-
lowing tentative conclusions can be drawn. Given its relatively slow rates
of biodegradation, photodegradation, and volatilization, and its only moderate
affinity tor particles, TBT is only slowly removed from the water column by
natural processes in the freshwater aquatic ecosystem relative to simple
hydraulic dillution. Under steady state conditions, the potentially most
significant routes of removal of TBT from the water column after hydraulic
dilution are photodegradation, biodegradation, and association with settling
particles and subsequent degradation and burial beneath the active sediment
layer. In turbulent systems having very slow rates of sediment accumulation,
burial would not be a significant route of removal from the system.
TBT is only slowly converted to dibutyltin (DBT) and then monobutyltin
(BT) via biodegradation, with the most significant rate of conversion occur-
ring in the sediments. Whether the rate of biodegradation determined in a
-------
TABLE I. TBT RELEASE RATE DATA
Component
Measurement
Reference
Existing Release Rate
20 ug/cm^-day
Anderson (1987)a
Number of Recreational
Vessels in L. St. Clair
95%
100%
14,334
Holecek and Brothers
(1983)
Number-weighted Average
Painted Surface Area
19
Author's calculation
Percentage of TBT Users
on Recreational Vessels
14% X 30%
(U.S. average)
Anderson (1987)a
Number of Ocean-Going Vessels 2.6/day
That Pass Through The System
Papineau (1987)b
Transit Time 5 hours
Number of Ocean-Going Vessels 5%
Docked in The System 100%
X 5
Papineau (1987)b
Armitage (1987)c
Average Painted Surface Area
Percentage ot TBT Users
on Commercial Vessels
9.4 X 103 m2
27%
(U.S. average)
Author's calculation
Anderson (1987)a
aAnderson, E., U.S. Environmental Protection Agency, personal communication.
bpapineau, P., Canadian Coast Guard, personal communication.
GArmitage, K., Harbor Master, Port of Sarnia, Ontario, personal communication.
-------
TABLE II. ESTIMATED LOADINGS OF TBT (kg/day) TO THE LAKE ST. GLAIR SYSTEM
FROM RECREATIONAL AND COMMERCIAL USES OF TBT-BASED ANTI-FOULING
PAINTS
Rate
TBT Uses
20 ug/cm^-day
Recreational Boats
0.44
2.2
Paint Chips in Sediments from
wet scraping (assume 20% of
TBT users scrape down huil
over water)
0.175
@ 0
Precipitation Runoff contribu-
tion from dry dock scraping
(assume 80% of TBT users
scrape down hull in dry dock)
0.35
@ 0
Ocean-going Commercial
Vessels
0.6
0.38
TOTAL
1.0
2.6
sediment slurry in vitro is applicable to river or lake sediments in situ can-
not be ascertained with certitude at present, but it would appear from the
observed ratios of TBT to DBT and DBT to BT that the rates of biotransformation
of TBT to DBT and DBT to BT are slower in natural sediment. Whether photolysis
of TBT yields DBT and thence BT or other photoproducts cannot be determined
from the available data.
Given the tendency of TBT to accumulate in the sediments due to its par-
titioning behavior, and given the relatively higher rate of biodegradation in
the sediments versus the water column, it would appear that the most signifi-
cant route of removal from the aquatic ecosystem is biodegradation in the
sediment. Once degraded from TBT to DBT and thence to BT, the rates of re-
lease of these species from the sediments could be greater than for TBT, both
because their sediment/water partition coefficients would be correspondingly
-------
smaller and their diffusion coefficients correspondingly greater. Thus, it
could well be argued that the sediment is a significant source of the DBT and
BT found in aquatic ecosystems.
The hydraulics and biodegradation rate constants for water and sediment
and the photolysis rate constant for water, as well as desorption kinetics
data, also were obtained from the studies of Maguire and Tkacz (1985). To
quantify the relationship between the loading rate and the biologically
available concentration in the water column, a simplified lake model was
calibrated to the important limnological parameters describing Lake St. Glair
and site-specific adaptations of the fate rate coefficients. The simplified
lake model was developed by HydroQual, Inc., in Mahwah, N.J. (DiToro et al.
19bU). Simple algebraic solutions to the time-dependent and steady state
equations were obtained by the authors at the expense of model accuracy. The
most important simplifying assumptions and approximations used in developing
the model were:
1. Water flow is constant.
2. Water column is homogeneous.
3. Suspended and settled particles are of uniform size, shape and chemi-
cal composition:
o one settling, resuspension and sedimentation velocity
o one particle/water partition coefficient
o one rate constant for each particle associated process
4. Suspended and settled particle concentrations are constant.
5. Homogeneous active sediment layer of constant depth.
6. The bed is stable.
7. Rates of adsorption and desorption are so rapid relative to other
processes that particle/water sorbtion equilibrium exists locally.
8. All toxicant concentration-driven rate processes are first-order
with respect to toxicant concentration.
9. No speciation of dissolved and/or partitioned toxicant occurs.
10. There is no contribution of organisms in direct contact with contami-
nated sediments (e.g., bottom feeding fish and benthic invertebrates)
to the food chain biomagnification of toxicant in top predators.
The model equations, the general algebraic solution for the steady state
condition, the model input data, and the model results are described in the
appendix. The results of this assessment tend to support the conclusion that,
in systems having a low sedimentation rate and short retention times, there
are no significant routes and rates of removal relative to simple hydraulic
dilution. Thus, for the Lake St. Clair example, the assumption of conserva-
tion of mass will introduce an error of only about 10% into the calculation
of the loadconcentration relationship for TBT.
8
-------
MASS BALANCING TBT IN THE ST. GLAIR-DETROIT RIVER SYSTEM
Maguire et al. (1985) studied the distributions of TBT, DBT, and BT in
the St. Glair River and Detroit River water and sediment. The Detroit River
was sampled in June 1983; the St. Glair River was sampled in October, after
the recreational boating season. The fast-flowing St. Glair River is not
nearly as popular as Lake St. Glair for recreational boaters, however, so the
difference may not be too great. Both rivers were sampled during the commer-
cial shipping season, however.
It is highly unlikely that a single sample of water collected and ana-
lyzed from a few stations along these rivers could adequately characterize
the TBT loads entering and leaving Lake St. Glair. On the other hand, with
diffuse and relatively constant sources of TBT, DBT and BT in the system, the
heterogeneity in loadings should not be nearly as great as that associated
with point source industrial cycles or storm water runoff events. Thus, while
the calculation of a mass balance for TBT in the system from so few data could
not be used for purposes of regulating sources, it can be used to gain insight
into relative source strengths in this screening level analysis.
TBT was not detected in subsurface samples in the St. Glair River during
the study by Maguire and co-workers. DBT and BT were detected. If the ratio
of the concentration of TBT to BT at the headwaters of the Detroit River (the
mouth of Lake St. Glair) is the same as at the mouth of the St. Glair River
(the headwaters of Lake St. Glair), then the concentration of TBT can be in-
ferred for the St. Glair River from that ratio and the concentration of BT de-
tected there. A concentration of 8.1 x 10E-12 moles Sn/L is calculated. Using
a flow rate for the system of 4.7 x 10E11 L/day, an upstream TBT loading rate
of about 1.1 kg/day is calculated. The concentration of TBT as organic tin
at the headwaters of the Detroit River was reported to be 6.90 x 10E-11
moles/L, corresponding to a downstream TBT loading of 9.4 kg/day. The dif-
ference between downstream and upstream loadings indicates a source strength
of about 8.3 kg/day in Lake St. Glair in June 1983.
It should be noted here that the estimate of the opening season release
rate from recreational vessels on Lake St. Glair, assuming one-third of the
14,300 vessels put in during any 2-week period for the first 6 weeks of the
season, is about 6.2 kg/day. The estimate of the opening season release rate
for commercial vessels, assuming that one-sixth of the vessels are re-painted
in any given year and that 50% of the newly painted boats travel to the Port
of Sarnia during the initial release rate period, the commercial vessels con-
tribution increases to about 0.6 kg/day. The time of sampling in June of
1983 may have been early enough to quantify the impact of this early release
rate phenomenon on Lake St. Glair and Detroit River water quality.
Whether the initial release rate phenomenon discussed above for commer-
cial vessels could account for the upstream loading rate observed in June of
1983 cannot be determined. An evaluation of the maximum contribution that
could be made to upstream loadings from sediment particle resuspension, pore
water exchange and desorption can account for only a fraction of the 1.1
kg/day estimated loading. The possibility of an upstream point source cannot
be ruled out.
-------
Due to the relatively short retention time in Lake St. Glair, the lake
rapidly approaches steady state conditions in response to this early season
loading. On the other hand, the sediment lags behind the water column,
approaching steady state only towards the end of the recreational boating
season of 180 days duration. The 0.6 kg/day loading rate would be sufficient
to raise the concentration in the water column above the 1 ng/L level of con-
cern and keep it there for the duration of the opening season, perhaps as
long as 30 to 45 days. Under the proposed 4 ug/cm^-day release rate limit,
the initial phase loading contribution from commercial vessels drops to about
0.27 kg/day, just above the loading rate necessary to cause the Detroit River
to exceed a 1 ng/L concentration under once in 10-year, 7-day low flow con-
ditions.
CONCLUSIONS
Based on the above analysis, to assure that susceptible harbors and con-
necting channels in the Great Lakes system not exceed the proposed 20 ng/L
level of concern, the use of TBT anti-foulant paints on the vast majority of
recreational vessels should be discontinued. Although the proposed maximum
release rate of 4 ug/cm2-day at steady state is not likely to pose a threat
to rivers and harbors having short retention times, the maximum initial release
rate of 168 ug/cm^-day from freshly painted vessels that put into the water
only shortly before entering the Great Lakes system could present an unaccept-
able risk to aquatic life during the months of April and May. Follow-up studies
should track the concentration profiles of TBT, DBT, and BT in the water column
and sediment from the opening of the commercial shipping season to its close.
If levels exceeding 1 ng/L are observed for any significant period of time sub-
sequent to virtually eliminating the recreational uses of TBT anti-foulant
paints, additional controls on commercial uses may be necessary.
RECOMMENDATIONS
1. Recreational uses of TBT-based anti-foulant paints in the Great Lakes
Basin should be virtually eliminated.
2. The effect of continued commercial uses of TBT-based anti-foulant paints
under the proposed restrictions should be carefully monitored over the
next 2 years.
3. Over the next 2 years, additional toxicological studies should be under-
taken to evaluate the adequacy of the proposed 20 ng/L TBT criterion to
protect aquatic lite, particularly with respect to the food chain contri-
bution to bioaccumulation.
10
-------
REFERENCES
Connolly, J. 1987. Presentation at Workshop on Modeling PCBs in Lake
Ontario. International Joint Commission, Toronto, Ontario, Canada.
February.
DiToro, D., D. O'Connor, R. Thomann and J. St. John. 1981. Analysis of
the Fate of Chemicals in Receiving Waters - Phase 1. Chemical Manu-
facturers Association, Washington, DC.
Holecek, D.F. and G. Brothers. 1983. Documentation and Analysis of Temporal
and Spatial Changes in Marinas Serving Michigan's Great Lakes. Michigan
Department of Natural Resources, Lansing, MI.
Laughlin, Jr., R.B., H. Guard and W. Coieman. 1986. Tributyltin in Seawater:
Speciation and octanol-water partition coefficient. Environ. Sci.
Technol. 20(2):201-204.
Leo, A., C. Hansch and D. Elkins. 1971. Partition coefficients and their
uses. Chemical Reviews. 71(6):525-616.
Maguire, R.J. 1984. Butyltin compounds and inorganic tin in sediments in
Ontario. Environ. Sci. Technol. 18(4):291-294.
Maguire, R.J. and R. Tkacz. 1985. Degradation of tri-n-butyltin species in
water and sediment from Toronto Harbor. J. Agric. Food. Chem. 33(5):
947-953.
Maguire, R.J., R. Tkacz and D. Sartov. 1985. Butyltin species and inorganic
tin in water and sediment of the Detroit and St. Glair Rivers, J. Great
Lakes Res. 11(3):320-327.
Smith, J., D. Bomberger, Jr., and D. Haynes. 1981. Volatilization rates of
intermediate and low volatility chemicals from water. Chemosphere.
10:281-289.
USEPA. 1987. EPA Proposes Restrictive Use for Tributyltin Antifouling
Pesticides. U.S. Environmental Protection Agency, Washington, DC.
press release, October 1. .
Weast, R. (ed.). 1977. CRC Handbook of Chemistry and Physics, 58th Edition.
CRC, Inc., Cleveland, OH.
11
-------
APPENDIX
TABLE A-l. PROPERTIES OF TBTO AND
Species
Sol.
H20&
log
Kowc
tl/2
biod
tl/2
vol
tl/2
hydrod
tl/2
photo
TBTO
0.75
4.0
1 mg/L
at pH
7.8
TBT
3.2e
20+/-
b wks
(H20)
16 +/
2 wks
(sed)
> 89 d
aMaguire and Tkacz (1985) in Toronto Harbor water and sediment at 20°C.
(Note: All measurements were made at 20°C unless otherwise indicated.)
bpH = 6.0-7.8 in phos. buffer (mg/L).
CpH = 6.0.
dpH = 2.9-10.3.
eLaughlin et al. (1986) report a log Kow value tor TBT of 3.845
12
-------
LAKE ST. CLAIR LIMNOLOGICAL CHARACTERISTICS
flow
5400 M3/s
4.7 x 1011 I/day
average depth of the water column
3.6 M
H2 - average depth of active sediment
layer
A = surface area of the lake
volume of lake
0.05 M
1250 km2
1.25 x 109 M2
3.7 x 109 M3
mla
mlb
mlc
w
rs
hydraulic retention time
the concentration of abiological
particles in the water column
the concentration of biological
particles (algae, etc.) in the
water column
the concentration of colloidal
(non-filterable organic) particles
in the water column
concentration of particles in the
active layer of the sediment
settling velocity of inorganic
particles
resuspension velocity of particles
from the active sediment layer
velocity of burial of inorganic
particles beneath the active sediment
layer
V/Q
8 days
11.2 x 10"6 kg/1
20 x 10"6 kg/1
20 x 10"6 kg/1
0.42 kg/1
1.2 M/day
3.2 x 10"5 M/day
t
@ 0 M/day
13
-------
LAKE ST. CLAIR LIMNOLOGICAL CHARACTERISTICS
TOC
total organic carbon
DOC
"dissolved" organic carbon
TOC of the inorganic particles in the water column and sediment is assumed
to be the average of the TOC values for the St. Clair River and Lake St.
Clair stations, roughly 0.35%. .
TOC of the biological particles in the water column is assumed to be
roughly 35% (Connolly, 1987).
TOC of the colloidal particles in the water column is assumed to be
equal to that for biological particles, or 35%.
9.2 x 10^ I/kg
9.2 x IQT I/kg oc x 0.0035
3.2 x 10^ I/kg
9.2 x 10* I/kg oc x 0.35
3.2 x 10^ I/kg
«lc
= 9.2 x 104 I/kg oc x 0.35
3.2 x 10^ I/kg
lbio
0.693
20 wks x 7 days/wk
4.95 x 10"3 day"1
0.693
16 wks x 1 days/wk
6.2 x 10~3 day
0.693
> 85 days
< 8.15 x 10~3 day
14
-------
LAKE ST. CLAIR LIMNOLOGICAL CHARACTERISTICS
= apparent dissolved fraction in the water column
(non-filterable TBT)
truly dissolved fraction in the water column
= fraction of TBT associated with inorganic particles in
the water column
fraction of TBT associated with biological particles in
the water column
fraction of TBT associated with colloidal particles in
the water column
fPla
mla
0.0015
fPlb
fPlc
fllb x
mlc
0.280
0.280
fdi
* =
fPlb + fPlc
0.4385
fdj* = fd* + fpic
1 - (
15
0.7185
-------
LAKE ST. CLAIR LIMNOLOGICAL CHARACTERISTICS
assume
'2
fp2 s fraction of TBT associated with sediment particles
H2 x
3.2 x 102 I/kg x 0.42 kg/1
1 + 3.2 x 102 I/kg x 0.42 kg/1
0.9926
- fP2
0.0074
16
-------
VOLATILIZATION RATE CALCULATIONS
Ky = 1/L [
RT
H Kwg (Drg)n
i-l
Smith et ai. (1981)
where:
the ratio of the molecular diffusivity of
the chemical to that of air in water
the ratio of the molecular diffusivity of the
chemical to that of water in air
mass-transfer coefficient for water vapor in the
air phase
mass-transfer coefficient for oxygen in the water
phase
m,n
In the range from 1.0- for stagnant conditions and
0.5 for very turbulent conditions
H = Henry's Law Constant, generally calculated as the
ratio of the vapor pressure to the solubility
the universal gas constant (0.082 1-atm/mole-0)
= the absolute temperature on the Kelvin scale
the depth of the water column
17
-------
PARAMETER VALUES FOR VOLATILIZATION RATE COEFFICIENT CALCULATION
K°l
K«
m
n
H
H2U
TCDD
02
0.43 M/day
504 M/day
0.7
1.0
8.42 x 10-1° atm /
0.239 cm2/s (80C)
Smith et al. (1981)
8 x 10~2 mole/M3 = 6.1 x 10'8 atm/mole-M3
2.1 M2/day
2.1 x 10"b cm2/s = 1.8 x 10"4 M2/day
'7 x 10~2 Cm2/s
"6
'1 x 10
H20 2/3
Di x ( MW H?0 )
MW 02
, 1977)
calculated by Smith
et al. (1981)
1.8 x 10"4 M2/day /18)2/3
= 1.2 x lO'4 M2/day
4.4 x 10"5 H2/day
1.2 x 10-4 M2/day
( 0.42 H2/day )
2.1 M2/day
0.37
0.2
1/3.6 M [ "O3 M/day (0.37)'7
0.36 x ID'3 day-1
0.82 L-atm. x 10'3 M3/l x 293 °K
mole-0 ]'
504 M/day (1.8 x 10"7 atn^M3)(0.2)1
moles
18
-------
SIMPLIFIED STEADY STATE TRANSPORT-FATE MODEL OF LAKE ST. CLAIR
CTI(OO) x QL [ i + KT vL/gL 3
B x (r2/r!) (
ks )
ws x fp2 @ 0
H2
kdl x fdl + kPl x fPl
kdl = klbio + klpho + kvol
kdl = 4.95 x 10"3 day"1 + 8.15 x 10~3 day"1 + 0.36 x 10"3 day"1
cdl
1.35 x 10"2 day'1
1.35 x 10~2 day"1 x 0.4385 + 6.2 x 10"3 day"1 x 0.0015
5.9 x 10"3 day'1
(2 '=. kd2 x fd2 +
8.1 x 10~3 day'1
x m x
fp2 x w\2 x
+ 8.15 x 10"3 day"1 x 0.9926
It Is assumed that the biological par-
ticles do not settle in Lake St. Clair
but are carried on through the system to
the Detroit River, because the settling
time of biological particles is longer
than the retention time. Thus, the only
communication between the sediment and
the overlying water column is via inor-
ganic particles. Therefore, fpj = fpia»
ml = mla» and fll =
19
-------
B
0.0015 x 0.42 x 103 kg/M3 x 0.05 M3
0.9926 x 11 x ID'3 kg/M3 x 3.6 M
0.8
If
» tnen
KT - K! + B x (r2/ri) (K2 + ks )
= 5.9 x lO'3 day'1 + 0.8 x 1 x ( 8.1 x 10'3 day"1 + 0 )
1.2 x ID'2 day'1
WT = cTi(oo) x QL C i + KT x VL/QL 3
= 20 ng/1 x 103 1/M3 x 4.7 x 108 M3/day x [ 1 + 1.2 x 10~2 day"1 x 8 days ]
= 9.4 kg/day [ 1 + 0.1 ]
*
= ' 10.3 kg/day
8 3
Under design drought flow conditions, QL = 108,000 cfs = 2.64 x 10 M /day,
and:
WT - 20 ng/1 x 103 1/M3 x 2.6 x 108 M3/day [ 1 + 0.1 ] ' '
* 5.2 kg/day [ 1 + 0.1 ] = 5.7 kg/day
If only 10,000 cfs is given over for dilution of any particular source
category in the Detroit River, as it would be if the source category
were point sources regulated under Michigan's NPDES Permit Program,
then:
WT = 20 ng/1 x 103 1/M3 x 2.45 x 107 M3/day [ 1 + 0.1 ]
= 0.49 kg/day [ 1 + 0.1 ] = 0.54 kg/day
20
-------
SOCIAL AND ECONOMIC ASPECTS OF WATER QUALITY MANAGEMENT
by
A.K. Kuzin, O.I. Kovaleva, and L.S. Garibova^
ABSTRACT
Economic feasibility is an important component of the decision-making
process in the water management area. In this paper, some suggestions are
advanced for determining the economic effectiveness of water management pro-
grams.
WATER QUALITY MANAGEMENT
As self-financing continues to be developed in the USSR, the role of
economic support for good water quality programs is increased. This is due
to the high outlays of funds for construction and exploitation of water man-
agement facilities. Several published scientific articles, recommendations,
and instructions are devoted to methodological issues in this area (Anon.
1981; 1983a,b; 1986a,b; 1987). Economic factors for the support of water
management programs are often carried out to the detriment of social and
economic programs, which creates erroneous assumptions for decision-making.
The role of economic feasibility within the decision-making process in the
water management area is often unclearly defined. In this paper, some sug-
gestions for determining the economic effectiveness of water management pro-
grams are advanced as component parts in the larger issue, taking into account
the social and ecological factors.
The componenets of social well being—living conditions, social environ-
ment, living environment—are derived to a lesser or larger degree from the
quality of water. Living conditions, which are determined by the degree to
which the population is supplied with material goods, are dependent on water
quality on two levels. On the one hand, when there are extremely high out-
lays of funds for the provision of high standards of water quality, there is
a reduction in the production of material goods and, as a rule, there is a
drop in the quality of life. On the other, when outlays to protect water
quality are low, water quality may deteriorate to the point that both capital
iAll-Union Research Institute for Water Protection, Kharkov USSR.
21
-------
and operating costs for water processing will rise sharply to the point that
this will also result in decreased allotments to the production of material
goods. Consequently, from the point of view of providing the best conditions
of life, a compromise on the state of water quality must correspond to the
optimum outlays to protect water quality.
The state of the social environment also is tied to water quality on two
levels. First, when water quality is poor, the health of the population de-
clines, which results in decreased income of the population and increased
costs for public insurance and preventive measures, as well as other negative
effects. Second, when water quality costs are high, closures of enterprises
may result because of their diminished effectiveness as a result of the high-
cost purifying facilities, or employment opportunities may be reduced because
of a reorientation of the enterprises. Thus, as mentioned above, a search
for a compromise on positions is necessary.
Water quality is also an important component of the state of the natural
environment, which serves as a source of aesthetic pleasure and a source of
satisfaction to man both spiritually and culturally. Thus, to a large degree,
the issue of establishing a specific category of natural water is reduced to
the development of an optimal correlation between the components that make up
social well-being—a state characterized by economic, social, and ecological
parameters. The determination of this correlation is difficult to formalize,
however, and the formalization process is imbalanced at different stages of
the decision-making process.
Let us note that the water quality category and the category of the
water body are determined by their purpose, because indicators for water com-
position and properties have been defined accordingly (drinking and household,
general use, fishing industry). Thus placement of water bodies in the appro-
priate category—or in other words, the procedure for establsihing usage
standards for water protection activity—is carried out jointly at the state,
republic, regional, and local levels.
Water quality on the state level is established for major water bodies
that have national significance, such as Lakes Baikal, Sevan, and Ladoga.
The basis for placing water bodies in this category may be their extremely
important economic significance, their exceptional value as natural environ-
ments, or their unique water quality. Also examined are such indicators as a
particular territory's supply of water, its agroindustrial potential, popula-
tion density, condition of the environment in the region, and other factors.
Concerning such priority bodies of water, end use is determined by the
water protection guidelines that take into account the management factors as
well as socio-economic values. When necessary data are available, the eco-
nomic feasibility of reaching the appropriate water quality category can be
determined and taken info account as a component for substantiating this
category. Also determined on the state level is water quality on boundary
sites of principal rivers that border with Union republics, as well as on
state boundaries (on the basis of inter-government agreements).
22
-------
The principal documents that contain decisions or substantiations for
the protection of water bodies of national significance are the decrees of
the USSR Council of Ministers, the General Plan for Utilizing and Protecting
USSR Water Resources, Basin Maps of Large Water Bodies, and intergovernmental
agreements on the protection and utilization of water adjacent to borders.
On a republic level, the end use standards for water quality are deter-
mined at the water body sites that form boundaries between territories, and
that are within the purview of neighboring basin management authorities for
water use and protection thereof, as well as at individual priority water
sites of significance to the republic.
End-use standards for water quality are determined in a similar fashion
on a regional level by a regional basin management that sets standards for
the boundaries between autonomous republics, as well as at water sites that
have the greatest significance for the region.
And finally, there are standards for all bodies of water that have man-
agement or socio-economic importance. These standards, which are set at the
local territorial level, are determined by administrative allotment (oblast,
kray, or autonomous republic).
The substantiations and decisions for setting end use water quality
standards for selected priority water bodies are maintained in the decrees of
the Council of Republic Ministers, the decrees of local management bodies,
and republic-level basin and oblast plans for the use and protection of water
resources. Decisions are made on the basis of management or socio-economic
factors. As the evaluations are substantiated, the accuracy for determined
economic feasibility in achieving the corresponding water quality levels is
improved, which makes it possible to include this component as well.
Calculation of economic feasibility for achieving the corresponding water
quality category and end use standards involves the comparison of results and
outlays involved with improving water quality, i.e.
where: 36+i *s the general economic effect as a result of improved water
quality, from <5 to i-category, millions of rubles per year; ^36+i is
the total effect for water consumers with improved water quality, from <5 to
i-category; millions of rubles per year; li^, ^ei+i is tne adduced costs
for the achievement of water protection measures, for maintaining the base
water quality category, and for achieving the i-category quality, millions of
rubles per year; n is the number of quality categories that are possible to
be attained.
Component indicators included in Formula 1 are selected on the basis of
the following considerations. It is presumed that necessary water protection
facilities have been established previously at this particular site, or a part
23
-------
of it, and that the operation of these facilities is tied to outlays Cfo to
operate them. It is thanks to them that 6 level water quality is maintained
presently.
To maintain water at this level in the future, it is necessary to allo-
cate capital investments K^ for the development of new facilities and to
anticipate increased exploitation costs of AC$ . Thus, maintenance of
the base water quality is tied to the cited 115 , that is,
(2)
where E is the normative efficiency coefficient for capital investments,
taken to be 0.12 with a 7-year amortization period.
If there are plans to achieve water quality above the i-category in the
future, it is necessary to take into account additional capital investments
and the exploitative expenditures of ACg+i, that is,
= "6
3. ~~
AC6+i
(3)
At the same time with the change in water quality category (as compared
to base), it may be possible to improve the economic effect as a result of
improved conditions of water usage. This effect is defined as the sum total
of the types of water utilization, each of which is calculated accordingly as
a sum total of benefits for individual sites or facilities, i.e.
A36+1
a= 1 e= 1
(4)
3. ~~
where: b is the number of usage types; k is the number of a-type usage
facilities and A.3(g)ae is tne economic effect for the e-facility/site
with a-type water usage, when the water quality category is raised (compared
to base), millions of rubles per year.
Although the determination of the economic effect of improved water
quality for various types of usage—household and drinking, fish industry,
general use—has its particular requirements, the common efficiency indicator
is the annual increase in income or the reduction of adduced costs. For
example, with household and drinking usage, changes in the water quality
indicators within the allowable limits results in changes in adduced costs
24
-------
relating to various high-cost water purifying equipment and the expenditures
to service it.
After a decision has been made realtive to the water quality at a given
water site, there arises an economic problem in determining a set of measures
that will assure the type of water quality indicators that will involve mini-
mum outlays. A number of modifications exist to deal with this problem, de-
pending on the details of the original data, familiarity with a particular
water resource, types of models utilized to transform the water quality, and
other conditions. A general view of the economic effect following an exten-
sive examination of the contamination sites and sources, and the selection of
an optimal system of measures is determined by the expression:
n
- min. £
i=l
(5)
where 3j is the economic effect following the development of a network of
inter-connected measures for the purpose of bringing a water body to a norm-
ative state, according to contamination indicator j, millions of rubles per
year; n is the number of sites at which water protection measures are being
implemented; m is the number of contaminating substances that are being
studied; uind.^, H^ are the adduced costs for the i-site for the j-
contaminant, according to individual checks for sources of water contamina-
tion, and the development of a system of inter- connected measures, millions
of rubles per year; P±j is the reduction level of contaminants at the i-site,
j -contaminant in percent, determined by the expression:
UH
Uij
- u99T
u
.H
• 100 at:
min
Dmax
where u±, ^±^CT are the initial and residual concentration of j-type con-
taminating substance (prior to runoff into the water body) for the i-site
[Note: Cyrillic letters "H" and "OCX" above formula indicate intiial and
residual concentrations, respectively]; P±max - the minimal acceptable and
maximum possible degree of contamination reduction of j -contaminant at the
i-site, respectively.
Limitations of contaminating substances may be defined according to:
- extent of contaminants in the runoff into the water body:
.„ <
based on observations of runoff conditions for the contaminants in
the gaging section of a water body:
(6)
25
-------
.
u
rdon.
(7)
based on observations for contaminant runoff in each gaging section
where runoff is contaminated:
.2don.
(8)
where: gj^ is the runoff flow at i-site containing j-contaminant; Wj on -
the allowable amount of contaminants in the runoff into the water body [Note:
Cyrillic "don" is abbreviation for allowable]; u?, ujdon are the calculated
and allowable concentrations of j-contaminant in the gaging section of a water
body, respectively; p are the parameters of the water body that are necessary
to construct a water quality transformation model.
There has been extensive experience in solving the problem that has been
formulated (equation 5). Data and software for various calculation needs
have been developed. In particular, solution of equation 5 defines the maxi-
mum allowable runoff [MAR] for contaminating substances.
An industrial enterprise that has been given a MAR restriction may
achieve the economic effect, according to the water protection regulation, by
reducing the cost of purification (without increasing the volume of contami-
nant in runoff), increasing the profitability of utilizing beneficial sub-
stances, and utilizing the payment factor for discharge of water effluent in
instances where payment has been established and where size depends on puri-
fication effectiveness. The dimensions of such an effect may be determined
according to the expression:
= Sg - min.
j
[iij(pj)+cc6rcymj]
(9)
where:
npi
t*ie economic effect derived by a company as a result of
rationalizing water-protection actions to reduce the j -contaminant ,
S^ is the principal outlays for protecting water against the j- contaminant,
j
is the adduced costs for reducing j-contaminant runoff; Cc6j is
outlays for discharge of j-contaminant; and Cymj
j-contaminant.
is income from utilizing the
A well-thought-out combination of MAR and payments for discharge of
effluents may interest a company financially to improve the results of water
protection activities.
Thus, when water protection programs are being developed, the economic
effect may be achieved through establishment of a water quality level for a
site, or part of it, thorough the development of a system of water protection
measures that assure that the desired wate quality is achieved, and also
through determination of the effectiveness for rationalizing water protection
26
;f
-------
facilities. Thus, if in the first instance, the economical effectiveness
indicators are only one of the factors for the decision-making process, then
in the second and third instances, the economic calculations can serve as the
basis for implementation of water protection measures. Taking into account
the economic factors, along with social and ecologic factors, when establish-
ing water protection parameters makes it possible to heighten the effective-
ness of water quality management.
REFERENCES
Anon. 1981. Methodological Recommendations for Economic and Non-economic
Evaluations of the Effects of Human Activity on the Environment. CMEA
publication, Moscow. 28 p.
Anon. 1983a. Methodology for Calculating Losses for the Government Caused
by Violation of Regulations. USSR IBNTI Minvodkhoz, Moscow. 84 p.
Anon. 1983b. Success of Capital Investments. Collection of Approved Meth-
odologies. EKONOMIKA, Moscow. 128 p.
Anon. 1986a. Standard Methodology for Determining Economic Feasibility for
Implementation of Environmental Protection Measures, and Evaluating the
Economic Losses from Environmental Pollution. EKONOMIKA. 94 p.
Anon. 1986b. Social, Ecological and Environmental Feasibility Study of
Water Protection Measures. Chelyabinsk, USSR Minchermet. 24 p.
Anon. 1987. Decree on Procedures for Determining the Economic Aspects of
Science and Research Work in the Area of Soil Improvement and Water
Management. Moscow. Soyuzgiprovodkhoz Institute, Moscow. 1987, 156 p.
27
-------
INTEGRATED SYSTEM OF STATE CONTROL FOR WATER USE AND
CONSERVATION BASED ON THE EXAMPLE OF THE
AZOV-BLACK SEA DIRECTORATE
by
L.P. Yarmak1
ABSTRACT
The Principles of Water Legislation in the USSR are discussed. These
Principles establish that all water bodies are subject to protection trom any
pollution, contamination, or depletion that could be harmful to the health of
Che population. Specific information is provided concerning water pollution
monitoring and control activities in the Azov-Black Sea Basin.
THE INTEGRATED SYSTEM
The state system of water conservation includes legal, organizational,
technical, and economic activities. Water use regulation in the USSR is
based on the Principles of Water Legislation of the USSR and the Union
Republics (December 1970), which took effect on September 1, 1971, and on
other acts passed in agreement with them.
The state water use and conservation directorate is composed of the
Councils of Ministers of the USSR, of the union republics, the autonomous
republics, the executive committees of the local Councils of Deputies, and
specially empowered organs, such as the Ministry of Melioration and Water
Management of the USSR, similarly named ministries and other Union-Republic
organs of melioration and water management of the union republics, basin
directorates and inspectorates named by them, as well as other local organs
of the system "MINVODKHOZ" of the USSR.
The legal basis of water conservation in the USSR is a statute of the
Constitution to the effect that all natural resources including water, are
the exclusive property of the state and are for its exclusive use. The
statute also forbids any acts that directly or indirectly infringe upon the
state's right to natural resources, including water.
lAzov-Black Sea Basin Directorate, Inspector of Use and Preservation of Water,
Krasnodar USSR.
28
-------
All waters of the USSR constitute the common state water resources. The
common state water resources include rivers, reservoirs, lakes, other surface
bodies of water, as well as the waters of canals and ponds, subsurface waters
and glaciers, internal seas and other internal sea waters of the USSR, and
territorial waters (the territorial sea) of the USSR.
Legal water conservation in the USSR includes preserving the state's ex-
clusive water rights and establishing the legal responsibility of water users
to use water rationally and protect it from pollution, contamination and de-
pletion.
The Principles of Water Legislation in the USSR and the Union Republics
establishes that all water (water bodies) is subject to protection from any
pollution, contamination, and depletion that could be harmful to the health
of the population. This applies as well to any acts that could cause a reduc-
tion of fish stocks, worsen the condition of the water supply, or have other
unfavorable manifestations as a result of changes in the physical, chemical,
and biological properties of the waters, lowering their natural purification
ability, or disturbing the hydrological .and hydrogeological regime of the
waters. Such protection is provided not only for water flowing into the
economic cycle, but also for unused waters.
The Principles of Water Legislation prohibit discharge of industrial,
domestic, and other types of water and refuse into water bodies. Use of
water bodies for discharge of wastewater is only permitted subject to all
requirements and rules reviewed by the legislature of the USSR and its re-
publics. Discharge of wastewater into water bodies can occur only with per-
mission from the water use and conservation regulatory agencies, after their
agreement with the agencies of state sanitary supervision and conservation of
fish resources and with other interested agencies.
Water users, utilizing bodies of water for industrial purposes, are re-
quired to take measures for the reduction and elimination of emissions of
wastewater by means of state-of-the-art production technology.
The Principles forbid initiating operations of facilities affecting local
water quality that are not equipped with systems to prevent pollution and con-
tamination. This must be coordinated with special agencies, the executives of
local Soviets of Deputies of workers, and other interested agencies.
According to water use and conservation regulations, geological survey
organizations are required to inform the agencies that monitor water use and
conservation immediately upon introducing exploratory efforts and to take
measures for the conservation of subsurface waters. Self-pumping wells must
be equipped with regulators and are subject to temporary or permanent shut-
downs.
The Principles provide for the payment of special water use fees accord-
ing to a fee schedule established by the Soviet of Ministers of the USSR.
The Principles also establish criminal or administrative responsibility
(in accordance with legislation of the USSR and union republics) for pollution
29
-------
and contamination ot waters by public utilities and other facilities lacking
treatment equipment to prevent pollution and contamination of the waters or
their harmful effects. Organizations and citizens may be required to provide
compensation tor damages resulting from violation of water legislation.
Guilty officials in such cases must bear material responsibility for their
acts. Insofar as the Principles are the fundamental legislative act, all
other standard setting acts are viable only so far as they do not contradict
the Principles. The Principles of Water Legislation of the USSR and the union
republics are the basic document tor the development of all water codes by
all republics in the union.
The Directorate of Water Quality in the USSR exists in accordance with
specially developed rules—"Rules for preventing surface water pollution by
wastewater" and "Rules for the prevention of coastal water pollution." These
rules establish the main requirements of water quality for two uses—for water
supplies for cities and population centers and for fisheries.
Water quality in bodies of waters is evaluated using physical-chemical,
biological, and microbiological indicators. The analysis of these establishes
compliance or noncompliance of a tested body ot water as to water use require-
ments according to functioning legislative acts. Criteria for evaluating
allowable polluting substance loads for water sources are the maximum allow-
able concentration (MAC) of harmful substances in bodies of water, as well as
their general sanitary characteristics. Requirements applying to water qual-
ity for rivers, lakes, and seas are developed in view of the MAC for water
supply sources located near population centers and also for bodies of water
significant as fisheries. In the Soviet Union, there is a single rule for
wastewater discharge into bodies of water tor rivers, for internal reservoirs,
and for seashores. These rules state the MAC for a large number of harmful
substances (over 500) as well as the calculated hydrological conditions for
water quality evaluation.
To provide for the needs of the population, the economy, and the preser-
vation or establishment of water quality and quantity in the USSR according to
the requirements of water legislation, a general plan of integrated use and
conservation of water resources of the USSR was developed. Also developed
were plans for integrated use and conservation of water resources by basins
(for basins of rivers and other bodies ot water) and by territories (for
republics, areas, regions).
The general plan will serve as the basis for water management for plan-
ning development and distribution of industry of the USSR. The plan is devel-
oped in relation to its interaction with the plan for industry and is the
interdepartmental document on which the distribution of water resources is
based. It also establishes the norms for maximum allowable discharge (MAD)
of wastewater pollutants into bodies of water in a given water management
region.
The integrated plan outlines measures for: (1) the management of ration-
al and economic water use based on improved production technology application
of low water and no water processes; (2) limiting irreversible water losses in
irrigation and water supply systems; (3) maximum possible use of local water
30
-------
resources for water control; (4) interbasin redistribution of river flows;
(5) limiting the discharge volume of untreated wastewater; and (6) the preven-
tion of flooding and submerging of cities, populated centers, agricultural
lands, and other areas.
Integrated plans are being developed for no less than 15 years (by 5-year
plans, with more emphasis on the first 5-year plan) according to river basins
as per the water management regionalization of the USSR. MINVODKHOZ of the
USSR provides for all integrated plans of national significance, determines
the managerial developers, and confirms the technical plan tor development.
The technical plan for development should include: (1) indicators of
basic directions of economic and social development of the USSR for the next
5-year plan and for the long term for the given project as presented to the
directing agencies, (2) the conception of development and distribution of
industry of the USSR, (3) the options for development and distribution of
production, and (4) the management of technical policy. All of these activi-
ties are to be resolved by the Central Committee (CC) of the Communist Party
of the Soviet Union (CPSU) and the Council of Ministers of the USSR.
Under the development of the general plan tor integrated use and con-
servation of water resources of the USSR and other plans of national signifi-
cance, the ministries and agencies of the USSR, based on their branch plans,
present data on production and distribution of water use and water transfer
to MINVODKHOZ according to specific categories of wastewater and discharge
of pollutants into bodies of water and to the type and volume of discharge
into territories of basins.
Water management activity of water use facilities is regulated by
permits of special water use given to basin directorates according to the
requirements of water legislation and confirmed by the plans for integrated
water use and conservation. These permits establish the limits of taking of
water from natural sources and norms for maximum allowable discharge of pol-
lutants with wastwater.
In the Azov-Black Sea Basin Directorate Zone, there are more than 2000
special water use facilities.
Besides normalizing water use conditions and organizing observation and
information collection, the factual conditions of use and quality of natural
surface waters are also the function of the basin directorate. For analysis
ot water management conditions, information about actual water expenditure in
the Kuban River Basin is collected and systematized as are the quality of
water at permanent observation points and the quality and quantity of waste-
water by the water-use facilities.
For this, the basin directorate establishes systematic laboratory moni-
toring of wastewater composition in industrial and agricultural facilities
using territorial hydrochemical laboratories. They also monitor the effect-
iveness of purifying facilities and stations, and the activity of departmental
31
-------
laboratories for research of wastewaters in industrial enterprises. Hydro-
chemical observation is carried out at 600 points set up on the water bodies
near the points of use. All complexes of treatment facilities are regis-
tered and are the responsibility ot the basic hydrochemical laboratories.
According to plan, the departmental laboratories (there are approximately 360
of them in the zone of activity) must submit written analysis of the treat-
ment facilities' work. The volume of analytic work for each departmental
laboratory is set in relation to the quantity of wastewater and the degree
of its pollution.
By analytic groups, the basin directorate annually takes more than 3000
samples of waste and natural waters, these samples contain up to 80 water
quality indicators, on which more than 50,000 determinations are made, includ-
ing determinations for more than 60 specific pollutants (metals, pesticides,
petroleum products, furfural, methanol, formaldehyde, cyanides, phenols, and
others.)
Hydrochemical laboratories run these tests utilizing contemporary meth-
odology, using such instruments as spectrophotocolorometers, spectrophotom-
eters, polarographs, and gas chromatographs.
The collection of initial information on water source quality is carried
out with varying frequency depending upon the purpose and importance of the
source—hydrochemical indicators are determined every 24 hours, every 10 days,
monthly, or by season. Because of this, it is also necessary to determine
the indicators of water composition such as suspended particles, color,
odor, temperature, pH, mineral compositionj oxygen content, biochemical re-
quirements in oxygen, electroconductivity, as well as specific pollutants
characteristic for the given type of production. A full analysis of the back-
ground area water quality tor water sources and rivers that are practically
free of economic activity or production wastewaters is done every 3 months
and more often for polluted areas, depending upon the work regime of the pol-
luting sources and the significance of the water source to the local economy.
A system ot state assessment for water use has been developed in order
to evaluate the status of water resources use. In accordance with this,
every enterprise annually submits a special form with a complete account in-
cluding data on the volume of water taken, how the water was used, and of the
water quality of the resulting wastewater.
These data are systematized and computer-processed according to branches
of industry, to river basins, and to regions. The results of the generaliza-
tion are used to determine the direction of water management development,
comparing it with the basic proposals of the plans for integrated water use
and conservation, the status of the area, the disturbances of special water
use, etc.
Based on the analysis of all information about water management and
water conservation activity of every water use facility so as to determine
its compliance with norms established by the basin directorate, a decision
is taken about the necessity of taking further measures for the improvement
of water use and the quality of discharged wastewater.
32
-------
Based on the orders of the basin directorate, the water using facility
develops a plan of organizational and technical measures including the intro-
duction of more progressive technical and technological advances in the area
of rational water use and conservation in order to meet the established norms.
After agreement with the basin directorate, these plans are introduced into
the state economic development plan and are provided with the necessary
capital.
To provide the economic stimulus for resolving the issues of rational
water use and prevention of water pollution in the USSR, provisions for eco-
nomic sanctions have been made.
Thus, an industrial enterprise pays for the water it uses from the water
management system. In a case where the established use limits are exceeded,
the payment is increased by a factor of five. For discharging pollutants
with wastewater in excess of the allowable norms, the enterprise is charged
with compensation for damages from pollution of water sources. Thus, for a
one-time discharge of 1 ton of pollutants over the allowable norm, the fine
varies trom 8 to 200,000 rubles; for a steady discharge, from 2 to 72,000
rubles. In a case where the discharge of pollutants may lead to serious con-
sequences, or in a case of systematic non-use of water conservation measures,
the enterprise may be shut down.
Provision also is made for criminal and administrative responsibility for
specific officials found to be in violation of water legislation. To provide
timely implementation of water conservation measures in various significant
regions, the USSR is issuing a legislative decree, establishing concrete
assignments to ministries and managements to construct facilities for water
recycling systems.
Thus, in 1976, the CK of phe CPSU and the Union of Ministers of the USSR
issued decree #42 "Regarding the Measures for the Prevention of Pollution of
the Basins of the Black and Azov Seas. In order to effect the decree, 182
purification complexes with a total capacity of 936,000 mz/day were construc-
ted and put into operation in practically every city and town of the Krasnodar
and Stavropol regions. In addition, for rational water resource use, 49 re-
cycling systems with a total capacity of 700,900 mz/day were put into opera-
tion. As a result of this, discharge of polluted wastewater in the Azov-Black
Sea basin has been practically eliminated, and the saving of fresh water
reached nearly 7.8 billion mz. The quality of the shoreline of the Azov and
Black Seas has improved in the Sochi, Adler, Bzugu, Anapa, Gelendzhik,
Mskares, in the rivers of the Adagum in the area of Krymsk, and of the Kuban
River in the area of Armavira and others.
The preservation of small rivers that flow, as a rule, in areas of inten-
sive agricultural productivity, from depletion and pollution presents a par-
ticular problem.
In the USSR, a special decree was issued, "On the Strengthening of the
Preservation of Small Rivers from Pollution, Contamination, and Depletion and
on the Rational Use of their Water Resources." This decree acknowledged the
usefulness of establishing shorelines and water conservation zones along the
33
-------
shores of all small rivers, ravines, and creeks. Conditions were confirmed
tor these shorelines whereby a special regime was established in order to
prevent pollution, contamination, depletion, and buildup of sludge in the
water bodies.
To this end, the shorelines are planted with long-living grasses. Along
the shores, forest belts are planted. Facilities with the potential tor nega-
tive effects on the water bodies are removed. Work is underway to clear the
rivers of sludge layers and brushwood.
The system of state control for the prevention of pollution of the seas
has some specific aspects, the main ones being the regular aerial visual and
instrumental observation of monitored zones. This system allows for provid-
ing effective control of the condition of the sea waters in order to discover
practically all incidents of pollution in a timely manner and stop their
sources, to take energetic measures to liquidate the pollution, and also to
receive and register initial data and evidence for the application of
sanctions.
Systematic aerial monitoring using aerial photography was instituted to
detect and register petroleum pollution. The introduction of severe sanctions
against guilty parties provides for significant reductions in total discharge
of petroleum products into the territorial waters and waters of the economic
zones of the USSR.
Thus, in 1981, 55 cases of pollution, discharging almost 38 tons and
accounting for 60 percent of all discharges of petroleum products were
accounted for. But in 1985, 23 cases, discharging a total of only 5 tons,
and 95 per cent of all petroleum products was accounted for.
In the next two years, the USSR will establish measures for improving
the system of managing the use of natural resources and protecting them from
pollution. Cooperation with other countries is of no small importance in
dealing with these tasks insofar as in recent years, the problem of protect-
ing the environment has taken on an international character.
34
-------
MATHEMATICAL MODELING OF RUNOFF OF SUBSTANCES
FROM AGRICULTURAL WATERSHEDS INTO BODIES OF WATER
by
L. M. Bondarenko, V. Z. Kolpar, and Yu. M. Plis1
ABSTRACT
An approach to the creation of a mathematical model of the runoff of
nutrients and pesticides from an agricultural watershed is presented. The
model takes into account the inhomogeneity of the underlying surface, the
application of chemicals in fields, the spatial inhomogeneity of
precipitation, and other important simulation factors.
MODELING AGRICULTURAL WATERSHEDS
At the present time, the problem of contamination and silting of bodies
of water by surface runoff from agricultural areas has assumed great urgency.
Agriculture is suffering major losses as a result of the washout of fertile
topsoil with the fertilizers and toxic chemicals introduced into it.
Efficient planning of antierosion measures during the establishment of
water conservation complexes of small river basins should be based on esti-
mates of soil erosion and evacuation of nutrient elements and pesticides with
the surface runoff, allowing for the nonuniformity of distribution of the in-
tensity of nonpoint sources of contamination over the agricultural watershed.
One of the effective methods of estimating the evacuation of substances from
a watershed by surface runoff is mathematical modeling, the use of which
makes it possible to consider a large number of natural and anthropogenic
factors and to ensure the formulation of important practical recommendations
in relatively short periods of time.
This paper discusses an approach to the creation of a mathematical model
of formation of the surface runoff of storm waters and evacuation of sub-
stances from an agricultural watershed with a dissected topography. Essen-
tially, this approach consists in the following. The data of a topographic
map of the watershed are used to construct a geometric model on the basis of
which the routes of the runoff to a given contour, for example, the left bank
iAll-Union Research Institute for Water Protection, Kharkov USSR
35
-------
of a river or the edge of lake water, are determined. Then the storm water
paths obtained are used to calculate the runoff hydrograph and the concentra-
tion of suspended matter, which in turn constitute the starting data for cal-
culating the evacuation of chemical substances into a body of water.
Also considered is the possible variety of runoff routes over the planes
comprising the geometrical model of the watershed, and the movement of the
flow over the thalweg system if such exists. This approach also makes it
possible to take into acount a variety of important factors, including the
inhomogeneity of the underlying surface, application of fertilizers and pest-
icides in fields, spatial inhomogeneity of precipitation, and so on.
A general diagram of the software used for the problem is shown in Figure
1. Block 1 is the steering program; blocks 2, 3, and 4 implement the computa-
tional algorithms tor constructing a geometric relief model, finding the run-
off routes of storm waters along the planes of the relief model, and finally,
considering the relationships of water flows moving along different runoff
paths. These blocks have the most complex software. They are standardized,
therefore, and the user need not change them. Blocks 5, 6, 7, and 8, as
follows from their names, implement the computational algorithms for calculat-
ing the runoff hydrograph, concentration of suspended matter in the flow, and
evacuation of nutrient elements and pesticides from the slope. As the mech-
anism of formation of storm water runoff and evacuation of substances from
the slope is studied, these blocks can be improved and introduced into the
software without changing Blocks 2, 3, and 4.
1
(1) Runoff
1
1
(2) Relief
(3) Lateral
inflow
(4) Runoff
path
(5) Flow
(6) Suspended
matter
(7) Nutrient
(8) Pesticide
Figure 1. Module scheme of software.
36
-------
Evacuation ot substances from the watershed is accomplished by the liquid
and solid components of the storm runoff. In turn, other things being equal,
the discharge of liquid runoff predetermines the solid runoff, i.e., the prod-
ucts of soil erosion. Therefore, the most significant part of the problem is
the accuracy and reliability of the modeling o± the liquid component of the
runoff. There are various approaches to the calculation of the runoff volume,
runoff hydrograph, and maximum discharge (Kuchment et al. 1983, Gudzon 1974).
The accuracy of the estimate depends substantially on the consideration of the
variety of factors involved in the formation of the runoff and on the reliabil-
ity of the initial data. In our view, the most acceptable method of solving
practical problems is the use of unidimentional differential equations describ-
ing the motion of a water flow down a slope with allowance made for the rough-
ness of the underlying surface, infiltration ot water into the soil, and chang-
ing rate of precipitation.
The mathematical model of formation of the liquid phase of surface runoff
is based on the unidimensional continuity equation (Moskovkin et al. 1983).
— + — = (R-I)
9 t 3x
(1)
6-104
where: u> is the cross-sectional area of the flow, n»2
q is the water discharge, m^/sec
R, I are, respectively, the rates of rain precipitation and
infiltration, mm/min
B is the average slope width, m
t .is the time, sec
x is the coordinate measured downslope, m
q^ is the lateral inflow per unit length of slope, m^/sec
Integrating Eq. 1 over the length of the slope X from 0 to £ and repre-
senting the water discharge at the exit from the slope by qf with the aid of
the Chezy and Manning formulas,
qf
B
(2)
37
-------
we obtain the following nonlinear ordinary differential equation in
qs
dt
R-l
6-104
B
5 2
~ _-~
3 B 3
(3)
where: q is the water discharge at the entrance to the slope, m
i is the average slope angle
il is the roughness coefficient
£ is the length of the slope
uj and B are, respectively, the cross-sectional area and flow width averaged
over length X,
' dx
To describe the filtration rate I, use was made of the Holton-Overton
equations, given by Beasley and Huggins (1982).
(4)
where: Is is the steady rate of infiltration (filtration coefficients),
mm/min
Im is the maximum rate of inhibition, mm/min
Tp is the total volume of pairs in the confines of the control zone
considered, mm
Pv = Tp - W is the layer of water that can accumulate in the control
zone before its saturation, mm
W is the flowing layer of water accumulated in the control zone during
precipitation, mm
38
-------
P is a coefticient relating the rate of decrease in infiltration rate to
the increase in the moisture content of the soil
In determining Pv, use may be made of the equation
dP
v
dt
= DZ-I
(5)
0, for Pv > G
j == "X
Im(l-Pv/G)3, for P < G
(6)
where: Dz is the rate of drainage of water from the control zone, mm/min
G = Tp - W-£ is the gravitational moisture capacity of the soil within
the confines of the control zone, mm
W-L is the initial moisture capacity in the control zone, mm
The calculation of the evacuation of soil particles from the slope is
based on the equation of sediment balance in the flow in the form (Moskovkin
et al. 1983)
3 3
— (C
-------
(8)
dc _
— + C(
dt
— + K
6-104
0
U)
— C6 + KsC
str
where: Ks, Ctr, GS, q<$ are, respectively, the quantities Ks, Ctr,
and q$ averaged over the length of the slope.
Ks is determined from the formula
u
—
KS = -^-
M'Crv,u
4gv
(9)
where: u is the particle fall velocity, m/sec
g is the gravitational acceleration, m/sec
h is the average depth of the flow, m
M
c - - h-1/6
°Ch n
0.7*Cch + 6, for Cch < 60
48, for cch > 60
is the Chezy coefficient,
Vs- h~2/3 f± i
n
speed of water down the slope, m/sec
The concentration Ctr is calculated from the formula
_ v
Ctr - -=-
h
(10)
40
-------
where E, in accordance with Zubkova (1977) may be calculated from the formula
E = 9.85 CCk* .
Thus to calculate the liquid and solid phases of surface runoff, we
arrive at a closed system of ordinary differential Equations 3, 5, and 8, the
solution of which is carried out numerically by means of the Runge-Kutta
method. The characteristics of the liquid and solid runoffs obtained by cal-
culation may be used to estimate the magnitude of the evacuation of agricul-
tural chemicals (fertilizers and pesticides from farm lands as a result of
washout by surface runoff).
The evacuation of agricultural chemicals by surface runoff is determined
by the results of the interaction of the water flowing down the incline of the
watershed with the top (active) layer of the soil and with the chemicals pre-
sent therein. The depth of the active layer (ha) is determined by the thick-
ness of the saturation zones, where a moisture content equal to the total
moisture capacity is established in the course of precipitation. According
to Rode (1965), the saturation zone is about 10 mm thick.
At the present time, there is no definitive theory providing a reliable
estimate of the evacuation of chemicals from a watershed having a complex
relief structure and a nonuniform soil fertility. At the same time, for
practical purposes, an attempt has been made, on the basis of an analysis of
factors affecting the rate of evacuation of agricultural chemicals from farm
lands, to establish the relationships permitting quantitative estimates of the
magnitude of evacuation of nutrient elements and pesticides with the solid and
liquid phases of runoff. These relationships relate the concentration of the
indicated substances to their content in the arable layer and to the physical
parameters and characteristics of hydrological conditions in the watershed.
The evacuation of dissolved agricultural chemicals can be determined from
the following formulas.
(11)
41
-------
2 Wa
Ph^ - -£)
3 W
0.717C- j)
0.717,
Pha(
2 Wa
3 r
for Wa <
, for | <_wa < -n
for -P £Wa <
for Wa =
(12)
H
W:
D = + 1,274
(13)
where: I>s ±s the evacuation of dissolved agricultural chemicals by the
surface runoff, kg/ha
m is the initial (before the start of rain) content of dissolved
agricultural chemicals in the arable layer of the soil, kg/ha
'ar
is the depth of the arable layer of the soil (in the calculation
of the evacuation of nutrient elements) or the depth of the soil
layer in which most of the pesticide is localized (in the calcu-
lation of the evacuation of pesticides), m
P is the porosity of the arable layer, fraction of unity
h^ is the layer of surface runoff, mm
H is the layer of rain precipitation, mm
W0 is the initial moisture content of the soil in the active layer,
d
fraction of unity
42
-------
Wc is the moisture content of the soil in the control zone before its
humidification, fraction of unity
Q is the volume of water expended in changing the moisture content
in the control zone before the appearance of surface runoff, mm
1 Wc
Q = Phc(- - — ),
3 2P
hc is the depth of the control zone, mm.
The quantity hc is assumed equal to the depth of soaking of the soil be-
fore the appearance of surface runoff, i.e., when the rate of imbibition in
the active layer of the soil (ha = 10 mm) becomes numerically equal to the
filtration coefficient (I = Ic); According to Rode's scheme of the infiltra-
tion process (1965), the depth of soaking varies between 30 cm and 75 cm.
The evacuation of agricultural chemicals with the solid phase of surface
runoff may be estimated from the following equations.
for nutrients
Db =
for pesticides
nb _
mcb*M
s
(14)
ar
M
(15)
where: D, and DP are, respectively, the evacuation of nutrient elements and
pesticides with the solid runoff, kg/ha
m and m P are, respectively, the content in the arable layer of
nutrient elements in exchange-absorbed form and of pesticides in
sorbed form, kg/ha
M is the rate of solid runoff, t/ha
.y is the volume weight of the soil of the arable layer, g/cm^.
Let us give a more detailed description of the software used in solving
the stated problem. We note that the literature gives a description of the
algorithmization of the solution by computer of the problem of runoff forma-
tion and soil erosion in a watershed, for example, the ANSWERS program
43
-------
(BeasLey and HuRgins 1982). When calculations are made according to this pro-
program, the direction ot water flow down the slopes ot the schematized
watershed is specified by the user. This visual specification of the direc-
tion ol the water runoff down a slope oriented arbitrarily in space does not
preclude the occurrence of subjective errors capable of resulting in an appre-
ciable redistribution of runoff over a watershed with a dissected topography.
In this paper, in order to determine the formation and movement of a
storm runoff over a watershed with a dissected topography, a programmed reali-
zation of the solution of this problem is proposed that eliminates the user's
participation in the determination of the runoff path.
Let us consider a watershed G, shown on the map, enclosed by a rectangu-
lar contour I1 (Fig. 2). We fix the surface of the specified watershed in a
rectangular coordinate system OXYZ, superimpose the horizontal plane XOY onto
the map reference plane, and direct the Z axis vertically upward through the
left upper top of contour I'. We draw vertical planes in this coordinate
system parallel to the XOZ and YOZ planes, respectively, with step AX along
thi- X axis and step AY along the Y axis. We thus obtain a grid region G'
2 3
m
-f-
G
-5
\
~r
1 - Boundary of watershed G
2 - River bed
3 - Schematized river bed
4 - Rectangular boundary of watershed F
5 - Fragmentation of schematized watershed into cells G'
i, j - Nodal point numbers
x, y - Direction of coordinate axes
Figure 2. Schematization of a watershed and river bed.
44
-------
consisting of nodes formed by the intersecting straight lines in the XOY
plane. From the data of the topographical map, using the method of least
squares, we calculate the heights of the relief at the nodes of grid region
G', the totality of which determines the geometric relief model.' Naturally,
the smaller kX. and AY are, the more accurately the geometric model will
correspond to the actual relief. The module implementing the solution of
this problem is called the RELIEF module.
The RUNOFF PATH module determines the path of a water flow for a contour
L specified in advance. In general, this contour is a broken line consisting
of sectors connecting the two adjacent nodes of grid region G1. For each of
the sectors of the contour, the RUNOFF PATH module plots tnc runof.[.-forming
chains of triangular planes and thereby defines the individual watersheds,
the runoff from which reaches contour L in a distributed manner, i.e., uni-
formly along the length of a sector. The nodes of the contour are also
analyzed. If the runoff from a thalweg enters the node of a contour, a
runoff-forming chain is formed (which may include a graph of the thalwegs),
and in this case, a concentrated discharge to contour L is considered. In
the computer representation, this will be an ordered recording of the numbers
of the triangles (this recording will be called the runoff-forming chain of
triangles RCT), which consists of four types:
(a) 0A, (b) BC, (c) D0, (d) 01
(16)
where A, B, C, and D represent the coded number of the traingle.
In Expression 16, the meaning of the symbols is as follows.
(a) The runoff from triangle A arrives at the triangle located under-
neath;
(b) The runoff from B arrives at C (among the elements of the RCT, there
must be one more symbol of type EC, B ^ E);
(c) The runoff from D enters either a thalweg or a sector of contour L;
(d) Indication of the end of RCT in sector L.
The RCT having a runoff into the node formed by two adjacent sectors is
simlarly constructed. The construction of the RCT considers all possible
cases of direction of motion of the flow at the models of the watershed:
runoff from one triangle to another
- runoff from one triangle to the two adjacent to it
- the triangle has an inflow from the two adjacent to it
- the two adjacent triangles form a thalweg, and the system of such
pairs of triangles forms the graph of thalwegs
45
-------
- two (three) thalwegs have a runoff into one node of contour L
- two (three) thalwegs originate from the same node
- the runoff from a thalweg (graph of a thalweg) reaches a triangle
The system of runoff-forming chains is the initial information in the
calculation of runoff hydrographs and concentration of substances in the run-
otf with the use of the LATERAL INFLOW module. The RUNOFF PATH and LATERAL
INFLOW modules realized in the YeS computer system in PL/1 language were set
up by the use of the principle of structural programming. The module con-
struction makes it possible, without changing the hierarchic structure of the
programs, to replace one module by another. For example, the runoff of water
from a slope can be described by the NEShA equation, the kinematic water
equation, and hydrodynamic equations. On the basis of each of these equa-
tions, one can construct a FLOW submodule that can easily be connected to the
LATERAL INFLOW module.
The RELIEF, RUNOFF PATH and LATERAL INFLOW modules are combined by the
RUNOFF basic steering program. The RUNOFF program functions as the software
for modeling the formation of storm runoff from agricultural watersheds. In
combination with two files of initial data, the program employed can be used
to consider the following.
- the spatial nonuniformity of the underlying surface
- the space-time variability of sediments above the watershed
- the concentrated (over the ravine grid) and distributed (over the
slopes) runoff of storm waters
- estimate of the effectiveness of antierosion measures (which include
water conservation measures in the watershed).
As an illustration, we will consider the RCT on contour L = T of a water-
shed fragment consisting of four cells (Fig. 3). Figure 3a shows the view of
the watershed from above, and the numbers along contour T indicate the relief
height at the nodes of region G'; number 101, 102, ... denote the consecutive
numbers of the triangles; the dashed line is the dividing line. Figure 3b
shows a model of the watershed fragment in perspective geometry. It is evi-
dent from Figure 3 that the watershed has thalwegs, triangles with two out-
flows: 101, 102, 103, and triangle 202 with two inflows. The RCT elements
are listed in Table 1.
It follows from Table 1 that the runoff from the watershed arrives in a
distributed manner at the sector connecting nodes (1, L) and (2, 1) and in a
concentrated manner down u thalweg into node (2, 1). There Is no ruiiott to
the remaining sectors and nodes of contour L.
Let us consider the specification of information and the course of the
computational process. Two files of initial data are specified. The first
46
-------
A
B
thalweg
Figure 3. Direction of motion of water flow
down the planes and thalwegs of the geome-
tric relief model—A, view of watershed
from above; B, perspective geometry of the
watershed.
file constitutes a table of surface level marks, which is obtained from data
of a topographic map. Steps AX, AY of grid region G' are established which,
as the final result, determine the degree of detail in the description of the
watershed relief. The greater the steps Ax and AY, the more generalized
the geometric model of the watershed will be. On the other hand, when the
values of AX and AY are taken to be small, there is an increase in the
number of cells of the watershed model and hence, in the investment of computer
time in the solution of the problem. Therefore, steps AX and AY should be
selected by considering the accuracy of representation of the actual watershed
relief by its geometrical model as well as the cost of the computing. The
second file also constitutes a matrix whose dimensions are determined by the
number of cells covering watershed G. For each such cell, it is necessary to
know the "passport," by which is meant a set of parameters: the roughness
coefficient, rate of precipitation, filtration coefficient, pesticide content
of the soil at the time of precipitation, etc., necessary for the calculation
using Equations 3, 5, and 8 and Relations 11 through 15.
As a rule, the execution of practical calculations is preceded by cali-
bration calculations, carried out for the purpose of determining the parame-
47
-------
TABI,K 1. Rtmol. l-tormiiitt chains ot triangles on contour L
Mo.
1
2
3
4
RCT
elements
102
1010000
1
102
No.
5
6
7
8
RCT
elements
1030000
1040000
104
204
No.
9
10
11
12
RCT No.
elements
2030202 13
2010202 14
2020000
102
RCT
elements
1010000
1
ters of the models, primarily, those of the hydrologlcal model. In this
case, the file of initial data will contain only the set of parameters for
calculations of the runoff hydrograph. During the calculation, the parameters
ot the model are determined more accurately in order to achieve agreement be-
tween the calculated hydrographs (runoff volumes) and the observed values.
Then, when necessary, calculations of soil erosion are performed until the
observed and calculated masses of the soil introduced into the body of water
are similar. Through an expansion of the second file of data, this procedure
is repeated for the nutrient elements and pesticides being modeled.
The calculation of liquid and solid runoff by means of Equations 3 and 8
is carried out for all the RCT elements. The calculation of the evacuation
ot chemical substances from a watershed is performed as follows. One calcu-
lates the area of the partial watershed adjacent to the sector or node of
contour L, the average content of the substance in the soil of the watershed,
the rate of solid runoff, and the layer of liquid runoff. These quantities
are substituted into Formulas 11 through 15 and thus is obtained the mass of
substances evacuated by storm runoff from partial watersheds.
Such calculations, associated with the use of well known agricultural
and hydraulic antierosion methods, make it possible to ensure the establish-
ment of an efficient, cost-effective and efficient set of measures aimed at
protecting bodies ot water from contamination by surface runoff from farm
lands.
REFERENCES
Beasley, D. and L. Huggins. 1982. ANSWERS (A real nonpoint source watershed
environment response simulation) user's manual. U.S. Environmental Protec-
tion Agency, Region V, Chicago, Illinois. EPA-905/3-82-001.
48
-------
Gudzon, N. 1974. Protection of soils and control of erosion. Kolos, Moscow.
304 p.
Kolpak, V. Z. and V. Ye. Lysenko. 1982. Water quality control of nonpoint
sources of contamination using mathematical modeling of surface runoff
formation. In: Quality control of natural and runoff waters. Sb. Nauchn.
Tr. VNIIVO, Kharkov.
Kuchment, L. S., V. N. Demidov, and Yu. G. Motovilov. 1983. Formation of
river runoff. Nauka, Moscow. 216 p.
Moskovkin. V. M., V. Ye. Lysenko, V. Z. Kolpak and A. Kh. Kurnosenko. 1983.
Modeling of the evacuation of sediments and pesticides by surface runoff
from agricultural watersheds. In: Protection of waters from contamination
by surface runoff. Kharkov, pp. 129-136.
Rode, A. A. 1965. Principles of soil moisture science. Volume 1. Gidromete-
oizdat, Leningrad. 663 p.
Zubkova, K. M. 1977. Study of the hydraulic and morphological characteristics
of surface flows. In: Collected papers on hydrology. Number 120. Gidro-
meteoizdat, Leningrad.
49
-------
AMMONIA DISTRIBUTION IN AND EXCRETION BY FISHES
by
D.J. Randall1 , R.C. Russo2, and R.V. Thurston3
ABSTRACT
Ammonia is widely found in natural water systems in both ionized and
un-ionized forms. This paper reviews research into the conditions affecting
ammonia concentrations in water and the distribution of the compound in fish
tissues. A discussion of ammonia excretion with emphasis on removal across
fish gills is included. Elevated environmental ammonia levels will reduce
excretion and result in ammonia accumulation in the body of the fish with
deleterious consequences.
INTRODUCTION
Ammonia can enter natural water systems from several sources, including
industrial wastes, sewage effluents, alternative fuel conversion processes,
and agricultural discharges. It is a natural biological degradation product
of nitrogenous organic matter.
To understand ammonia distribution in and excretion by fishes, it is
important to understand its chemical equilibrium in water. In aqueous solu-
tions ammonia assumes two chemical forms, illustrated by the following
equation.
NH
nH20 = NH3.nH20
OH + (n-l)H20
(1)
These species are the un-ionized form (NH3), hydrogen-bonded to at least three
(n £ 3) water molecules (Butler 19b4), and the ionized form (NH^+). Total
ammonia is the sum of NHo and NH* , and it is total ammonia that is measured
analytically in aqueous solution.
The relative concentrations of ionized and un-ionized ammonia in a given
solution are principally a function of the ph, temperature, and ionic strength
of that solution. As pH increases, the equilibrium is shifted toward the
^Department of Zoology, University of British Columbia, Vancouver, BC, Canada;
^Environmental Research Laboratory, U.S. Environmental Protection Agency,
Athens, GA, USA;
^Fisheries Bioassay Laboratory, Montana State University, Bozeman, MT, USA.
50
-------
un-ionized species, and the concentration of NH3 increases while that of NH^
decreases. For example, in aqueous solution a pH increase from 7.0 to 8.0
within the temperature range 0 to 30°C results in a nearly tenfold increase
in the concentration of NH3 (Emerson et al. 1975, Thurston et al. 1979).
Temperature increase also favors the NH3 species, but to a lesser extent; a
temperature increase of 5 degrees between 0 and 30°C at pH 7.0 results in an
NH3 concentration increase of 40 to 50% (Emerson et al. 1975). An increase
in ionic strength, at low concentrations, favors the NH^ species. In natural
waters with low to moderate amounts of dissolved solids (200 to 1000 mg/L),
this effect will slightly lower the concentration of NH3, and the magnitude of
this effect will vary with the composition of the water (Thurston et al. 1979).
Ammonia is a toxic end-product of protein metabolism and, therefore, must
be excreted or converted to less toxic compounds, such as urea or glutamine.
Ammonia is, however, both a substrate and a product of protein metabolism and
in some tissues it may be utilized rather than produced. Elevated concentra-
tions of environmental ammonia will reduce excretion and result in ammonia
accumulation within the body of the fish.
AMMONIA DISTRIBUTION
The pK of the ammonia/ ammonium reaction is around 9.5, so at the pH of
fish tissues nearly all of the ammonia will be as NH^ . Cameron and Heisler
(1983) found that ammonia was slightly more soluble in fish plasma than in
water, and they also constructed a nomogram to describe the effects of ionic
strength and temperature on the pK of the NH^/NH^ reaction (see also Kormanik
and Cameron 1981, Boutilier et al. 1984).
Ammonia gas (N^) diffuses at about the same rate as, but is much more
soluble than, C02, so it will rapidly equilibrate between different tissue
compartments and be excreted across the gills. As a result, the body concen-
trations of NHo are low. The concentrations of NH^ , however, can be sev-
eral orders of magnitude higher than those of NH3, bringing total ammonia
levels into the mMol range. Concentrations of NH^ often reflect the pH of
the compartment (Randall and Wright 1987); tissues with a lower pH having
higher total ammonia concentrations. In many instances, however, NH^ is
distributed according to membrane potential, indicating a considerable mem-
brane permeability to NH^+. For example, Wright et al. (1988b) have shown
that the distribution of NH^+ in skeletal muscle, heart muscle „ and brain of
the lemon sole (Parophrys vetulus) is related to membrane potential rather
than pH, indicating a relatively high NH^ permeability across the barriers
between these compartments and blood. Thus, NH^ concentrations are greater
than expected from that due to pH. The consequence will be a continual pro-
duction of H within the cell with the cycling of NH^ into the cells and the
diffusion of NH3 out of the cell.
Wright et al. (1988c) also found that ammonia was in equilibrium across
red blood cell membranes and that NH^+ was distributed between plasma and red
blood cells according to membrane potential. In this case, however, the
results need not indicate a high NH^ permeability, as hydrogen ions also are
distributed according to membrane potential.
51
-------
AMMONIA EXCRETION
Most of the ammonia produced by the fish is excreted across the gills,
and the amount is variable, depending on the state of the animal, envi-
ronmental conditions, and the species. Ammonia excretion by the dogfish
(Scyliorhinus stellaris) in seawater is unaffected by temperature change,
exercise, hyperoxia, hypercapnia, or the infusion of HC1, NaHC03, or anything
that induces acid-base stress (see Heisler, 1984, for review). Ammonia excre-
tion tripled in sockeye salmon (Oncorhynchus nerka) following daily feeding
(Brett and Zala 1975) but remained low and unchanging during 22 days of
starvation.
In freshwater fishes, ammonia excretion increases in response to exer-
cise (Sukumaran and Kutty 1977, Holeton et al. 1983), long-term acid exposure
(McDonald and Wood 1981, Ultsch et al. 1981), hypercapnia (Claiborne and
Heisler 1984), and Nt^Cl infusion (Hillaby and Randall 1979). In contrast,
short-term exposure to acid or alkaline water caused a decrease in ammonia
excretion in trout (Wright and Wood 1985). It is not known whether these
changes in excretion reflect changes in the rate of ammonia production or in
the total ammonia content (NHo + NH^+) ot the body.
The total ammonia content of fish is likely to be about the equivalent
ot the ammonia excreted in 2 hours, most of the total ammonia being stored in
muscle. Blood levels are around 0.2 to 0.3 mMol, but muscle has concentra-
tions up to 1 mMol. Thus a 1-kg fish contains about 0.5 to 0.7 mMol of total
ammonia and has an excretion rate of about 0.3 mMol per hour.
There is an increase in blood ammonia during starvation (Hiilaby and
Randall 1979, Morii 1979), which is perhaps surprising because at the same
time ammonia excretion declines (Brett and Zala 1975). Carbon dioxide ex-
cretion is reduced during starvation and this may account for this reduction
in ammonia excretion and concommittent elevation in blood ammonia concentra-
tion.
Blood ammonia concentrations also increase with increases in temperature
(Fauconneau and Luquet 1979) and exposure to air (Gordon 1970), or to in-
creased ammonia concentrations in water (Fromm and Gillette 1968, Guerin-Ancey
197b). This is associated with a rise in urea production in many fishes. Un-
like the above studies, Buckley et al. (1979) found no change in total ammonia
in the blood when coho salmon (Oncorhynchus kisutch) were exposed to elevated
ammonia concentrations in the environment. They did observe a significant
rise, however, in plasma sodium indicating some coupling between sodium uptake
and ammonia excretion.
The excretion of ammonia is largely a function of the NH3 gradient across
the gills (Hiilaby and Randall 1979, Kormanik and Cameron 1981, Cameron and
Heisler 1983, Wright and Wood 1985), as is ammonia entry into the fish (Wuhr-
mann et al. 1947, Wuhrmann and Woker 1948, Fromm and Gillette 1968). The ex-
cretion ot NU^+ in freshwater fishes is strongly coupled to the movement of
other ions. Membranes, including the gills, are not very permeable to cations,
and NH.+ is probably transferred across the gill epithelium via carrier-media-
ted processes. Potassium can be displaced by Nll^ in many membrane processes,
52
-------
for example in squid giant axon (Binstock and Lecar 1969), and this may be
the reason that elevated ammonia concentrations cause convulsions in so many
vertebrate's.
In fish gills, it is possible that NH^+ can substitute for potassium in
oubain sensitive sodium/potassium exchange, and also substitute for protons
in amiloride sensitive Na+/H+ exchange—the former moving NH^+ from blood into
the gill epithelium, and the latter exchanging NH^+ for sodium on the outer
surface of the gill epithelium (Maetz and Garcia-Romeu 1964, Evans 1977,
Girard and Payan 1980, Wright and Wood 1985). Either acid conditions or
amiloride in the water inhibit sodium influx across the gills, and both these
conditions result in a reduction of ammonia excretion (Wright and Wood 1985,
Randall and Wright, 1986). In addition, ammonia infusion will stimulate
sodium influx, even in salt water fish (Evans 1977).
Cameron and Heisler (1983), however, could account for ammonia excretion
in trout, under most conditions, by the diffusion of NH3, but in the presence
of high external ammonia, Na /NH^+ exchange may counterbalance the diffusive
uptake of ammonia. Indeed, this would explain the unchanged blood ammonia
levels but increased sodium levels in coho salmon exposed to elevated ammonia
concentrations (Buckley et al. 1979).
A more detailed understanding of ammonia excretion requires a more de-
tailed analysis of pH gradients across the gills. The amount of C(>2 excreted
across the gills usually exceeds that of ammonia. Water flow over the gills
is laminar and there are boundary layers next to the epithelial surface.
Mucus and dead cells contribute carbonic anhydrase to this boundary layer.
Fish excrete molecular C02 rather than HCOg" across their gills. The
CO2 entering the boundary layer will be rapidly hydrated to HCO-j", and this
will acidify the water boundary layer next to the gill surface (Wright, et al.
1986). Ammonia entering the boundary layer will be trapped as NHA+, maintain-
ing the ammonia gradient across the gills. A reduction in C(>2 excretion will
reduce the acidity of, and therefore raise ammonia levels in, the boundary
layer and hence the blood. An elevation of blood ammonia is seen in starving
fish and this is associated with a reduction in C02 production.
More recently Wright et al. (1988a) have shown that manipulations that
reduce the acidity of the boundary layer have a marked effect on ammonia
transfer. Thus, C(>2 and ammonia excretion are coupled in the fish gill as in
the kidney, although the nature of the coupling is somewhat different.
Carbon dioxide in the water affects ammonia toxicity; if C02 levels are
raised, total ammonia toxicity is decreased (Alabaster and Herbert 1954).
This is because an increase in CC>2 causes a fall in pH and decreases the
proportion of NEhj in solution. The un-ionized form has a greater toxic effect
because ammonia must enter the fish to exert its toxic action, and lipid mem-
branes are much more permeable to ammonia as NH3 (Wuhrmann et al. 1947, Wuhr-
mann and Woker 1948, Thurston et al. 1981).
53
-------
Lloyd and Herbert (1960), however, found that although total ammonia tox-
icity was reduced at high C02 levels, the inverse was true when considering
NH3 alone. More N% is required in low CC^-high pH water to exert the same
toxic effect as is seen in fish in high C02-low pH water. The explanation
presented by Lloyd and Herbert (1960) for the decreased toxicity of NH3 in
low C02 water was that CC>2 excretion across the gills would reduce pH and
therefore the concentration of NH3 in water flowing over the gills. This is
consistent with the conclusions of Wright et al. (1988a), with the addition
that these effects are occurring in the boundary layer of water next to the
gills rather than in the bulk flow.
REFERENCES
Alabaster, J.S. and D.W.M. Herbert. 1954. Influence of carbon dioxide on
the toxicity of ammonia. Nature: 174:404.
Binstock, L. and H. Lecar. 1969. Ammonium ion currents in the squid giant
axon. J. Gen. Physiol. 53:342-361.
Boutilier, R.G., T.A. Heming, and G.K. Iwama. 1984. Appendix: Physico-
chemical parameters for use in fish respiration physiology. Pages 403-
430 In: Fish Physiology, Volume X(A), W.S. Hoar and D.J. Randall, eds. ,
Academic Press Inc., New v.ork, N.Y. , USA.
Brett, J.R. and C.A. Zala. 1975. Daily pattern of nitrogen excretion and
oxygen consumption of sockeye salmon (Oncorhynchus nerka) under con-
trolled conditions. J. Fish. Res. Board Can. 32:2479-2486.
Buckley, J.A., C.M. Whitmore, and B.D. Liming. 1979. Effects of prolonged
exposure to ammonia on the blood and liver glycogen of coho salmon
(Oncorhynchus kisutch). Comp. Biochem. Physiol. 63C:297-303.
Butler, J.N. 1964. Ionic Equilibrium. Addison-Wesley Publishing Co., Inc.
Reading, Mass., USA.
Cameron, J.N. and N. Heisler. 1983. Studies of ammonia in the rainbow trout:
physico-chemical parameters, acid-base behaviour, and respiratory clear-
ance. J. Exp. Biol. 105:107-125.
Claiborne, J.B. and N. Heisler. 1984. Acid-base regulation and ion trans-
fers in the carp (Cyprinus carpio) during and after exposure to environ-
mental hypercapnia. J. Exp. Biol. 108:25-43.
Emerson, K., R.C. Russo, R.E. Lund, and R.V. Thurston. 1975. Aqueous ammonia
equilibrium calculations: Effect of pH and temperature. J. Fish. Res.
Board Can. 32:2379-2383.
Evans, D.H. 1977. Further evidence for Na/NH4 exchange in marine teleost
fish. J. Exp. Biol. 70:213-220.
54
-------
Fauconneau, B. and P. Luquet. 1979. Effect of temperature rise on blood
aminoacids and ammonia in rainbow trout (Salmo galrdneri R«) after feed-
ing. Ann. Bioi. anim. Bioch. Biophys. 19(4A):1063-1079. [Translated
from French.]
Fromm, P.O. and J.R. Gillette. 19b8. Effect of ambient ammonia on blood
ammonia and nitrogen excretion of rainbow trout (Salmo gairdneri). Comp.
Biochem. Physiol. 26:887-896.
Girard, J.P. and P. Payan. 1980. Ion exchanges through respiratory and
chloride cells in freshwater- and seawater-adapted teleosteans. Am. J.
Physiol. 7:R260-R268.
Gordon, M.S. 1970. Patterns of nitrogen excretion in amphibious fishes.
Pages 238-242 In Urea and the Kidney, Proceedings of an International
Colloquy, Sarasota, FL, USA, September 9-12, 1968. B. Schmidt-Nielsen,
ed., Excerpta Medica Foundation, Amsterdam.
Guerin-Ancey, 0. 1976. Experimental study of the nitrogen excretion of bass
(Dicentrarchus labrax) during growth. III. Effects of water volume and
initial ammonia concentration on the excretion of ammonia and urea.
Aquaculture 9:253-258. [Translated from French]
Heisler, N. 1984. Acid-base regulation in fishes. Pages 315-401 In: Fish
Physiology. Volume X(A), W.S. Hoar and D.J. Randall, eds. Academic
Press Inc., New York, USA.
Hillaby, B.A. and D.J. Randall. 1979. Acute ammonia toxicity and ammonia
excretion in rainbow trout (Salmo gairdneri). J. Fish. Res. Board Can.
36:621-629.
Holeton, G.F., P. Neumann, and N. Heisler. 1983. Branchial ion exchange and
acid-base regulation after strenuous exercise in rainbow trout (Salmo
gairdneri). Resp. Physiol. 51:303-318.
Kormanik, G.A. and J.N. Cameron. 1981. Ammonia excretion in animals that
breathe water: a review. Mar. Biol. Letters 2:11-23.
Lloyd, R. and D.W.M. Herbert. 1960. The influence of carbon dioxide on the
toxicity of un-ionized ammonia to rainbow trout. Annals Appl. Biol.
48:399-404.
Maetz, J. and F. Garcia-Romeu. 1964. The mechanism of sodium and chloride
uptake by the gills of freshwater fish, Carassius auratus. II.
Evidence for NH4+/Na+ and hCOo~/Cl~ exchanges. J. Gen. Physiol. 47:
1209-1227.
McDonald, D.G. and C.M. Wood. 1981. Branchial and renal acid and ion
fluxes in the rainbow trout, Salmo gairdneri, at low environmental pH.
J. Exp. Biol. 93: 101-118.
55
-------
Morii, H. 1979. Changes with time of ammonia and urea concentrations in the
blood and tissue of mudskipper fish, Periophthalmus cantonensis and
Boleophthalmus pectinirostris kept in water and on land. Comp. Biochem.
Physiol. 64A:235-243.
Randall, D.J. and P.A. Wright. 1986. Ammonia production and excretion by
fish. Pages 19-30 In: Proceedings of the US-USSR Symposium on Problems
of Aquatic Toxicology, Biotesting, and Water Quality Management. July 30-
August 1, 1984, Borok, Jaraslavl, USSR. R.C. Ryans, ed. U.S. Environ-
mental Protection Agency, Athens, GA, USA. EPA/600/9-86/024.
Randall, D.J. and P.A. Wright. 1987. Ammonia distribution and excretion in
tish. Fish Physiol. Biochem. 3:107-120.
Sukumaran, N. and M.N. Kutty. 1977. Oxygen consumption and ammonia excretion
in the catfish Mystus armatus, with special reference to swimming speed
and ambient oxygen. Proc. Indian Acad. Sci. 868:195-206.
Thurston, R.V., R.C. Russo, and K. Emerson. 1979. Aqueous ammonia equilib-
rium—tabulation of percent un-ionized ammonia. U.S. Environmental Pro-
tection Agency, Duluth, MM, USA. EPA/600/3-79/091.
Thurston, R.V., R.C. Russo, and G.A. Vinogradov. 1981. Ammonia toxicity to
fishes. Effect of pH on the toxicity of the un-ionized ammonia species.
Environ. Sci. Technol. 15:837-840.
Ultsch, G.R., M.E. Ott, and N. Heisler. 1981. Acid-base and electrolyte
status in carp (Cyprinus carpio) exposed to low environmental pH. J.
Exp. Biol. 93:65-80.
Wright, P.A., T.A. Heming, and D.J. Randall. 1986.
water flowing over the gills of rainbow trout.
Downstream pH changes in
J. Exp. Biol. 126:499-512.
Wright, P.A. and C.M. Wood. 1985. An analysis of branchial ammonia excre-
tion in the freshwater rainbow trout: effects of environmental pH change
and sodium uptake blockade. J. Exp. Biol. 114:329-353.
Wright, P.A., D.J. Randall, and S.F. Perry. 1988a. Fish gill water boundary
layer: a site of linkage between carbon dioxide and ammonia excretion.
J. Comp. Physiol. In press.
Wright, P.A., D.J. Randall, and C.M. Wood. 1988b. The distribution of ammonia
and H+ between tissue compartments in the lemon sole (Parophrys vetulus):
the effects of hypercapnia and exercise. J. Exp. Biol. In press.
Wright, P.A., C.M. Wood, and D.J. Randall. 1988c. An in vitro and in vivo
study of the distribution of ammonia between plasma and red cells of
rainbow trout (Salmo gairdneri). J. Exp. Biol. 134:423-428.
56
-------
Wuhrmann, K. and H. Woker. 1948. Contributions to the toxicology of fishes.
11. Experimental investigations on ammonia and hydrocyanic acid poison-
ing. Schweiz. Z. Hydrol. 11:210-244. [Translated from German.]
Wuhrmann, K., F. Zehender, and H. Woker. 1947. Biological significance for
fisheries of ammonium ion and ammonia content of flowing bodies of
water. Vierteljahrsschrift der Naturf. Gesellschaft in Zurich. 92:198-
204. [Translated from German.]
57
-------
EFFECT OF AMMONIUM IONS ON MINERAL EXCHANGE IN
FRESHWATER FISH AND CRUSTACEANS
by
G.A. Vinogradov*
ABSTRACT
The effects of ammonium salts (NH^Cl) and (NI^^SC^ at different pH
values of the medium on the regulation of sodium exchange in the crucian carp
are examined. Results indicate disturbances in ion exchange at sublethal
concentrations of total and ionized ammonium.
INTRODUCTION
Ammonia, the end product of protein metabolism, has a toxic effect dur-
ing its accumulation in the fish organism. The percentage dissociation of
free ammonia from ammonium salts in aqueous solutions is substantially depend-
ent on the pH and temperature of the water. When the solution pH increases,
the equilibrium shifts toward the formation of free ammonia (NH3>. In the
range of pH acceptable to aquatic animals, a one-unit increase in pH raises
the NHg concentration approximately ten-fold. Raising the temperature also
promotes the formation of NH3, but to a lesser degree. Increasing the ionic
strength at low concentration promotes the formation of the NH^ species
(Thurston et al. 1979).
The concentration of ammonium ions (NH^ ) in fish blood normally amounts
to 0.2 to 0.3 mmole/L. Most of the ammonia produced by a fish is eliminated
through the gills (Para and Prekup 1960). At physiological pH values, prac-
tically all of the ammonia in the organism is in the ionized state. The gills
of fish are not very permeable to NH^+. At the present time, there is every
reason to assume that ammonia is eliminated through the gills, not only be-
cause of passive diffusion of gaseous ammonia (Ntkj), but also as a result of
NH^+/Na+ exchange. Most of the data indicating NH^ /Na+ exchange in the gills
of freshwater fish were obtained on the crucian carp and fish of the salmon
family (Maetz 1973, Wright and Wood 1984, Randall and Wright 1986).
1Institute of Biology of Inland Waters, USSR Academy of Sciences, Borok
USSR.
58
-------
RESULTS
Our experiments dealt with the effect of ammonium salts (N^Cl) and
at different pH values of the medium on the regulation of sodium
exchange in the crucian carp (Carassius carassius). The introduction of
ammonium salts into aquariums containing fish acclimated for 2 and 7 days at
pH 5.5 and 4.5 to a concentration of 5 and 10 mmole/L did not cause any dis-
ruption in Na+ uptake (Thurston et al. 1979).
The concentration of Na+ and Cl~ ions in the blood of the crucian carp
and pond carp placed in water with different concentrations of (NH4)2S04 up
to the maximum tolerable value did not differ reliably from the control.
Sodium exchange in the crucian carp in alkaline medium (pH 7.8) changes
insignificantly at a very high concentration of NH^* (Figure 1). Immediately
after the fish are placed in water containing 2.5 mmole/L NH4C1, the net up-
take of sodium becomes negative as a result of an increase in the passive
yield of this ion, which in our view is due to the ammonia stress factor.
After 1 day and during the subsequent acclimation, the uptake of pure sodium
became positive (Figure 1). Placing the fish after 4 days of acclimation to
2.5 mmole/L of Nl^Cl in water containing no ammonium ions slightly increased
the sodium loss in distilled water, but the overall Na+ balance remained
positive. During the subsequent deacclimation, small variations in sodium
exchange caused by a change in Na+ loss were observed. Ammonium ions in
high concentrations had no reliable effect on sodium exchange in the gills
or in other species of freshwater fish—the perch (Perca fluviatilis) or the
roach (Rutilus rutilus).
Acclimation
Deacclimation
Total calcium uptokex
5'f 7 11 15 19
Sodium losses in/
400/i.mole/L of CaCI
Days
Figure 1. . Effect of ammonium (2.5 mole/L NtfyCl) on the
uptake of calcium and sodium and loss of sodium in the
crucian carp (pH 6.8, temperature 15 °C).
59
-------
A possible cause of the discrepancy between our results and the data of
other investigators may be the fact that injections into the organism of
ammonium solutions, which were used to provide Na+/NH^+ exchange, caused the
development of appreciable acidosis (Cameron and Heisler 1983), which stimu-
lated Na"l"/H+ exchange. In earlier studies, this fact was not considered in
the interpretation of the results, and therefore, the increase in sodium
transport after the injections was regarded as stimulation of Na/NH^ exchange.
Nevertheless, it should be noted that in fish of the salmon family, there is
a relationship between the excretion of NH^+ and sodium uptake (Wright and
Wood 1984, Randall and Wright 1986). In a weakly acidic or alkaline medium,
the flow ratio of NH^+ and Na+ is 1:1, and a substantial portion of the total
ammonia is eliminated from the organism in the form of NH^ (Randall and
Wright 1986).
According to our data, the presence of NlfyCl in the water disturbs the
K+ balance. An excess of potassium loss over its uptake is observed during
the entire acclimation period.
izes the potassium exchange.
The return of the fish to clean water normal-
NH^ ions depress the transport of calcium in the crucian carp starting
at a concentration of 0.25 to 0.5 eq/L (Figure 2). The inhibiting effect of
NH^* also is preserved during the acclimation of the fish to water with an
NH,+ concentration of 1 to 5 eq/L. Transferring the fish into clean water
after their exposure to ammonia stimulates the uptake of calcium by 150 to
1 - Atlantic salmon (fingerlings)
(temp. 10°C, pH 6.5)
2 - Brachydanio rerio
(temp. 22°C, pH 7.8)
3 - Crucian carp
(temp. 21°C, pH 6.5)
0.5 1.0 20
Concentration, mmole/L
Figure 2. Effect of Nlfy4" on Ca2+ uptake in
fish in 0.5 mmole/L CaCl.
60
-------
200% in comparison with the control. Lowering the pH of water to 4.0 to 4.5
does not appreciably affect the calcium exchange. The above results speak in
favor of Ca /NH^"1" exchange. Apparently, this exchange is also characteris-
tic of other bony fish living in fresh water.
The inhibition of calcium uptake by ammonium ions takes place in the
immature Atlantic salmon - parr and "aquarium fish" (Brachydanio rerio). The
Ca /NH^ exchange also is supported by the results of a study of the excre-
tion of total ammonia in the perch. In distilled water, the yield of NH3 +
NH4 is 0.057+0.18 ueq/gh, in CaCl2 solution 20 ueq/L 0.125+0.046, in 50
meq/L 0.163+0.031, and in 100 - 0.212+0.044. The calcium uptake in these
solutions is 0.021+0.010, 0.06+0.02, and 0.130+0.40 ueq/gh, respectively. As
the calcium content of the water increases further, the yield of total ammonia
either remains unchanged or decreases slightly. The results ot this study
show that the calcium ion may be exchanged for two ammonium ions. A breakdown
of this ratio during an intensive uptake of calcium is probably due to an
insufficiency of the ion exchange stock of NH^+.
The acclimation of crucian carp to a higher content of ammonium ions re-
vealed a number of significant aspects of calcium exchange. It was found that
when the NH^ content of the water is 2.5 and 5.0 meq/L, the calcium transport
practically stops after 30 minutes of action. It then increases slightly
during the next 16 to 24 hours of acclimation. Subsequently, the level of
calcium uptake remains almost unchanged, and after 4 to 5 days, the acclima-
tion amounts to 10 to 30% of the original value (Figure 1). At a lower con-
centration (1 meq/L of NH^+), the degree of the original inhibition of calcium
inflow is about 40%. During the subsequent acclimation, the calcium uptake
from water is not completely restored (Figure 3).
When the crucian carp are returned to clean water after acclimation to
an increased ammonium content, a rapid increase in Ca^"1" uptake is observed.
CL CD
13 "5
0.28
0.24
0.20
o £ 0.16
o ^ 0.12
«- £ 0.08
° *O 0.04
(5 0
2 3 4 5 6 7 8 9 10 11 12 13 14 15
Days
Figure 3. Effect of ammonium (1 mmole/L NH^Cl) on the
total uptake of calcium in the crucian carp (tempera-
ture 18 °C, pH 6.8, 0-13 days - medium with NH4+, 13-
15 days - pure water).
61
-------
During the entire first hour of acclimation, the Ca^+ uptake from water in-
creases substantially and reaches a value that is 1.5 times the level of
normal calcium transport. After 3 to 4 hours of deacclimation, the calcium
uptake by the organism decreases again. During the next 16 to 20 hours, the
calcium exchange becomes normalized.
A study of the influence of ammonium salts on the dynamics of uptake and
loss of sodium in the narrow-clawed crayfish (Astacus leptodactilus) showed
that, at ammonium concentrations in excess of 1 meq/L of NH^ , the equilibrium
between the loss and gain of sodium is disturbed. In the first 2 to 3 hours,
a depression of sodium uptake and an increase of its discharge into the sur-
rounding medium are observed. The degree of deviation from the balanced state
depends on the ammonium concentration (Figures 4 and 5). Subsequently, there
is a gradual restoration of the rate of sodium transport and rate of its
escape into the surrounding medium.
After 24 to 36 hours, both processes become stabilized, and do not change
significantly for the next 3 days of acclimation. After the crayfish are put
back in clean water (deacclimation), a marked increase in sodium uptake is
noted, the magnitude of which substantially exceeds the initial level. Then
the sodium transport gradually decreases, and after 48 hours the deacclimation
reaches normal values. During the deacclimation, a reduction of total sodium
loss is noted.
In the course of longer periods of acclimation to increased ammonium con-
centration, it was noted that an equilibrium in sodium exchange was estab-
0.30
o
CO
co^.
CO ^
3 ~o 0.15
c
o
o
-*-»
ex
Acclimation
Deacclimation
I 1 - Sodium loss
2 - Total sodium uptake
3 - Sodium loss
4 - Total sodium uptake
I
24 48 72 96 120 144
Hours
Figure 4. Na+ exchange in the narrow-clawed cray-
fish (Astacus leptodactylus.) during acclimation
to 2.5 mmole/L NlfyCl (temperature 12 °C, pH 7.2-
7.4—lines 1 and 2) and 5 mmole/L (Nlfy^SCty (temp-
erature 18-21 °C, pH 7.2-7.4—lines 3 and 4).
62
-------
lished at a level that was lower than the initial one. After 14 days, the
inflow of sodium was equal to its loss. After the crayfish were transferred
to clean water, as in the case ot a shorter acclimation period, a manyfold
increase in sodium uptake and decrease in its loss took place.
Placing the crayfish in an isotonic solution of NaCl (190 meq/L) in-
creases the sodium transport three-fold during the first few hours of the
experiment. After a day of exposure to the salt solution, the sodium uptake
practically ceases. The Na+-dependent ammonium yield, calculated from the
difference between the excretion in animals with maximum sodium transport
and nontransporting animals, amounts to about 60% of the total ammonium
excretion.
A study of the effect of dinitrophenol, which is a respiratory phosphory-
lation inhibitor, on sodium uptake and ammonium excretion made it possible
to establish the interdependence of these processes. Inhibition of active
transport of ions through the gills substantially reduces the level of
cation-dependent ammonium excretion. Good agreement is observed between the
degree of inhibition of ion transport and ammonium excretion. A three-fold
depression of sodium transport and, apparently, other cations leads to an
analogous reduction of cation—dependent excretion.
In the amphipod (Gammaracanthus lacustris), the effectiveness of the
..action of ammonium ions on the uptake and loss of sodium is somewhat lower
than in the freshwater crayfish.
The results of experiments on the uptake of K+ and Ca"1"1" showed that, in
solutions containing NH^ in concentrations of 0.25 to 2.0 meq/L, a depres-
sion of potassium uptake from the medium and an increase of its yield from
the organism are observed as early as the first 30 minutes. Similar, but
more pronounced, changes take place in calcium exchange. The addition of 2
OT
C
I
&
240
200
160
120
80
40
0
1 - Sodium loss
2 - Chlorine loss
3 - Chlorine uptake
4 - Sodium uptake
12345
NH^CI Concentration, mmole/L
Figure 5. Effect of NlfyCl on sodium and chlorine exchange
in the narrow-clawed crayfish (temperature 18-21 °C, pH
7.6).
63
-------
TABLE 1. EFFECT OF FUNCTIONAL ACTIVITY OF Na+-TRANSPORTING SYSTEM ON THE
AMMONIUM EXCRETION OF THE CRAYFISH (after Vinogradov et al. 1982)
Time of
exposure
to isotonic
solution
(hours)
Control (0)
0.5
24
Surrounding
medium
0.2 mmole/L of
NaCl
distilled water
190 mmole/L NaCl
190 mmole/L NaCl
+ dinitrophenol
190 mmole/L NaCl
190 mmole/L NaCl
+ dinitrophenol
Total Na+
uptake,
ueq/gh
0.185+0.022
0
0.540+0.030
0.188+0.027
0
0
NH4+
excretion,
ueq/gh
0.267+0.27
0.157+0.036
0.384+0.039
0.232+0.056
0.155+0.042
0.140+0.061
Na -dependent
NH^+ excretion,
ueq.gh
0.110
0
0.229
0.092
0
0
+ - confidence interval for p = 0.05
mmole/L of ammonium fluoride to the medium causes a three—fold increase in
calcium loss, whereas its uptake is almost completely suppressed (Figure 6).
When the crayfish are kept in a 2 mmole/L NH4C1 solution for 7 days, uptake
of calcium from water is lower than in the control and amounts to 33, 39, and
43% of the control in the 36th, 96th, and 168th hour, respectively.
The rate of ammonium excretion depends on the concentration of potassium
and sodium in the ambient medium and correlates with their rate of uptake,
indicating the existence in the crayfish of a mechanism of ammonium excretion
associated with calcium transport (Figure 6). A comparison of the rates of
sodium and calcium uptake and rate of ammonium excretion shows that calcium
can be exchanged for ammonium in the ratio of 2:1.
Another factor in the transport of cations in addition to ammonium can
be the hydrogen ion (Ehrenfeld 1974). In this case, an increase in the ex-
ternal concentration of hydrogen ions should cause a depression of calcium
and potassium uptake. In our experiments, a decrease in water pH from 6.5
to 4.0 did not lead to any changes in the transport and escape of potassium
or in the uptake of calcium. The passive loss of calcium on acidification
of the medium sharply increases. This effect is probably due to an increase
in the solubility at low pH values of calcium carbamate, which enters into
the composition of the exoskeleton.
64
-------
o o 30°
o o
o o
200
-0.3
o~§ 100
1 - Calcium loss in distilled water
2 - Calcium uptake from a CaCI solution of 150/xmole/L
3 - Potassium loss in distilled water
4 - Potassium uptake from a KCL solution of 50/z.mole/L
0 0.5 1.0 1.5 2.0
NH4CI Concentration in
Ambient Medium, mmole/L
Figure 6. Effect of ammonium on potassium and calcium
exchange in the narrow-clawed crayfish (temperature
18-21 °C, pH 6.5).
In crayfish placed in water at pH 4.5, the rate of the total loss of
calcium increased by a factor of 8-9 in comparison with the control and re-
mained unchanged during the entire experiment (7 days). The critical pH
value at which the transport of calcium is unable to offset its losses amounts
to 4.6 to 4.7. It should be emphasized that under standard conditions, the
loss of calcium amoaunts to only 1/7 to 1/8 of its uptake. Thanks to this
characteristic of calcium exchange, a constant accumulation of calcium in the
organism is evidently ensured.
DISCUSSION
Numerous literature data generalized in review articles (Liebmann 1960,
EPA 1977, Thurston et al. 1979 and 1982, EPA 1983) attest to a high toxicity
of the NH^ species. In this connection, the criterion of the European Con-
sultative Commission on Fish Breeding in Inland Waters and of the U.S. Environ-
mental Protection Agency is based on the premise that NH^ is not very toxic
to aquatic organisms. There is no question that the acute toxicity of ammonium
(NHg + NH^ ) is determined mainly by the concentration of NHo. In estimating
water quality, however, it is necessary to consider the NHg content as well.
Our studies in fish showed that ammonium ions (NH^ ) inhibit the uptake
of calcium. This effect manifests itself when the concentration is only 4.5 to
65
-------
18.0 mg/L NH/1". It is well known that the normal course of many metabolic re-
actions depends on the availability of calcium to the fish organism (Romanenko
et al. 1982).
It also is known that the concentration of calcium in water, and hence,
the rate of its uptake by the organism determine the development and growth
of fish to a considerable extent (Yeleonskiy 1932, Skadovskiy 1954). It was
found that immature fish are most sensitive to a calcium deficiency. This
applies particularly to crustaceans. A calcium deficiency decreases the
growth of fingerlings by slowing down the mineralization of the skeleton
(Kaplanskiy and Boldyreva 1934, Bodrova and Krayukhin 1962).
In this connection, it is interesting to note that a similar negative
influence on fish is exerted by sublethal ammonium concentrations, as is
manifested most clearly in juveniles (Thurston 1981, EPA 1983). A recalcu-
lation of sublethal NH3 concentrations for different species of fish (0.05
to 0.15 mg/L) cited in literature sources, for the ionized form of ammonium
indicates that the NH^* content in these experiments usually amounted to
4 to 30 mg/L. That is, the NH^+ concentration in water was sufficient to
disturb the calcium exchange. In our view, this fact accounts for much of
the negative influence of sublthal ammonium concentrations on fish and inver-
tebrates, especially in the early stages of ontogeny.
In discussing the toxicity of ammonium to aquatic animals, we would like
to emphasize, in particular, that the results of our studies clearly indicate
disturbances in ion exchange at sublethal concentrations of total and ionized
ammonium. The results obtained and analysis of literature data regarding the
effect of sublethal ammonium concentrations lead to the conclusion that the
chronic toxcity of ammonium contamination and the negative influence of sub-
lethal values of the total ammonium content of water are apparently determined
not only by the NH^ species but also by NH^ .
Without belittling the value of the reported data and our own results
for NH^+/Na exchange in fish and crustaceans (Shaw 1960, Maetz and Garcia 1964,
Maetz 1973, Evans 1975, Payan 1978, Vinogradov 1981, Vinogradov et al. 1983,
Randall and Wright 1986), we should note in conclusion that apparently, in
stenohaline freshwater fish, sodium uptake is primarily associated with the
excretion of H+, and the removal of ammonium ions from the organism may take
place as a result of exchange for Ca2+. A diagram supplementing the concept
of ion exchange processes in the gills of freshwater fish and crustaceans is
given in Figure 7.
REFERENCES
Bodrova, N.V. and B.V. Krayukhin. 1962. The importance of calcium for
immature carp. Some aspects of digestive physiology and metabolism in
fish. Kiev.
Cameron, J.N. and N. Heisler. 1983. Studies of ammonia in the rainbow
trout: physiochemical parameters, acid-base behavior and respiratory
clearance. J. Exp. Biol. Vol. 105.
66
-------
H*
Na"1
H
NH
NH
H
HH
HCO
Gill epithelium
Water
Figure 7. Diagram illustrating the participation of
ion exchange in the gills of freshwater fish and
crustaceans in the maintenance of acid-alkali and
osmoionic homeostasis ( ionic and molecular
diffusion, ion exchange mechanism).
Ehrenfeld, J. 1974. Aspects of ionic transport mechanisms in crayfish
Astacus oeptodactilus. J. Exp. Biol. Vol. 61.
EPA. 1977. Ammonia toxicity to fish. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina.
EPA. 1983. Chronic toxicity to animals. Ammonia. U.S. Environmental Pro-
tection Agency, Duluth, Minnesota.
Evans, D.H. 1975. Ionic exchange mechanisms in fish gills. Comp. Biochem.
Physlol. 51A(3).
67
-------
Kaplanskiy, S. and N. Boldyreva. 1934. Concerning the regulation of mineral
metabolism in homoosmotic fish in the presence of different mineral
compositions of water. Fiziol. Zh. 17(1).
Kersteller, T.H., L. Kirschner, and D. Rafuse. 1970. On the mechanism of
Na+ ion transport by the irrigated gills of rainbow trout (Salmo
galrdneri). J. Gen. Physiol. 56(2).
Kormanik, G.A. and J.N. Cameron. 1981. Ammonia excretion in the catfish:
the role of diffusion. Amer. Soc. Zool., 21(4).
Liebmann, H. 1980. Toxicologie des Abwassers. Handbuch der Frischwasser-
und Abwasser - Biologie.
Maetz, J. and R.F. Garcia. 1964. The mechanism of sodium and chloride up-
take by the gills of a freshwater fish, Carassius auratus II. Evidence
for NH4+/Na+ and HC03~/C1~ exchanges. J. Gen Physiol. 47:1209-1227.
Para, A. Ye. and 0. Prekup. 1960. Study of excretory processes in fresh-
water fish: Report 2. Effect of temperature of the medium on excretory
processes in the pond carp and crucian carp.
Prosser, L. 1977. Comparative animal physiology. Moscow. Vol. 1.
Randall, L.D. and P.A. Wright. 1986. Production and excretion of ammonia
in fish. Problems in aquatic toxicology, biotesting and control of
water quality. Leningrad.
Romanenko, V.D., O.N. Arsan, and V.D. Solomatina. 1982. Calcium and phos-
phorus in the vital activity of aquatic organisms. Naukova dumka. Kiev.
Shaw, J. 1960. The absorption of sodium ions by the crayfish Astacus pallipes
Lereboullet III. The effect of other cations in the external medium.
J. Exp. Biol. 37(3).
Skadovskiy, S.N. 1954. Hydrobiological studies of fishery management of
ponds. Vestn. Mosk. No. 5.
Thurston, R.V., R.C. Russo, C.M. Metterolf, T.A. Edsall, and V.M. Barber.
1979. A review of the EPA Red Book: Quality criteria for water. Amer.
Fish. Soc., Bethesda, Maryland.
Thurston, R.V., G.A. Vinogradov, V.T. Komov, and V. Ye. Matey. 1979. Effect
of low pH values, ammonium salts and desalting on the activity of en-
zymes, sodium exchange in gills, and ultrastructure of chloride cells
in freshwater fish: report 1. Biology of Inland Waters: Inform. Byul.
No. 43.
Thurston, R.V. 1981. Factors affecting the toxicity of ammonia to fish.
• Theoretical aspects of aquatic toxicology: Proceedings of Third U.S.-
Soviet Symposium. Leningrad.
68
-------
Thurston, k.V., R.C. Russo, and E.L. Meyn.
Abstracts. Bozeman, Montana.
1982. Ammonia toxicity.
Vinogradov, G.A. 1981. Processes of ion regulation in freshwater animals
in the presence of anthropogenic contamination of bodies of water.
Biology of Inland Waters: Inform. Byul., Leningrad. No. 51.
Vinogradov, G.A., Ye. S. Dal', and V.T. Komov. 1982. Effect of ammonium
salts and acidification of the medium on metabolic processes in fresh-
water animals. Study of NHA+/Na+ and H+/Na+ exchange in the gills of
the crayfish and stickleback.
Leningrad. No. 56.
Biology of Inalnd Waters: Inform. Byul.
Vinogradov, G.A., Ye. S. Dal', and V.T. Komov. 1983. Study of the basic
functions of freshwater crayfish gills acted upon by ammonium salts and
acidification of the medium. The reaction of aquatic organisms to con-
tamination. Moscow.
Yeleonskiy, O.N. 1932. Principles of fish breeding. Moscow. Vol. 2.
69
-------
ON-SITE TOXICITY TESTING; APPLICATIONS
IN THE UNITED STATES AND SOVIET UNION .
by
M.G. Henry*, B.A. Flerov2, V.T. Komov2, and
T.A. Heming3
ABSTRACT
On-site toxicity tests using the cladoceran Ceriodaphnia dubia were con-
ducted in the USA and USSR by the same team of investigators. A non-point
source situation (Detroit River, USA) and a point source spill situation
(Rybynsk Reservoir, USSR) were assessed. The test method was successfully
applied in both instances. The resultant compatability of approach affords
opportunities to accurately compare future data sets from both countries.
INTRODUCTION
One of the focuses of Project 13 of the USA-USSR scientific exchange pro-
gram for cooperation in the field of environmental protection is the develop-
ment of aquatic toxicity test methods that can be field applied in both coun-
tries. Species differences, changes in the life histories of similar species
due to latitudinal differences, and inconsistencies in available equipment
present formidable obstacles in the development and application of common
methods. Use of the cladoceran Ceriodaphnia dubia in on-site assessments
offers an opportunity to overcome these difficulties. C. dubia is widely dis-
tributed in both countries, has a short life-cycle, and is sensitive to low
levels of contaminants. Moreover, the test protocol developed for this spe-
cies by Mount and Norberg (1984) is relatively simple, requires no specialized
equipment, and can easily be modified to compensate for temperature-related
differences in life-cycle (McNaught and Mount 1986, Cowgill et al. 1985).
The objectives of this research were twofold: to establish healthy cul-
tures of Ceriodaphnia dubia in the USSR and to use the same team of American
and Soviet investigators to conduct on-site toxicity tests in both countries
so that dissemination of the method and compatibility of approach could be
accomplished.
^•National Fisheries Center-Great Lakes, U.S. Fish and Wildlife Service, Ann
Arbor, MI USA.
^Institute of Biology of Inland Waters, USSR Academy of Sciences, Borok,
Jaroslavl Oblast USSR.
^University of Texas Medical Branch, Galveston, TX USA.
70
-------
MATERIALS AND METHODS
The Great Lakes and their connecting channels are located in the north
central part of the United States. They constitute an important commercial
and recreational resource and are heavily used by industries. The Detroit
River, which connects Lake St. Clair and Lake Erie, is no exception. Several
chemical and steel manufacturing plants are located along the banks of the
Detroit River. Numerous organic and inorganic contaminants have been de-
tected in Detroit River sediments but the toxicity of Detroit River water
(DRW) needs needs closer examination. We selected a representative site in
the Trenton Channel of the Detroit River (Figure 1) as our test site. This
site was located downstream from steel and concrete manufacturing plants.
Sediments collected at this site have been found to contain elevated concen-
trations of certain polyaromatic hydrocarbons and metals (Giesy et al.
1988).
The on-site toxicity test was conducted during August 1986. Because of
the nonpoint source nature of contaminants in the Detroit River, a continuous
water sampling scheme was established at the test site. Samples were taken
3.5 m below the water surface with a shore-mounted Masterflex Composite
Sampler. A 500-ml aliquot of DRW was sampled every 15 min for 7 days. Inte-
grated 24-h samples were collected each morning for use in the static renewal
Ceriodaphnia test.
The Ceriodaphnia test followed methods outlined by Mount and Norberg
(1984). The test was conducted in a constant temperature (24-25°C) environ-
mental chamber at the USEPA Large Lakes Research Station, Gross lie, Michigan.
Test animals were cultured in reconstituted dilution water. The dilution
water was a mixture of well water and reverse-osmosis water with approximately
the same hardness as Detroit River water. Ten beakers with one neonate (less
than 24-h old) in each were used at each exposure level. Six exposure levels
were assessed: 100% Detroit River water, 50% DRW; 25% DRW, 12.5% DRW, 6.25%
DRW, 0% DRW (100% dilution water). Mortalities and number of neonates produced
were monitored daily. The results were statistically analyzed using methods
developed by Kaiser and Finger (1986).
On test days 1, 3, and 6, aliquots of the integrated 24-h water sample
were prepared for organic analysis by fixing 3-liter subsamples with 150 ml
of dichloromethane. Two 200-ml samples were also taken on each of these days
for inorganic analysis. One was filtered and the other was left unfiltered.
Both were fixed with 2 ml of ultra-pure nitric acid. Organic samples were to
be analyzed by gas chromatography/mass spectroscopy and inorganics analyzed
using atomic absorption spectrophotometry.
USSR TEST
In January 1987, a wastewater spill occurred at a metallurgical plant
located on the northern shores of the Rybynsk Reservoir at Tcherepovets,
adjacent to station 1 (Kabachino, Figure 2). Thousands of gallons were re-
leased, causing immediate fish mortality and water quality problems - (Pant-
zirev 1987). At the first available opportunity after ice melt, grab samples
71
-------
LtktSt. Clair
Figure 1. Sampling station location in the Detroit River for
the USA ori-site toxicity test conducted in 1986.
72
-------
of water were taken at nine stations along a north-south transect moving away
from the spill site to assess the persistent effects of the spill. Water
samples were taken 4 m below the surface using a van Dorn bottle.
The Periodaphnia test was conducted at the Freshwater Institute for the
Biology of Inland Waters at Borok in June 1987. Static renewal was used and,
because of the lower-than-recommended temperature of available reservoir
water (21 to 22°C), the test was run for 10 days. All other methods were
those recommended by Mount and Norberg (1984). Water from the southernmost
tip of the Rybynsk Reservoir at Borok below station 10 (Volga, Figure 2) was
used as control and culture water. Chemical analyses of water, fixed and
prepared as described above, were conducted using atomic absorption spectro-
Koschta
Jagorba
Kabachino
ach
Figure 2. Location of sampling stations in the Rybynsk
Reservoir for the USSR on-site toxicity test conducted
in 1987.
73
-------
photometry and gas chromatography. Results of the toxicity test were analyzed
using analysis of variance.
RESULTS
A 48-hr acute lethality bioassay conducted in conjunction with the USA
test indicated that 81% DRW induced mortality in 50% of the adult test popu-
lation. Complete mortality also was induced during the 7-day on-site test in
the 100% and 50% DRW treatments. Results of the 7-day reproduction test were
considered to be inconclusive because the requirement for production of three
broods of at least nine young in the control (Mount and Norberg 1984) was not
achieved.
Despite the lack of protocol-required reproduction, the data suggest that
there was a dose-response relationship in mortality and reproduction associ-
ated with DRW dilution (Table 1). Complete mortality was induced in the two
highest dilutions, 100% and 50% DRW within 2 and 4 days, respectively. The
25% DRW treatment induced 30% mortality and, of the remaining survivors, two
broods were produced with only a mean number of 3.1 neonates in each. Mortal-
ity in the remaining three treatments was insignificant and statistically in-
distinguishable from each other. Three broods were produced in the 12.5% DRW
treatment, whereas four were observed in both the 6.25% and control treatments.
Due to lower neonate/brood production, the results of the bioassay were con-
sidered incomplete and water samples were not chemically analyzed.
Initial cultures of £. dubia were established in the USSR in July-August
1984, using animals collected from the Rybynsk Reservoir. These cultures
were subsequently found to be contaminated with JC. reticulata. Stable cul-
tures of pure C. dubia were achieved at the Freshwater Institute for the
Biology of Inland Waters by August 1986.
Results of the Rybynsk Reservoir on-site test were more complete than
the USA test. A definite dilution effect was apparent, even 4 months after
the spill (Table 2). Complete mortality was induced by water collected at
four sites; partial mortality was induced by water from the other six sites.
Three sites where 100% mortality was induced were upstream from the spill
site. Control reservoir water did not induce any mortality. Reproduction,
of course, was non-existent in treatments where 100% mortality occurred but,
of the survivors in other treatments, reproduction was completely truncated
in water from station 1 and was statistically indistinguishable from control
levels in waters from stations 5, 6, 7, 8, and 10. Neonate production approx-
imated the protocol-recommended levels. Chemical analyses of water samples
are currently being completed.
DISCUSSION
Although some of the test results were inconclusive, this does not dim-
inish the significance of the research. This is the first time identical
species have been used in common bioassays conducted in the USA and the USSR.
Because pollution abatement and monitoring is a global problem, the compati-
74
-------
TABLE 1. EFFECTS OF DETROIT RIVER WATER ON CERIODAPHNIA SURVIVAL AND
REPRODUCTION.
$ Mortality/Mean Neonates per Adult
Day
Detroit River
Water
6
100% 60/0
50% 20/0
25% 10/0
12.5% 0/0
6.25% 0/0
0% 10/0
100/0
20/0
10/0
0/0
0/0
10/0
50/0
30/0
0/3.0
10/3.4
10/3.1
100/0
30/0
0/6.2
0/6.2
10/6.5
30/0
0/0
0/0
10/0
30/0
0/0
0/11.7
0/12.4
30/3.1
0/1.0
0/5.7
10/6.6
bility of approach used here affords us opportunites to accurately compare
future data sets. It further emphasizes that this on-site method can be
applied successfully in point source and non—point source situations.
Establishment of healthy cultures of test organisms is paramount if the
required number of broods and neonates per brood are to be produced. Trans-
port stress to animals may impede reproductive capability and, consequently,
should be minimized. On the other hand, diet and culture water quality also
have been implicated in determining the health of test organisms (DeGraeve
and Cooney 1987). Because brood production is species- and temperature-de-
pendent, maintaining pure cultures under recommended conditions is important.
Furthermore, although taxonomic distinction between C. dubia and C. reticulata
can be difficult, culture purity must be periodically monitored because a lack
of culture purity can influence the test results.
Despite the lack of protocol-required reproduction in the Detroit River
test, the data indicate that Detroit River water is toxic to £. dubia. This
emphasizes the need for further research.
Mortality patterns observed in Periodaphnia exposed to water from the
Rybynsk Reservoir spill appear to have been influenced by the hydrology of
the reservoir. The flow patterns of water in the northern portion of the
reservoir are such that the plant wastewater moved slightly north before mov-
ing south through the central basin (Flerov 1987). The mortality observed in
animals exposed to water from station 9, however, is not explainable on the
basis of water currents. Perhaps results of chemical analyses will aid in
75
-------
q
ed
w
CA
1
CO
w
CO
o
w M
H Ej
M «$
CO H
I lo
BS
fa J
O M
D-(
CO CO
H
J H
a z
CO •<
a
CM
(^
PM
r-l
«
(8 3 0)
o) o o,
a o M «
a) o a) a
CO >
CO
CO
00
CO
3
8
•H (8
&
?l • II tl
OO OO
ON CM
00 CO 00 OO
CM
-------
the interpretation of this portion of the test results. Areas requiring
further development are 1) adapting available technology in the Soviet Union
for taking integrated water samples and 2) writing a manual for circulation
in the USSR so that personnel associated with water management offices and
sewage treatment facilities also can employ this versatile method.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the field assistance and cooperation
of Steve Smith, NFRC-GL. Without his help, this project would never have
been possible. Additional thanks go to Susan Finger for her assistance with
test protocol recommendations and data analyses and to Mark Kaiser for data
manipulation and statistical analyses.
REFERENCES
Cowgill, U.M., K.I. Keating, and I.T. Takahashi. 1985. Fecundity and lon-
gevity of Ceriodaphnia dubla affinis in relation to diet at two differ-
ent temperatures. Journal of Crustacean Biology. 5:420-429.
DeGraeve, G.M. and J.D. Cooney. 1987. Ceriodaphnia: an update on effluent
toxicity testing and research needs. -Environmental Toxicology and
Chemistry. 6:331-333.
Flerov, B.A. 1987. Personal communication. Institute of Biology of Inalnd
waters, Borok, Jaroslavl, USSR.
Giesy, J.P., C.J. Rosiu, J.L. Newsted, R.L. Graney, A. Benda, R.G. Kreis and
F.J. Horvath. 1988. Detroit River sediment toxicity. Journal of Great
Lakes Research. In press*
Kaiser, M., and S. Finger. 1986. Personal communication. National Fisheries
Contaminants Research Center, Columbia, Missouri.
McNaught, D.C., and D.I. Mount. 1986. Appropriate durations and measures
for Ceriodaphnia toxicity tests. ASTM STP 891, R.C. Banner and D.J.
Hansen, Eds., American Society for Testing and Materials, Philadelphia,
Pennsylvania, pp. 375-381.
Mount, D.I. , and T.J. Norberg. 1984. A seven-day life-cycle cladoceran
toxicity test. Environmental Toxicology and Chemistry. 3:425-434.
Pankratov, A., and V. Pantzirev. 1987. The day the fish came to the surface.
Komsomolskay Pravda. 1-3.
77
-------
VARIOUS RESISTANCE MECHANISMS OF CARP (Cyprinus Caprlo L.) AND
PERCH (Perca Fluviatilis L.) TO DDVF ORGANOPHOSPHORUS COMPOUNDS
by
G.M. Chuyko*
ABSTRACT
Differences in the resistance of carp and perch to the effects of
dichlorofos in acute tests are examined. The experiments demonstrate that
differences are determined by the toxin's different rate of penetration into
the fish and the uneven intensity of subsequent detoxification processes.
INTRODUCTION
An increase in world-wide farming production implies a greater reliance
on pesticides, among which insecticides and. acaricides play a major role.
The principal volume of these falls in the organophosphorus category that,
in comparison to other types of chemical substances has a broader selectivity
range for animals (Melnikov et al. 1977, Melnikov, 1981). The synthesis of
compounds that are highly selective and narrowly directed in their effects is
based on a thorough knowledge of the resistance mechanism of various animal
groups (O'Brien 1964, Rozengart and Sherstobitov 1978).
Fish have varying degrees of resistance to organophosphorus. As data of
various authors indicate, the most resistant to a majority,of compounds in
this group are members of the Cyprinidae family, and specifically,, the
Salmonidae and Percidae families (Gantverg 1985, Macek and McAllister 1970,
Johnson and Finley 1980). The selective effect mechanisms of organophosphorus
on fish is practically unknown. Current understanding has been based on the
research on mammals and arthropods (Rozengart and Sherstobitov 1978). Accord-
ing to their work, the most probable causes of varying resistance levels in
fish may be type-specific characteristics of the toxicant's rate of penetra-
tion into the organism and their sensitivity to the "target" acetylcholines-
terase (ACE), and its metabolic intensity. No complete study of the selec-
tivity of organophosphorus on fish has been done to date, although certain
aspects of this issue have been examined in a number of studies (Gantverg
1985, Murphy, 1966, Macek and McAllister, 1970, Fujii and Asaka 1982). For
1Institute of Biology of Inland Water, USSR Academy of Science, Borok USSR.
78
-------
this reason, the intent of this paper is to study the primary reasons for the
varying degree of resistance to one of the organosphosphorus compounds,
"dichiorophos" (DDVP).
DATA AND METHODOLOGY
Two types of fresh water fish were utilized in the experiments, each
with varying degrees of resistance to organophosphorus compounds—the carp,
Cyprinus carpio L,. (genus Cyprinidae) and the perch, Perca fluviatis Ij.
(genus Percidae). For the tests, carp were obtained from the All-Union Fish
Experimental Facility, (Moscow Oblast); the perch were caught by net from the
Rybinsk Reservoir. The carp were 1-year and 2-year old fingerlings (weight
5.7+0.4, 10.1+1.4 and 65.0+6.1 g); the perch were 2-year and 4-year old finger-
lings (weight 3.6+0.3, 22.7+0.8 and 75.2+9.4 g). When studying the ACE sensi-
tivity to DDVF, fish were selected immediately after catching. For other
experiments, fish were adapted to laboratory conditions for no less than 7 to
10 days prior to the start of the experiments.
The toxic compound consisted of purifying 97% DDVF preparation: 0.0-
dymethyl-0-(2.2-dichlorovinyl)phosphate. Concentration was calculated accord-
ing to the active ingredient.
Fish resistance to the effects of the toxin was evaluated in acute tests
using assay-analysis inherent modification (Nepomnyashchikh and Chuyko 1986),
based on the lethal concentration LK.5Q or LD5Q dosage, causing 50% mortality
of the test fish over a 48-hour period. Six or eight fish were placed in 30-L
aquariums, with settled river water at 8.1 to 8.3 pH, 16 to 18°C, and 8.7 to
9.8 mg/L oxygen content. The toxic compound was preliminarily dissolved in
acetone, and then, depending on experimental conditions, was placed into the
water or into the physiological solution for the fish. End volume concentra-
ration of the acetone in each aquarium did not exceed 0.1%. To determine
LD5Q the DDVF physiological solution was introduced intraventrally into the
abdominal cavity, calculated per 1 kg of live weight. The injected solution
was 0.5 ml, with the amount of acetone per fish not exceeding 0.05 ml. Fish
injected with a physiological solution with acetone, but without the toxin,
served as control.
Sensitivity of fish ACE to DDVF was evaluated in vitro according to
magnitude of the biomolecular constant of the inhibition rate k^ (Yakovlev
1965). Ferment activity was determined in brain homogenates using Ellman's
method (Ellman et al., 1961) and the Maslova and Reznik modification (1976).
Acetylthyocholine bromide served as substrate. Homogenates were prepared
using a phosphate buffer with a 7.5 pH, 1:10 ratio. After homogenization, the
samples were centrifuged at 5000 rpm for 10 minutes. The supernatant fluid
was analyzed.
The content of DDVF in the blood and liver of the fish was determined by
chromatographic self-modification. DDVF extraction was carried out using
hexane with additional evaporation of the extract under vacuum at 40°C. DDVF
extraction from blood comprised 91.6+0.6% and 87.2+2% from liver. Quantita-
tive content of DDVF was evaluated according to peak level, using absolute
79
-------
calibration methods. A Gazokhrom-1109 liquid gas chromatograph with an
electron-grip detector, and a packed glass spiral column 1000 x 3 was util-
ized. Nitrogen served as gas carrier. Five percentile SE-30, N-AW chroma-
tone was utilized as stationary phase. Chromatography conditions were an
electrometer scale 2 x lO^^A, column temperature 110°C, detector 130°C,
evaporator 200°C, nitrogen blow-through velocity 73 cm^/min, speed ot record-
ing unit 200 mm/hr, hold time for DDVF 2 minutes and 35 seconds. Sensitivity
detection was 0.5 ng. One to two ml of the extract were placed into the
chromatograph.
Two series of tests were conducted in order to study the rate at which
DDVF enters the fish organism from the time they are kept in the toxic solu-
tion, as well as the concentration of the toxin in water. In one series, the
fish were kept from 5 to 60 minutes in a 21.y mg/L DDVF solution; in another,
the fish were placed for 5 minutes in solutions containing 5, 10, 15, and
21.9 mg/L of toxin and a measured amount of DDVF in the blood of the fish.
Blood was taken from tail blood vessels after caudotomy.
Fermentative destruction of DDVF was determined on the basis of differ-
ences in toxin content after incubation at 30°C for 2 hours with 2 ml of in-
tact and inactivated fish liver homogenates (thermal processing for 10 minutes
at 80°C). Original content of DDVF in sample was 45 mkg. Rate of the fer-
mentative destruction of the toxin was expressed in DDVF mkg, 1 g raw tissue
destroyed in 1 hour. Liver homogenates were prepared like the brain homoge-
nates—1:5 ratio for perch and 1:10 for carp.
All data were statistically processes at p=0.05 and represented in the
form of means (x), their errors (mx) and confidence interval boundaries.
Accuracy of results were calculated according to (Stedent's) criteria (Lakin
1969). Each experiment was repeated no less than 2 times, and the number of
fish per point (n) was no less than 5.
RESULTS AND DISCUSSION
COMPARATIVE RESISTANCE OF FISH TO DDVF
The toxicological experiments establsihed that the carp is significantly
superior to the perch in its resistance to DDVF. The value LK0 comprised
21.9 and 0.59 mg/L, respectively (Table 1). Our data correspond to results
derived by other researchers on the same types of fish (Gantverg 1985,
Johnson and Finley, 1980, Svobodova 1980). The carp is also more resistant
than the perch when the toxin is injected intra-abdominally. LD5Q values
are 292.0 mg/kg for carp and 30.4 mg/kg for perch (Table 2).
When resistance ot fish to the effects of the toxin is compared, the
selective coefficient Ks is utilized which is represented by the relationship
of values LK5Q (LD50) for the more resistant fish, to LK.50 (LD50) for the
less resistant fish (Perevoznikov 1979, Gantverg 1985). It is believed that
the comparison of Ks in direct contact with the toxin and intra-abdominal or
intravenous injection makes it possible to evaluate separately the role of
penetration processes, and the "internal" reasons for the toxic selectivity
80
-------
TABLK 1. VALUES OF DDVF LK50 FOR PERCH AND CARP FOR 48-hr EXPOSURE
Type
of Fish
Number
of Fish
mg/L
Boundary of
confidence
interval LK50, mg/L
Kc
Carp
Perch
38
42
21.9
0.59
10.2 - 23.8
0.54 - 0.64
37.1
effects that are connected with the fate of the toxin inside the organism.
Apparently, if the toxin's rate of penetration into the organs of the fish
does not differentiate, when it is injected intra-abdominally or intraven-
ously, which eliminates the toxin's penetration through the external tissues,
then the Ks value must remain close to that of the value that results from
external contact with the toxic medium.
In many other cases, when Ks is increased or decreased, one of the causes
of selectivity appears to be the inter-genus differences in the toxin's rate
of penetration. This method is widely utilized in the study of the selectiv-
ity of organophosphorus in mammals and arthropods (O'Brien 1964, Rozengart
and Sherstobitov 1978). Use of this method with respect to fish has proven
that the determining factor of higher toxicity in the Amur lamprey (Misgurnus
anguillicaudatus) to DDT and Dieldrin, in comparison to "monokrotofos" and
"dikrotofos" is the higher speed of penetration of the former into the fish
organism (Yang and Sun 1977).
TABLE 2. VALUES OF DDVF LK50 FOR PERCH AND CARP FOR 48-hr EXPOSURE
Boundary of
Type
of Fish
Number
of Fish
LK50>
mg/L
confidence
interval LlOjQ , mg/L Ks
Carp
Perch
20
42
292.0
30.4
254.0 - 336.0
23.0 - 40.1
9.6
81
-------
In our tests, the value Ks for carp and perch when placed in a toxic solu-
tion was 37.1, and 9.6 when injected intra-abdominally (Tables 1 and 2), that
is, Ks is reduced by four times when DDVF is introduced ventrally. It follows
then that the higher resistance of carp in comparison to. the perch may be
explained by the inter-genus differences in the DDVF penetration process into
the fish organism.
PENETRATION OF DDVF INTO FISH ORGANISMS
Results derived from experiments to determine DDVF content in the blood
of carp and perch after being kept in a toxic solution confirm our explanation.
It has been establsihed that 5 minutes after fish are placed into a 21.9 mg/L
toxin solution, which corresponds to value LK5Q for carp when exposed for 48
hours, there is a significant amount of DDVF in the blood—1.3 mg/L for the
carp and 5.1 mg/L for the perch (Figure 1). Note that the content is 4.5
times higher for the perch than for the carp. This value is close to the one
derived by comparing Ks for the different methods by which toxin is acquired.
As exposure time for the fish in a DDVF solution is extended, DDVF con-
tent in the blood rises and the establshed differences are maintained for the
duration of the entire period of observation (60 min.). For the perch, the
21.9 mg/L concentration was extremely lethal because fish died 45 minutes
after start of tests. This, apparently, explains the tact that during the
period immediately preceding death, when breathing in fish becomes less
active, the amount of DDVF in the perch's blood stops rising.
•o 20
o
_
16
.2 12
-*-»
E
•+J
c 8
-------
Also studied was the relationship of the initial content of DDVF in the
blood of carp and perch to DDVF concentration in water. For this purpose,
extreme lethality data for a 48-hour exposure period to the toxin was selec-
ted. The lowest possible concentration is taken as being a bit higher than
LK.5Q values for the perch, because with lower concentrations, a precise deter-
mination of DDVF concentration in the blood by LGC methods is somewhat compli-
cated. It has been established that with all the researched toxin concentra-
tions in water, DDVF content in the blood of the perch is definitely higher
than in the carp (Table 3). These data, shown as a graph, indicate that the
relationship of toxin content in the blood to toxin content in the water has
a linear nature (Figure 2).
Thus, the results attest that one of the reasons for the varying resis-
tance to DDVF for the carp and perch is the unequal rate of penetration of
the toxin into the blood of the fish. This rate is four to five times higher
for the perch than for the carp. The estabished pattern is maintained for
the lethal range (LK^Q for 48 hours) of concentrations, 5 to 21.9 mg/L.
The discovery of the differences in DDVF penetration rates for the perch
and carp may be explained from the point of view of the laws of permeability.
It is known that organic-origin toxins, which include organophosphorus com-
pounds, penetrate the fish directly from the water through tissue barriers
(Tinsley 1982, Marcelle and Thome 1984). Penetration of these toxins through
TABLE 3. CONTENT OF DDVF IN CARP AND PERCH BLOOD 5 MINUTES AFTER START OF
EXPERIMENT, WITH TOXIN CONCENTRATIONS OF 5, 10, 15, AND 21.9
mg/L IN THE WATER
Type
of fish
Amount of DDVF in blood (mg/L) with following toxin
concentrations in water
5 mg/L
10 mg/L
15 mg/L
21.9 mg/L
Carp
0.19 + 0.03
(0.13-0.25)*
n = 7
0.46 + 0.03
(0.40-0.53)
n = 5
0.61 + 0.05
(0.52-0.70)
n = 6
1.30 + 0.12
(1.06-1.54)
n = 8
Perch
1.69 + 0.07
(1.56-1.82)
n = 5
2.40 + 0.19
(2.03-2.77)
n - 5
3.48 + 0.13
(3.23-3.73)
n = 4
5.11 +_ 0.23
(4.66-5.56)
n = 5
*Parentheses indicate boundaries of confidence interval at p = 0705
n - number of fish per section
83
-------
C7»
O
O CO
Perch
Carp
0 5 10 15 20 25
DDVF Concentration in Water, mg/L
Figure 2. Relationship of DDVF accumulation in
perch and carp blood as a result of the toxin
concentration in water.
biologic membranes is determined by the ability of these compounds to dis-
solve in lipids; this occurs through the means of ordinary diffusion accord-
ing to Pick's law.
According to this law, the rate of diffusion of a particular compound,
in the presence of a stable concentration in an external medium, depends on
the area and thickness of the barrier through which diffusion occurs—the
larger the area, and thinner"the barrier, the more rapid the diffusion of the
compound (Tinsley 1982). The principal tissue barriers in the fish that
separate the internal medium (blood) from the external medium through which
toxic compounds found in the environment may be transported are the epithelia
of the gills and the skin. Due to their morphofunctional peculiarities, the
gills are the point of principal contact between the internal and external
media of the organism (large surface area, thin tissue barrier and high de-
gree of vascularization); the skin is of secondary importance. Consequently,
gills have an important role in such processes as osmoregulation and ion ex-
84
-------
change (Evans 1979). Based on the morpholunctional features of the gills,
and the nature of the diffusion process, It can be considered that the pene-
tration of organic toxins into fish, including organophosphorus compounds,
occurs primarily through the gills. Other researchers have similar beliefs
(Tinsley 1982, Yang and Sun 1977).
Gills in different types of fish have varying characteristics. In par-
ticular, it has been established that the overall surface of the gills in
highly active, fast swimming fish are larger than in slow-swimming bottom
fish. Differences can be measured by many factors of ten. The differences
in tissue barrier thickness are also on the same order (Amineva and Yarzhombek
1984, Hughes 1984). It is precisely these differences that may be the cause
of the unequal rate of DDVF penetration into the perch and carp organisms.
The fact that the more active fish are less resistant to toxins, as compared
to less active ones, may serve as an indirect confirmation of this conclusion.
The derived results only partially explain the selective effects of DDVF
on the perch when toxin contact is external, and do not explain at all the
higher resistance of the carp when toxin is injected intra-abdominally. In
this case, the selectivity is determined only by internal causes, which in
turn determine the fate of the toxic substance in the body of the animal.
Some of the main reasons are the different levels of sensitivity to the toxic
substance by the fish's ACE in the nervous system, and unique features of its
metabolism.
DDVF INHIBITION OF FISH BRAIN ACE IN VITRO
Organophosphorus compounds are neuroparalytic poisons. ACE serves as a
principal target of their effects inside animal organisms with "cholinergic"
systems. The symptoms of acute poisoning in amlraals by organophosphorus com-
ounds are tied to the depression of this ferment (O'Brien 1964, Rozengart and
Sherstobitov 1978).
Various researchers have expressed the opinion that differences in ACE
sensitivity of the nervous system in fish may be a determining factor in their
different resistance to organophosphorus compounds (Lukyanenko 1983, Macek
and McAllister 1970). For this reason, the sensitivity of the brain of the
two fish to DDVF also was examined. It was established that the value k^ for
both fish is close in magnitude and is on the same order, which attests to the
fact that both are equaly sensitive to the toxin (Table 4). There are small
differences (p = 0.05), but these are not important enough to be the cause of
the resistance difference in the fish.
The absence of inter-genus differences of brain ACE sensitivity to a number
of organophosphorus compounds, including DDVF, by other types of fish is sup-
ported by several authors (Hogan and Knowles 1968a, Hogan 1971), and our own
efforts (Chuyko 1987). Thus the sensitivity of brain ACE in the examined fish
does not play a significant role in the DDVF selective mechanism in the perch.
A study of the nature of resistance in fish to "karbofos" brought identical
results (Gantverg 1985).
85
-------
TABLE 4. RATE OF INHIBITION CONSTANTS (ku) FOR DDVF IN VITRO IN BRAIN ACE
OF CARP AND PERCH
Type of fish
Number fish
mole 1 x 1 x tnin~l
Carp
Perch
7
7
9.1 + 0.3
5.6 + 0.1
DDVF DESTRUCTION BY FISH LIVER HOMOGENATES IN VITRO
As we know, organophosphorus compounds may undergo metabolic transforma-
tions inside the fish organism, as a result of which detoxification occurs
(Gantverg 1985, Fujii and Asaka 1982). This process generally occurs in the
liver, and DDVF is not an exception. It has been established that, in the
liver of the canal catfish, Ictalurus punctatus, and another type of perch,
Lepomis macrochirus, DDVF is destroyed at a rate that significantly exceeds
the rate of spontaneous destruction. Here the variances in the efficiency of
fermentative DDVF destruction are insignificant; the rate of toxin deteriora-
tion is only 1.3 times higher for L. macrochirus, than for I_. punctatus
(Hogan and Knowles 1986b).
Tests were carried out on the detoxifying system of the carp's and
perch's liver showed the presence of fermentative systems in both fish, ones
capable of destroying DDVF. The content of toxin in the samples after incuba-
tion with intact homogenate was significantly lower than with inactivated
homogenates (Table 5).
Based on the data, the specific rates of DDVF destruction by liver homo-
genates of these two fish were shown to be 12.7 for the perch, and 38.8 mg/
g-hr for the carp. These data attest that the specific rate of DDVF destruc-
tion by the carp liver homogenates is three times higher than that of the
perch. As some researchers indicate, in comparison to other types of fish,
the carp also has a higher rate of destruction of such organophosphorus com-
pounds as "diazokson" and "karbofos" (Gantverg 1985, Fujii and Asaka 1982).
At the same time, the fact of a higher specific rate of DDVF destruction in
the carp does not completely explain the almost 10-fold difference in the re-
sistance of this fish when toxin was injected intra-abdominally. Apparently,
the absolute detoxifying capability of the liver depends both on the specific
activity of its fermentative systems, as well as on its overall mass. Study
of the absolute and relative mass of the carps' and perchs' livers (rounded
off to unit of body weight) showed that for the carp this factor is two times
higher than for the perch. Table 6 shows morphometric indicators of the
fish, which were used in the DDVF detoxification experiments.
86
-------
TABLE 5. DESTRUCTION OF DDVF liY INACTIVATED AND INTACT CARP AND PERCH LIVER
HOMOGENATES
Carp
DDVF content, % of
original
Type
of fish
Number
of Fish
Inactivated
homogenate
Intact
homogenate
Specific rate
of fermentative
destruction,
mkg/g/hr
80.0 + 4.7
45.2 + 7.9
38.8 + 5.1
Perch
75.9 + 1.9
54.5 + 2.6
12.7 + 1.4
Other authors show that the relative mass of the liver of these two types
of fish varies depending on the metabolic activity of the liver, the season
of the year, and the degree of ontogenesis; however, mass is greater in the
carp than in the perch (Makarova 1973, Dobrinskaya 1984). It can be thus con-
cluded that the absolute capability of the carp's liver to destroy DDVF is
six times higher than for the perch. This information makes it possible to
explain the differences that were establsihed with the intra-abdominal toxin
injections.
CONCLUSION
As a result of the experiments, it has been established that the differ-
ing resistance of the carp and perch to the effects of DDVF in acute tests
does not depend on the variant peculiarities of fish brain ACE sensitivity
to the toxin but, rather, is determined by the toxin's different rate of
penetration into the fish, and the uneven intensity of processes of the
subsequent detoxification.
REFERENCES
Amineva, V.A. and A.A. Yarzhombek. 1984. Physiology of Fish, Moscow.
Chuyko, G.M. 1987. Biochemical and Physiological Mechanisms of Varying
Resistance of Freshwater Fish to the Effects of "Khlorofos" and "Di-
khlorofos." Abstract of dissertation. Leningrad.
Dobrinskaya, L.A. 1984. Correlation of Increase in Body Weight and Other
Organs in Young Carp (probably Gyprinus carpio ]L.) and Crusian Carp
(Carassius carassius); A Morphobiologic Analysis. Sverdlovsk.
87
-------
Dokin, O.K. 1973. Biometry. Moscow.
Eilman, J.L., K.D. Courtney, I.R. Andres, and R.M. Featherstone. 1961. A
new and rapid colorimetric determination of acetylcholinesterase activity,
Biochem. Pharmacol. 7(1). ,
Evans, D.H. 1979. Fishes. Comparison of Osmoregulation in Animals. London.
Fuji!, Y. and S. Asaka. 1982. Metabolism of diazinon and diazoxon in fish
liver preparation. Bull Environ. Contam. Toxicol. 29(4).
Gantvery, A. 1985. Characteristics of Resistance of Freshwater Fish to Car-
bofos and Other Organophosphoric Pesticides. Dissertation abstract.
Leningrad.
Uogan, J.W. 1971. Brain acetylcholinesterase from cutthroat trout. Trans.
Am. Fish. Soc. 100(4).
Hogan, J.W. and C.O. Knowles. 1968a. Some enzymatic properties of brain
acetylcholinesterase from bluegill and channel catfish. J. Fish. Res.
Bd. 25(4).
Hogan, J.W. and C.O. Knowles. 1986b. Degradation of organophosphates by
tish liver phosphateses. J. Fish Res. Bd. Can. 25(8).
Hughes, G.M. 1984. Measurement of gill area in fishes: Practices and prob-
lems. J. Marine Biol. ASSOC. UK. 64(3).
TABLE 6. MORPHOMETRIC INDICATORS FOR CARP AND PERCH
Indicators
Number of fish
Carp
Perch
Body length, mm
Body weight, g
12
12
141.0 +_ 4.0
64.0 + 5.9
160.0 + 6.0
73.6 + 10.9
Liver weight, mg
Relative liver mass,
mg/g
12
12
1293.0 + 77.0
21.6 + 1.5
745.0 + 12.0
9.9 + 0.7
88
-------
Johnson, W.W. and M.T. Finley. 1980. Handbook of Acute Toxicity of Chemicals
to Fishes and Aquatic Invertebrates. Washington.
Lukyanenko, V.I. 1983. General Ichthyotoxicology. Moscow.
Macek, K.J. and W.A. McAllister. 1970. Insecticide susceptibility of some
common fish family representatives. Trans. Am. Fish Soc. 99(1).
Makarova, N.P. 1973. Seasonal changes of certain physiological indicators
in perch Perca fluviatulis 1^. found in the Ivankovskiy reservoir.
Voprosy Ikhtiologii. 13(5).
Marcelle, C. and J.P. Thome. 1984. Relative importance of dietary and environ-
mental sources of lindane in fish. Bull. Environ. Contam. Toxicol. 33(4).
Maslova, M.N. and L.V. Reznik. 1978. Suppression of cholinesterase activity
in rat brains with organophosphoric inhibitors with various hydrophobic
characteristics. Ukrainskiy biokhimicheskiy zhurnal. 48(4)
Melnikov, N.N. 1981. Principal Directions in the Search for New Pesticides.
Problems of Hygiene and Pesticide Toxicology. Papers of the VI All-Union
Scientific Conference, Kiev.
Melnikov, N.N., A.I. Volkov, and O.A. Korotkova. 1977. Pesticides and the
Environment. Moscow.
Murphy, S.U. 1966. Liver metabolism and toxicity of thiophosphate insecti-
cides in mammalian, avian and piscine species. Proc. Soc. Exper. Biol.
Med. 123.
Nepomnyashchikh, V.A. and G.M. Chuyko. 1986. , Determination of Effective
Dosages in Toxicologic Tests. USSR Academy of Sciences Institute of
Internal Water Biology. Borok.
O'Brien, R. 1984. Toxic Ethers of Phospheric Acid. Moscow.
Park, D.V. 1973. Biochemistry of Foreign Compounds. Moscow.
Perevoznikov, M.A. Ichthiocidal properties of "karbofos." Papers of
GosNIORKh, Leningrad, No. 146
Rozengart, V.I. and O.Ye. Sherstobitov. 1978. Selective Toxicity of Organo-
phosphoric Insecto-acaricides. Leningrad.
Svobodova, 2. 1980. Akutni toxicita pesticidu pro ryby. Agrochemia. 20(11).
Tinsley, I. 1982. Behavior of Contaminants in the Environment. Moscow.
Yakovlev, V.A. 1985. Kinetics of Fermentative Catalysis. Moscow.
Yang, C.F. and Y.P. Sun. 1977. Partition distribution of insecticides as a
critical factor affecting their rates of absorption from water and rela-
tive toxicities to fish. Arch. Environ. Contam. Toxicol. 6(3).
89
-------
RELATION OF TRACE METAL BODY BURDENS AND GILL DAMAGE IN FISH
TO SURFACE WATER ACIDIFICATION FROM ATMOSPHERIC DEPOSITION
by
T.A. Hainesl, C.H. Jagoe^, F.J. Dwyer^, and D.R. Buckler^
ABSTRACT
The possible relationship between trace metal body burden in fish and
various physical and chemical factors in lakes was examined. The effect of
acidity-related variables on body burdens in brook trout wasdetermined. Also
investigated were the effects of aluminum on gill structure of Atlantic salmon
under laboratory conditions to determine whether a link exists the form and
concentration of aluminum and damage to gill structures involved in ion
regulation.
INTRODUCTION
Atmospheric deposition is a major source of the strong acids and trace
metals reaching surface waters in the northeastern United States (Chan et
al. 1986, Summers et al. 1986). Deposition of the acids may also increase
the translocation of some metals, especially aluminum, from terrestrial to
aquatic systems (Campbell and Stokes 1985, Goyer et al. 1985). Such metals
may be directly toxic to fish or may be accumulated by fish to produce body
burbeds that are toxic to tertiary consumers including humans.
A number of investigators have reported that body burdens of trace
metals are higher in fish from acidic lakes than in those from neutral or
alkaline lakes (Dickson 1980, Fjerdingstad and Nilssen 1983, Hakanson
1980). Fish may accumulate metals directly from the external environment
through the body integument (especially through the gill), or by way of the
diet. Food chain uptake is especially important for metals occurring in
organic form, such as raethylmercury (Komarovskiy and Polishchuk 1981).
Uptake across body membranes depends on the chemical form of the metal.
lu.S. Fish and Wildlife Service, National Fisheries Contaminant Research
Center, Orono Field Station, Department of Zoology, University of Maine,
Orono, Maine, USA;
^Department of Zoology, University of Maine, Orono, Maine, USA;
-*U.S. Fish and Wildlife Service, National Fisheries Contaminant Research
Center, Route 1, Columbia, Missouri, USA.
90
-------
Free ions are apparently more readily transported across membranes than are
ions complexed to various ligands, and are generally more toxic (Campbell
and Stokes 1985; Hutchinson and Sprague 1987). In acidic lakes the metal
most toxic to fish appers to be aluminum; it is accumulated in the gills
(Brumbaugh and Kane 1985, Karlsson-Norrgren et al. 1986), where it seemingly
disrupts ion transport processes essential to osmoregulation (Leivestad et
al. 1987, Neville 1985).
In the present study we investigated the possible relation between fish
trace metal body burden and various lake physical and chemical factors, to
determiine whether acidity-related variables affect body burdens in brook
trout (Salvelinus fontinalis), a common species in lakes in the northeastern
United States. We further investigated the effects of aluminum on gill
structure of Atlantic salmon (Salmo salar) under labortory conditions, to
determine whether there is a link between the form and concentration of
aluminum and damage to gill structures involved in ion regulation.
MATERIALS AND METHODS
Six lakes were selected in Maine and New Hampshire along a southwest
to northeast deposition pH gradient. The lakes were similar in size and
elevation, and in having low specific conductance, but differed in acidity
(Table 1). Each lake was visited twice; water samples were collected during
both visits, and fish only during the second.
Water samples were collected with a plastic Van Dorn-type bottle at a
depth of 0.1 m and at 1 m above the bottom in the deepest area of the lake.
Table 1. Physical and chemical characteristics of lakes examined. Color,
specifc conductance, and pH values are means of four samples per lake
(two samples on two dates).
Lake
Area Max. Depth Elevation Color Sp. Cond, pH
(ha) (m) (m) (Hazen Unit) (pS/cnT )
East
Chair back
Ledge
Mountain
Coburn
Mountain
Rangeley
Solitude
Speck
18
3
2
17
2
4
18
6
3
12
6
8
466
911
871
733
729
999
• 0
48
20
15
33
22
18
17
19
23
20
18
5.18
5.43
6.63
5.80
4.95
5.08
91
-------
They were transferred to polyethylene bottles that had been acid-washed and
distilled water-rinsed, placed on ice, and returned to the laboratory for
analysis. One aliquot of each sample was preserved with nitricc acid and
stored for later determination of metals. The pH was measured with a digi-
tal meter (Orion model 611) equipped with a liquid-filled glass combination
electrode (Orion Ross). Specific conductance was measured with a tempera-
ture-compensating meter (Markson model 10). Apparent color was obtained by
visually comparing the color of a 25-ml unfiltered water sample to platinum
cobalt standards. We determined alkalinity by inflection point titration,
following the method of Gran (Strumm and Morgan 1981). All analyses were
performed in duplicate. Concentrations of elements were determined as
follows: Na, K, Mn, and Zn by air-acetylene flame atomic absorption spectro-
photometry (AAS; Perkin Elmer model 703); Ca and Mg by nitrous oxide-acety-
lene flame AAS; and Pb and total Al by graphite furnace AAS (Perkin Elmer
model HGA 2200).. Concentrations of 804, N03, and Cl were determined by ion
chromatography (Dionex model 2110i). Anion samples were analyzed within 24
hr after collection.
Brook trout were collected with experimental (graded mesh) nylon gill
nets fished overnight in the deepest area of each lake. After fish were
removed from the net, they were placed on ice and returned to the laboratory,
weighed (to the nearest gram), measured (total length, to the nearest milli-
meter), scale sampled for age estimation^ and frozen individually in plastic
bags. Whole 1-year-old fish were pooled in groups of three for homogeniza-
tion, digestion, and analysis. The fish were chopped into small sections
with a large knife and blended for 20 min in a polypropylene container fitted
with a Teflon blade assembly. All equipmnt was washed in the following
sequence between samples: detergent and tap water, 10% HC1, ultrapure water,
distilled acetone. About 100 g of homogenate was lyophilized and then dry-
blended for 10 min. Samples were stored in polyethylene bags in a desiccator
until digested. Duplicate 0.5 g subsamples of dry homogenate were digested
in 3 ml of 16 M HN03 in sealed pressure reaction tubes heated to- 70°C for 48
hr. Digestates were diluted to 50 ml with 1% HC1.
We determined aluminum, cadmium, lead, and manganese in fish by
graphite furnace AAS; selenium by graphite furnace AAS following hydrogen
selenide formation in an automated hydride generator; mercury by automated
cold vapor AAS; and copper and zinc by routine flame AAS. The quality
assurance program involved analysis of U.S. National Bureau of Standards
certified reference materials (tuna, oyster, pine needle, orchard leaves),
in-house reference materials, and spiked samples. The results were within
accepted limits for all metals. Fish trace metal concentrations were log
transformed to approximate a normal distribution. Analysis of variance and
Duncan's New Multiple Range Test were used to determine if trace metal
concentrations in fish differed among lakes, and stepwise multiple regression
was used to determine if lake physical or chemical factors were correlated
with trace metal concentrations.
Laboratory exposures were conducted with a flow-through proportional
diluter system supplied with reconstituted soft water as described by
Cleveland et al. (1986). Total alkalinity of the test water was 200 ueq/L
and calcium concntration about 3 mg/L. A 0.206 N mixture of sulfuric and
92
-------
nitric acids (2:1) was added with an automated pipetting system to maintain
the desired pH in the experimental aquaria. In the first year (1986) eggs
and fry of Atlantic salmon were continuously maintained without aluminum at
pH 7.2, 6.5, 6.0, 5.5, 5.0, and 4.5 in one diluter system. Simultaneously,
other groups were maintained in a second diluter system at pH 5.5 with
nominal aluminum concentrations of 38, 75, 150, and 300 ug/L. Aluminum was
added as aluminum sulfate, and pH was adjusted as described above. Controls
for this series were exposed to pH 5.5 with no aluminum, and unacidified test
water at pH 7.2. All aquaria were maintained at 8°C. Eyed Atlantic salmon
eggs were obtained from the Craig Brook (Maine) National Fish Hatchery. Eggs
and the resulting fry were held under test conditions until 60 days past the
median hatch date. Exposure procedures followed those described by Cleveland
et al. (1986). Beginning at swim-up, the fish were fed an Atlantic salmon
diet (ASD2-3 of the U.S. Fish and Wildlife Service) ad libitum three times
daily, supplemented twice daily with live nauplii of brine shrimp (Artemia).
Because the fish developed slowly at this low temperature, they did not swim
up and begin feeding until near the end of the study. Five fish from each
replicate were sampled for histological examination at 15, 30, and 60 days
after the median hatching date.
In the second year, a series of four sequential replicated experiments
were performed with post swim-up fry of Atlantic salmon to evaluate the
effects of dissolved organic acids on aluminum toxicity. Fish were exposed
to water at pH 5.7, without aluminum and with nominal aluminum concentra-
tions of 52, 86, 140, and 24000 ug/L. They were sampled after 6 days of
exposure. For the second experiment, 10 mg/L humic acid (Aldrich Chemical
Co.) was added to the stock water. Fish were exposed to water at pH 5.7,
without aluminum and with nominal aluminum concentrations of 780, 1300, 2160,
and 3600 ug/L. They were sampled after 6 days of exposure. For the third
experiment, 5 mg/L humic acid was added to the water, and fish were exposed
to pH 5.7 water with nominal aluminum concentrations of 0 or 780 ug/L. The
final experiment replicated the first (no humic acid added) to allow correc-
tion for the effects of increasing fish age on sensitivity for aluminum.
Observed mortalities during the first and fourth experiments did not differ
significantly.
Only live fish were selected for histological examination. Heads were
severed just posterior to the opercular coverings, and immediately fixed in
an ice cold mixture of 1% gluteraldehyde and 4% formaldehyde in 0.1 M phos-
phate buffer (pH 7.4), containing 10% sucrose (McDowell 1978). Tissues were
stored under refrigeration in this solution until used. Two or three indivi-
duals from each treatment at each sampling date were randomly selected for
examination by scanning electron microscopy (SEM).
Individual gill arches were dissected free of the branchial basket under
a dissecting microscope, rinsed in 0.1 M phosphate buffer (pH 7.4) with 10%
sucrose, and then postfixed in cold 1% osmium tetroxide in the same buffer
for 1 hr to increase specimen conductance. Tissues were dehydrated through a
graded ethanol series, and critical point dried under carbon dioxide. The
arches were then glued to stubs with silver paste, and sputter coated with a
gold palladium mixture to a nominal thickness of 250 angstroms. Specimens
were examined with an AMR 1000A scanning electron microscope operating at 5
KV, and photographed on Polaroid film.
93
-------
RESULTS AND DISCUSSION
The lakes ranged from 5.0 to 6.6 in pH and from -9 to 45 ueq/L in alka-
linity (Table 2). The major cation was calcium and the major anion was sul-
fate, in contrast to many remote lakes where the major anion is bicarbonate
(Wright and Henriksen 1983). The most likely source of the sulfate is atmos-
pheric deposition, which annually ranges from 15 to 22 kg/ha in these water-
sheds (National Atmospheric Deposition Program 1987). Alkalinity declines of
50 to 70 ueq/L in these lakes were estimated by the acidification model of
Wright and Henriksen (1983), suggesting that in the absence of acidic deposi-
tion, these lakes would have positive alkalinity and pH above 6. There was
no consistent relation between trace metal concentration in water and lake
acidity. We measured total rather than free ionic concentrations of metals,
which have been reported to increase at low pH (Borg 1983; Schindler et al.
1980).
Concentrations of aluminum, copper, mercury, manganese, and selenium
varied significantly among fish populations (Table 3). Stepwise multiple
regression of fish metal burdens on lacustrine physicochemical factors
indicated that chemical factors accounted for significant portions of the
variance in metal concentrations in fish (Table 4). The metals accumulated
by fish that were affected by lake acidity were mercury and manganese.
Accumulation of mercury in brook trout increased at lower pH, and lake pH
accounted for 71% of the variation in fish mercury concentration among lakes.
The addition of specific conductance (a measure of ionic strength) to the
regression increased to 83% of the proportion of variance accounted for. For
manganese, the most important variables were manganese in lake water, which
Table 2. Major ion and trace metal concentrations in lake water.
Values are means of four samples (two samples per lake on two
dates).
Lake
pH Alk Ca Mg 'Na K Cl SO, Al Mn Pb Zn
< Meq/1 > < yg/1 >
E. Chairback
Ledge
Mtn, Coburn
Mtn. Rangeley
Solitude
Speck
5.18
5.43
6.63
5.80
4.95
5.08
-6
4
45
13
-9
3
29
60
85
53
46
50
27
16
32
28
14
18
20
15
21
14
22
18
2
4
6
4
5
5
19
13
-
15
14
9
77
99
-
93
102
88
91
286
245
45
169
287
66
31
20
30
62
70
2
<1
1
<1
1
5
8
5
9
8
6
12
94
-------
Table 3. Mean concentrations, yg/g wet weight, of trace metals in age 1
fish. A sample consisted of three fish pooled. N is the number of
pooled samples analyzed. In each column, means with the same
superscript letter were not significantly different (p <0.05; one way
ANOVA, Duncan's New Multiple Range Test).
Lake
E. Chairback
Ledge
Mtn, Coburn
Mtn. Rangeley
Solitude
Speck
N
1
2
2
1
2
2
Al
3.25a'b
8,86b'C
2.22a
6.05a'b
14.73C
4.89a'b
Cu
0.543
1.62b
0.51a
0.47a
1.27b
0.48a
Metal
Hg
0.049a'b
0.046a'b
0.025b
0.060a'b
0.130C
0.085a
Mn
11. 2a
2.22C
2.88°
2.36°
7.34b
6.85b
Se
0.3203
0.155b
0.080b
0.160b
0.105b
0.205a'b
accounted for 74% of the variance in manganese in fish, and pH, which increased
the variance explained to 89%. For the remaining three metals, magnesium in
water was the most important variable for aluminum in fish, zinc in water for
copper in fish, and sulfate in water for selenium in fish. Water concentration
of aluminum was the second most important variable for both copper and selenium
in fish.
Acidity can influence trace metal uptake by fish through several
mechanisms. For mercury, the form most readily accumulated by fish is
methylmercury. The uptake is not increased by low pH (Rodgers et al. 1987),
but bacterial raethylathion of mercury in the water column is enhanced at
reduced pH (Xun et al. 1987). This increased methylathion would result in
increased concentrations of methylmercury in lake water at reduced pH and
could account for the increased concentration of mercury in fish from acidic
lakes. Increased concentrations of mercury have been reported in fish from
soft water lakes (Rodgers and Beamish 1983; Scheider et al. 1979), which may
account for the importance of specific conductance in the regression model.
Ionic strength of water may affect fish mercury concentration because reduced
concentration of ions in water, especially divalent cations, increases the
permeability of gill membranes (Franzin and McFarlane 1980; McFarlane and
Franzin 1980; Part et al. 1985). Certain metals, including manganese, become
more soluble as Ph decreases, and the free ionic species predominates at pH
<6 (Campbell and Stokes 1985). These factors would make manganese more
available for uptake across the gill surface by fish inhabiting acidic lakes.
Magnesium concentration in water probably affects trace metal content
of fish by increasing membrane permeability, as discussed above. Winner and
95
-------
Table 4. Results of stepwise multiple regression analysis of fish tissue
trace metal concentrations as dependent variables and lake physical and
chemical factors as independent variables. The best one and two
variable models are given for metals that varied significantly among
lakes.
Metal
Al
Cu
Hg
Mn
Se
Number of
Variables
1
2
1
2
1
2
1
2
1
2
3.
4.
1.
1.
0.
-2.
-1.
Model
18 -
60 -
24 -
85 -
78 -
28 -
05 +
-4.89 +
7.
9.
66 -
89 -
1
2
1
1
0
0
1
2
4
5
.85
.15
.55
.23
.37
.39
.05
.07
.31
,06
log
log
log
log
pH
Mg
Mg - 0.46 log Al
Zn
Zn - 0.68 log Mg
pH + 2.50 log cond.
log
log
log
log
Mn
Mn + 0.40 pH
SO
SO^ - 0.58 log Mg
r
0,
0.
0.
0.
0.
0.
0.
0.
0.
0.
2
74
87
81
94
71
83
74
89
63
77
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
P
0015
0002
0004
0001
0022
0019
0013
0005
0061
0056
Gauss (1986) found that increased concentration of calcium or magnesium
reduced the biaccumulation of copper, cadmium, and zinc by Daphnia sp. Part
et al. (1985) reported that calcium and magnesium in water reduced the trans-
fer of cadmium from water into perfused fish gills, probably by changing
the permeability of the gill epithelium. Conversely, water hardness and
alkalinity had no effect on the uptake of copper by rainbow trout, salmo
gairdneri (Lauren and McDonald 1986). The effect of zinc or sulfate
concentration in water on trace metal uptake by fish is unknown.
Gills of fish reared under control conditions appeared normal, as
described by Laurent and Dunel (1980), and free of lesions when examined by
SEM (Figure 1A)• Chloride cells were present in both primary and secondary
lamellar epithelium and were distinguished by a denser surface pattern of
shorter microridges and microvilli (Figure IB). Exposure to low pH alone
for up to 60 days post-hatch resulted in little change in gill morphology.
At pH 4.5, the lowest pH tested, there was some apparent swelling of epithe-
lial cells along the primary lamellae and increased amounts of mucus pro-
duced (Figure 2A). Exposure to aluminum concentrations of 2.75 ug/L in the
absence of humic accid produced serious alterations in gill structure. At 75
ug/L, secondary lamellae did not develop normally, and large areas of the
filament were devoid of them (Figure 2B). These present were concentrated at
the distal end of the filaments, and appeared small and stubby. Exposure to
96
-------
Figure 1. Scanning electron micrographs of normal gills of post-swim up
Atlantic salmon fry exposed to pH 7.2 water without aluminum or humic
acid. A: Gill filaments or primary lamellae extending from gill arch.
Bar = 100 microns. B: Chloride cell on primary lamellar epithelium.
Bar = 10 microns.
150 ug/L caused epithelial proliferation, especially of chloride cells,
leading to fusion of adjacent primary lamellae (Figure 2C). Exposure to 300
ug/L caused a similar, but more severe effect (Figure 2D). At the highest
concentration, individual filaments were barely distinguishable.
Alterations in morphology similar to those produced in fry during the
1986 experiments were also seen in fish from the 1987 experiments that were
exposed to 240 ug/L aluminum in the absence of humic acid. Individual gill
filaments were recognizable, but many sections were devoid of secondary
lamellae (Figure 3A). Apical crypts were observed in chloride cells. Gill
structure was normal in fish exposed to humic acid in the absence of aluminum
(Figure 3B). In water containing 10 mg/L humic acid exposure to total alutni-
97
-------
Figure 2. Scanning electron micrographs of gills of 60 day old Atlantic
salmon fry exposed to low pH and aluminum without humic acid (all bars
s 100 microns). A: pH 4.5, no aluminum. B: pH 5.5, 75 ug/L aluminum.
C: pH 5.5, 150 ug/L aluminum. D: pH 5.5, 300 ug/L aluminum.
num concentrations as high as 2160 ug/L—-nearly 10 times more aluminum than
was necessary to produce striking morphological alterations in the absence of
humic acid—produced no visable morphological effects (Figure 3C). Exposure
to 3600 ug/L total aluminum in water containing 10 mg/L humic acid produced
lesions identical to those produced by lower aluminum levels in the absence
of humic acid (Figure 3D)—loss of secondary lamellae, epithelial hyper-
plasia, and development of apical crypts on chloride cells.
At lower concentrations of humic acid (5 mg/L), lower concentrations of
aluminum caused morphological effects. Exposure to 78 ug/L total aluminum
produced slight but noticeable alterations (Figure 3E). Secondary lamellae
98
-------
Figure 3. Scanning electron micrographs of gills of post-swim up Atlantic
salmon fry exposed to low pH and aluminum with and without humic acid
(all scale bars = 100 microns). A:' pH 5.7, 240 ug/L aluminum, no humic
, acid. B: pH 5-.7, no aluminum, 10 mg/L humic acid. C: pH 5.7, 2160
ug/L aluminum, 10 mg/L humic acid. D: pH 5.7, 3600 ug/L aluminum, 10
mg/L humic acid. E: pH 5.7, 780 ug/L aluminum, 5 mg/L humic acid.
99
-------
were fused at the distal ends of the primary amellae, giving the ends of the
filaments a clubbed look.
Apical crypts of chloride cells, normally present after seawater
acclimation, were observed in fish exposed to pH 4.5 in the absence of alumi-
num (Figure 4A). A similar,effect was previously reported by Leino and
McCormick (1984). Exposure to aluminum at concentrations bove about 150
ug/L at pH 5.5 also caused the development of apical crypts in the absence
of humic acid. Pores of larger diameter were also observed in these aluminum-
exposed fish (Figure 4B). In cross section, these were found to be lined
with chloride cells (Jagoe et al. 1987). Humic acid (10 mg/L) prevented the
development of apical crypts at aluminum concentrations upp to 2160 ug/L
(Figure 4C), but apical crypts were numerous in fish exposed to 3600 ug/L
aluminum (Figure 4D). Some apical crypts were also observed after exposure
to 780 ug/L aluminum in water containing 5 mg/L humic acid.
It seems likely that only free ionic hydroxy aluminum produces the
morphological abnormalities. At high humic acid and low aluminum levels,
virtually all of the aluminum is probably chelated to,organic ligands;
consequently the hydroxy aluminum fraction is low and damage is prevented.
As the aluminum content of the water increases, the humic acid binding
sites become saturated and abnormalities again appear as the hydroxy aluminum
concentration increases. The observation that gill damage begins to appear
at higher total aluminum levels with increasing humic acid concentration is
consistent with this idea. Our attempts to confirm this hypothesis by
analytically fractionating aluminum into chelated and ionic forms failed,
possibly because of the high concentrations of organic acids or aluminum, or
both.
The relatively greater morphological alteration observed in the younger
Atlantic salmon (1986 study) supports the belief that young fish are the less
dependant on gill respiration (El-Fiky et al. 1987). A similar degree of
gill abnormality would kill an older fish that depended on gill respiration,
whereas younger fish can survive, at least until they come to rely more on
gill respiration. Chloride cell alterations observed in this study are
consistent with increased (or attempts to increase) ionic transport activity
in- the gill. It appears that some aluminum finds its way into chloride cells
(C. H. Jagoe, unpublished data; Youson and Neville 1987), which may affect
chloride cell function to some degree. If each chloride cell is functioning
less well, the only possible strategy open to the individual fish may be to
form more cells, leading to epithelial hyperplasia and the observed abnormal-
ities.
LITERATURE CITED
Borg, H. 1983. Trace metals in Swedish natural fresh waters. Hydrobiologia
101: 27-34.
Brumbaugh, W., and D. Kane. 1985. Variability of aluminum concentrations
in organs and whole bodies of smallmouth bass (tlicropterus dolomieui).
Environ. Sci. Technol. 19: 828-831.
100
-------
Figure 4. Scalling electron micrographs of gill epithelial surface of
Atlantic salmon fry exposed to various levels of pH, aluminum and humic
acid. A: pH 4.5, no aluminum, no humic acid (br = 10 microns). B: pH
5.5, 150 ug/L aluminum, no humic acid (bar = 10 microns). C: pH 5.7,
2160 ug/L aluminum, 10 mg/L humic acid (bar = 100 microns). D: pH 5.7,
3600 ug/L aluminum, 10 mg/L humic acid (bar = 10 microns).
Campbell, P., and P. Stokes. 1985. Acidification and toxicity of metals to
aquatic biota. Can. J. Fish. Aquat. Sci. 42: 2034-2049.
Chan, W., A. Tang, D. Chung, and M. Lusis. 1986. Concentration and
deposition of trace metals in Ontario-1982. Water Air Soil Pollut.
29: 373-389.
Cleveland, L., E. Little, S. Hamilton, D. Buckler, and J. Hunn. 1986.
Interactive toxicity of aluminum and acidity to early life stages of
brook trout. Trans. Am. Fish. Soc. 115: 610-620.
101
-------
Dickson, W. 1980. Properties of acidified waters. Pages 75-83 in D.
Drablos and A. Tollan, editors. Ecological impact of acid precipita-
tion. Acid precipitation - Effects on Forest and Fish Project, Aas,
Norway.
El-Fiky, N., S. Hinterleitner, and W. Wieser. 1987. Differentiation of
swimming muscles and gills, and development of anaerobic power in the
larvae of cyprinid fish (Pisces, Teleostei). Zoomorphology 107:
126-132.
Fjerdingstad, E., and J. Nilssen. 1983. Heavy metal distribution in
Norwegian acidic lakes: a preliminary record. Arch. Hydrobiol.
96: 190-204.
Franzin, W., and G. McFarlane. 1980. Fallout, distribution and some
effects of Zn, Cd, Cu, Pb and As in aquatic ecosystems near a base
metal smelter on Canada's Precambrian Shield, Pages 32-303 in D.
Drablos and A. Tollan, editors. Ecological impact of acid
precipitation. Acid precipitation - Effects on Forest and Fish
Project, Aas, Norway.
Goyer, R., J. Backmann, T. Clarkson, B. Ferris, J. Gram, P. Mushak, D. Perl,
D. Rail, R. Schlesinger, W. Sharpe, and J. Wood. 1985. Potential
human health effects of acid rain: report of a workshop. Environ.
Health. Perspec. 60: 355-368.
Hakanson, L. 1980. The quantitative impact of pH, bioproduction and
Hg-contamination on the Hg-content of fish (pike). Environ. Pollut.(B)
1: 285-304.
Hutchinson, N., and J. Sprague. 1987. Reduced lethality of Al, Zn and Cu
mixtures to American flagfish by complexation with humic substances in
acidified soft waters. Environ. Toxicol. CChem. 6: 755-765.
Jagoe, C., T. Haines, and D. Buckler. 1987. Abnormal gill development in
Atlantic slamon (Salmo salar) fry exposed to aluminum at low pH.
Annls. Soc. R. Zool. Belg. H7(Suppl. 1): 375-386.
Karlsson-Norrgren, L., W. Dickson, 0. Ljungberg, and P. Runn. 1986. Acid
water and aluminum exposure: gill lesions and aluminum acccumulation in
farmed brown trout Salmo trutta L. J. Fish Dis. 9: 1-9.
Komarovskiy, F., and L. Polishchuk. 1981. Mercury and other heavy metals
in the water: migration, accumulation and toxicity to aquatic organisms
(a review). Hydrobiol. J. 17: 51-62.
Lauren, D., D. McDonald. 1985. Effects of copper on branchial ionoregula-
tion in the rainbow trout, (Salmo gairdneri) Richardson. J. Comp.
Physiol. B 155: 635-644.
Laurent, P., and S. Dunel. 1980. Morphology of gill epithelia in fish.
Am. J. Physiol. 238: R147-R159.
102
-------
Leino, R., annd J. McCormick. 1984. Morphological and morphometrical
changes in chloride cells of the gills of Pimephales promelas after
chronic exposure to acid water. Cell Tissue Res. 236: 121-128.
Leivestad, H., E. Jensen, H. Kjartansson, and L. Xingfu. 1987. Aqueous
speciation of aluminum and toxic effects on Atlantic salmon. Annls.
Soc. R. Zool. Belg. 117 (Suppl. 1): 387-398.
McDowell, E. 1978. Fixation and processing. Pages 113-139 in_: B. Trump
and R. Jones, editors. Diagnostic Electron Microscopy, Vol. 1. J.
Wiley and Sons, New York, NY.
McFarlane, G., and W. Franzin. 1980. An examination of Cd, Cu, and Hg
concentrations in livers of northern pike, Esox lucius, and white
sucker, Catostomus commersoni, from five lakes near a base metal
smelter at Flin Flon, Manitoba. Can. J. Fish. Aquat. Sci. 67:
1573-1578.
National Atmospheric Deposition Program. 1987. NADP/NTN Annual Data
Summary: Precipitation Chemistry in the United States. 1986. Natural
Resource Ecology Laboratory, Colorado State University, Fort Collins,
CO. 363 pp. '
Neville, C. 1985. Physiological response of juvenile rainbow trout, Salmo
gairdneri. to acid and aluminum - Prediction of field responses from
laboratory data. Can. J. Fish. Aquat. Sci. 42: 204-2019.
Part, P., 0. Svanberg, and A. Kiessling. 1985. The availability of cadmium
to perfused rainbow trout gills in different water qualities. Water
Res. 19: 427-434.
Rodgers, D., and F. Beamish. 1983. Water quality modifies uptake of
waterborne methylmercury by rainbow trout, Salmo gairdneri. Can. J.
Fish. Aquat. Sci. 40: 824-828. ~
Rodgers, D., T. Watson, J. Langan, and T. Wheaton. 1987. Effects of pH and
feeding regime on methylmercury accumulation within aquatic microcosms.
Environ. Pollut. 45: 261-274.
Scheider, W., D. Jeffries, and P. Dillon. 1979. Effects of acidic
precipitation on precambrian freshwaters in southern Ontario. J. Great
Lakes Res. 5: 45-51.
Schindler, D., R. Hesslein, R. Wagemann, and W. Broecker. 1980. Effects of
acidification on mobilization of heavy metals and radionuclides from
the sediments of a freshwater lake. Can. J. Fish. Aquat. Sci. 37:
373-377.
Stumm, W., and J. Morgan. 1981. Aquatic chemistry. 2nd ed. J. Wiley and
Sons, New York, NY.
103
-------
Summers, P., V. Bowersox, and G. Stensland. 1986. The geographical
distribution and temporal variations of aquatic deposition in eastern
North America. Water Air Soil Pollut. 31: 523-535.
Winner, R., and J. Gauss. 1986. Relationship between chronic toxicity and
bioaccumulation of copper, cadmium and zinc as affected by water
hardness and humic acid. Aquat. Toxicol. 8: 149-161.
Xun, L., N. Campbell, and J. Rudd. 1987. Measurements of specific rates of
net methyl mercury production in the water column and surface sediments
of acidified and circuraneutral lakes. Can. J. Fish. Aquat. Sci. 44:
750-757.
Youson, J., and C. Neville. 1987. Deposition of aluminum in the gill
epithelium of rainbow trout (Salmo gairdneri Richardson) subjected to
sublethal concentrations of the metal. Can. J. Zool. 65: 647-656.
104
-------
BUFFER CAPACITY OF FRESHWATER ECOSYSTEMS FOR HEAVY METALS
AND HYDROBIOLOGICAL PARAMETERS DETERMINING IT
by
A.M. Nikanorov1, A.V. Zhulidov1, N.A. Dubova1,
•V.F. Gekov1, I.Y. Kamov1
ABSTRACT
Relationships between seasonal and yearly dynamics of phytoplankton bio-
mass with the background concentration of copper in the river Usman are ex-
amined. The investigation seeks to directly estimate the degree and charac-
ter of hydrobiota influence on the chemical composition of freshwater systems
because developments in this area are of special importance for understanding
the buffer capacity of freshwater systems for heavy metals.
ECOSYSTEM BUFFER CAPACITY
The problem of determination of freshwater
which is the parameter allowing one to estimate
ing an aquatic ecosystem that does not seriously
of functioning of the whole ecosystem, is one of
modern water chemistry and ecology (Izrael et al
Izrael 1984, Konemann 1984). (This formulation
includes elements of uncertainty in relation to
"func tioning.")
ecosystem buffer capacity,
the dose of a pollutant enter-
disturb the natural mode
the most urgent problems in
1985, Parker et al. 1982,
is preliminary because it
the options "seriously" and
Based on theoretical works (Ghilarov and Timonin 1972, Breymeyer 1979,
Braginsky 1981, Trojan 1979 and 1984, Ravera 1984, Kimmerer 1984, Brown 1984,
Kolasa and Biesiadka 1984, Remmert 1984, Barbauli 1985, Rygg 1985, Stenseth
1985), one may assume that freshwater ecosystem buffer capacity in the above
context is related to stability (homeostatis) of these ecosystems and, in
particular, to species diversity, trophic levels and size structure of popu-
lations. In practice however, only species composition of some communities
or trophic levels but not ecosystems on the whole is usually known rather
well. Within the limits of individual trophic levels, stability is probably
determined by interspecific competition (Murdoch et al. 1984, Thorp and
Gothran 1984, Reice 1985). Thus, different trophic levels, as communities,
l-Hydrochemical Institute, Rostov-on-Don, USSR.
105
-------
may reach equilibrium with habitat independently of each other (Breymayer
1979, Matsunaga et al. 1984).
It also turns out that there is a positive regression between freshwater
ecosystem area and species diversity of hydrobiota and the structure of mac-
rophyte cenosis (Bronmark 1985). Besides, in a number of cases when determin-
ing the structure and stability of communities, the leading species or complex
o± abundant species (Burkovsky and Azovsky 1985) included in the composition
of biogeocenosic nucleus are very important. Variation in species composition
of ecosystems mostly occurs unevenly, changing trends of many important bio-
logical processes (Futuyma 1973, Zhirmunsky et al. 1981).
Other difficulties involve the necessity of analysis of the processes
influencing population dynamics and processes of interaction between community
components (competition, predation, etc.), the need for information about the
ecosystem informative structure, the difficulties in determination of tax-
onomic affiliation of a number of groups of invertebrates and lower plants,
and also the poor knowledge of homeostasis mechanisms' (and even individual
organisms) comparing to the most important pollutants (Trojan 1979 and 1984,
Nikanorov et ai. 1982, Peterfi et al. 1983, Smalls and Cannon 1983, Nikanorov
and Zhulidov 1984, Kimmerer 1984, Smith and Kallf 1985, Maksimov et al. 1985,
Barbault 1985). The above problems do not allow one to quantify buffer
capacity through species diversity even in general now, although we think
that the prospects of such an approach are rather good. Wood (1983) showed,
however, that the influence of metals upon phytoplankton is in close correla-
tion with parameters characterizing complexation ability and does not depend
on the composition and structure of a community.
It does not seem possible at present to consider the option "buffer capac-
ity" and to compare it with the stability concept mainly because of the termi-
nological uncertainty of this term (Futuyma 1973, Feodorov 1975, Solntzeva
1982, Nazaros 1983, Sergeev 1984). In conformation with dynamic systems in
general, stability may be defined as an ability of a moving system to deviate
slightly from some track at perturbations. If we consider perturbation as
variation in primary condition of the system and deviation as a measure for
differentiating between perturbed motion (functioning) and unperturbed motion,
we come to the option of uniform and orbital stability according to Lyapunov.
If the tracks still possess some properties of unperturbed tracks, we come to
stability according to either Lagranzh or Puasson. Thus, in an attempt to
formulate the option "stability," it is necessary to understand which type of
non-sensitivity is of greatest interest or which responses to perturbation
are equivalent (or not equivalent).
The option "stability" developed in physico—mathematical fields of
natural science may be used also for ecological system (Kindlmann and Leps
1985). In this case, however, one should keep in mind that, in the selection
of parameters and variables on the basis of which the concept of stability
is considered, the subsystem of a certain ecosystem must be relatively auton-
omic so that we can neglect the effect of the variables not used. Besides,
time and space scales within the limit of which the concept of stability is
considered must be logically correlated with the selected type of equivalence
of ecosystem responses on perturbation. The option "ecosystem stability"
106
-------
makes no sense if these facts are not taken Into account. It should be
specially indicated that arbitrary assumptions result from the fact that the
concept of stability turns out to be inapplicable to real ecosystems due to
the following reasons (Kindlmann and Leps 1985, Nunney 1985).
- It is assumed that the system in normal conditions is in equilibrium
which is not true.
- System stability without information on the size of the field is in-
vestigated.
- Perturbations of permanent character in ecosystrems are not taken into
account.
Owing to the above reasons, the option of stability in ecology and geo-
graphy is of non-formalized character (Futuyma 1973, Armand 1983, Preobrazhen-
sky 1983, Svetlosanov 1983, Ulrich 1984, Gurtz and Wallace 1984, Noon et al.
1985). We must keep in mind that heavy metals (for instance, lead in concen-
trations up to 64 ppb when its background concentration in water is 0.1-0.6
ppb) stimulates reproduction of some invertebrates (Berglind et al. 1985).
Taking into account the above, the authors attempted to estimate buffer
capacity through the interaction of the most important parameters determining
behavior of pollutants (heavy metals, in particular) in freshwater ecosystems
(Izrael et al. 1985). Parameters conditionally attributed to hydrobiological
factors are discussed below.
Laboratory investigations indicated that hydrobiota (shell-fish, in par-
ticular) affect the forms of occurrence of extremely toxic heavy metals in
natural waters (Nikanorov et al. 1985). Because laboratory results can be
transferred to the processes occurring within natural ecosystems to quite a
limited degree, however, continuous natural investigations at the river Usman
in the Voronezh Biosphere Reserve were carried out. The reserve was chosen
for long-term investigations because of the specificity of the problems of
investigation carried out. Besides, Voronezh Reserve is located in a natural
zone that is convenient for such investigations—forest-steppe characterized
by optimal conditions for development of all biological components of an eco-
system. It is also important that there are laboratories in the Reserve pro-
viding short-term initial processing of the information. The area scheme,
hydrographic network, and water regime of the river Usman are presented in
the work of Nikanorov et al. (1985).
Because all the trophic groups of hydrobiota cannot be covered by long-
term field investigations, it was decided to bend every effort to phytoplank-
ton producers, often determining a whole number of the most important hydro-
chemical and hydrobiological parameters. Copper was used for this complex of
investigations. Selection of copper for modeling purposes was related to the
high biological activity of the element and its distinct ability to form com-
plexes with dissolved organic matter (Brown and Rattigan 1979, Sposito 1981,
Giesy 1983, Giesy et al. 1983, Borgmann and Ralph 1983 and 1984, Reddy and
Rao 1983, Cairns et al. 1984, Winner 1984 and 1985, Borgmann and Charlton
1984, Rygg 1985, Blaylock et al. 1985).
107
-------
The object of our investigations was to study relationships between sea-
sonal and yearly dynamics of phytopiankton biomass in model biological cur-
rent of the Usman with the background concentration of dissolved copper in
its water.
The information on seasonal and yearly dynamics of phytoplankton biomass
and dissolved copper concentration in the Usman is presented in Figures 1 and
2 and Table 1.
The information presented (in Figures 1 and 2) proves that seasonal dy-
namics of phytoplankton biomass in the Usman may reflect the following regu-
larity—growth of biomass from winter to summer (autumn) and decrease by
winter. Seasonal dynamics of dissolved copper concentration in the Usman
is characterized, however, by an inverse relationship: copper concentration
in water is maximum in winter and minimum in summer (autumn).
Quantification of the relationship between dissolved copper concentra-
tion and phytoplankton biomass is realized by processing of the results of
natural investigations with mathematical statistics. To adequately describe
the phenomenon studied, however, it is necessary to determine the form of
the relationship between the variables studied. The form ot the relationship
to
600
520
440
g 360
E
S 280
.1
•c 200
_D
QL
Q-
120
40
636
621
0.9 ^
0.7 "Q.
Q.
0.5
re*
O.3
0.1
CD
O.
Q.
o
T3
>
O
to
Q
1977 1978 1979 1980
Year
1981
1983 1984
Figure 1. Monthly average of many-year variability of phytoplankton biomass
and dissolved copper in the Usman River in the Voronezh Biosphere Reserve.
108
-------
ro
600
520
44°
36°
I
bo
o 280
c:
o
200
CL 120
40
0.9
0.7
CL
ex
O)
CL
CL
0.5 8
0.3
0.1
d)
_>
O
(/>
tn
Q
1 3 579 n
Month
Figure 2. Confidence interval of 95% precision for many-
year average of phytoplankton biomass and dissolved
copper concentration in the Usman water (1977-1981,
1983, 1984).
depends on the character of the process studied and is determined by investi-
gation of physico-chemical and biological regularities. Then, parameters of
the constraint equation are quantified on the basis of statistical processing
of field results.
We may assume that copper (M) entering a stream consists of five com-
ponents:
- MB - is copper left in water in dissolved form
- Mfc - is copper sorbed and accumulated by phytoplankton (by all
hydrobiota in more general case)
- MA - is copper that moved to bottom sediments in the process of
sedimentation of suspended organic and mineral material
~ MaA ~ is dissolved copper sorbed at the surface of bottom sedi-
ments directly from water
109
-------
TABLE 1. AVERAGE CHANGE OF DISSOLVED COPPER CONTENT AND PHYTOPLANKTON
BIOMASS IN THE USMAN, 1977-1981 AND 1983-1984
Month
January
February
March
April
May
June
July
Augus t
September
October
November
December
Dissolved copper,
ppb/L
0.637+0.12
0.636+0.107
0.593+0.129
0.536+0.056
0.386+0.098
0.319+0.079
0.332+0.084
0.316+0.007
0.307+0.091
0.401+0.123
0.545+0.106
0.706+0.058
Phytoplankton biomass,
mg/L3
139.8+59.7
143.7+55.1
184.8+96.3
238.5+87.3
357.4+117.7
371.1+91.2
471.9+77
504.4+79.1
472.9+82.4
341.7+103.4
252.9+114.4
144.0+58.2
- is copper adsorbed by suspended solids and contained in them.
In this case, the many-year relationship between dissolved copper con-
tent and phytoplankton biomass is considered. So, the processes determining
copper behavior in water may be neglected.
On the whole, the copper material balance equation may be written:
M = MB + M^ + M^ + M6 + MA (D
As for quantity of copper adsorbed by bottom sediments, it is proportion-
al to metal concentration in river water - CMg.
110
-------
or
M
aA
v
MaA =a'GMB = a
a
M-
v
v
(2)
so we finally obtain:
M
aA
a M
B
(3)
As a first approximation, the quantity of metal absorbed at suspension is
proportional to suspension quantity and metal concentration in water. Taking
into account physico-geographic and geological features of the Voronezh Re-
serve area and specific features of the Usman, however, suspension content
in river water may be taken as a value slightly varying during a year. (Al-
though in the flood period, suspension content increases.) Equality deter-
mining quantity of copper absorbed at suspended solids in unit water volume
is written as:
M
aB
or
MB
(4)
where y is a proportionality factor.
Hydrobiota accumulation ability in relation to dissolved metals is usu-
ally expressed by the accumulation factor, K^, (Nikanorov et al. 1985). The
accumulation factor is assumed to be the ratio between the metal concentration
in the tissue of species studied (per unit dry weight) and the content of the
same metal in similar water volume taken from the hydrobiota habitat. Thus,
copper content in photoplankton with regard for accumulation factors may be
written as:
(5)
where <1 is phytoplankton mass in unit water volume.
The quantity of copper moved to bottom sediments with dead phytoplank-
ton is estimated as:
MA = KHMBliC6
(6)
where £ is the mortality factor.
Ill
-------
With regard for Equations 3 through 6, Expression (1) may be written as:
or
(7)
M=MB[I+u
Using copper concentration per unit water volume and introducing the
terms:
M MB
% - - ; cMB = — s a " I+U+Y; » s KH(I+IJ) (8)
V V
we get the equation relating dissolved copper concentration to phytoplankton
blomass in a unit water volume:
CMB
cM
a+B • C/r
(9)
or for convenience of statistic processing dividing
I
CMB -
(10)
A+BC,
-------
Thus, we have
CMB
0.94+0.00477 Ct
(12)
The correlation factor between many-year average values based on the months
is 0.91.
Relationship 12 is shown in Figure 3. The off-side point (shown as a
cross in Figure 3) corresponds to April, the period of the Usman ice cover
break and flood when the content of suspension in the river and, hence, the
content of copper sorbed on this suspension is increasing. Because the por-
tion of copper sorbed on suspension may be considerable, this stimulates dis-
turbance of Relationship 10 obtained if the inorganic suspension concentra-
tion during a year is uniform.
We must note that, in case you are looking for Relationship 11 based on
direct measurement of C^g^/^ values in the samples taken on the same day
(total number of such pairs is 107), the regression relation of these parame-
ters is poor (correlation factor approaches 0.2). This result reflects the
general fact of relationship distortion with inaccurately present regressors
(Davies and Hutton 1975). In our case, the interferring factors in the pro-
cess of determination of interrelationships between processes are consider-
able non-uniformity of dissolved copper distribution and, in particular,
phytoplankton in space and time and, hence, poor representativeness of some
samples.
0.8
.a
8:0.6
Q.
8-0.4
O
T3
CO
O 0.2
if)
0 100 300 500
700
Phytoplankton Biomass,
mg/m3
Figure 3. Many-year-average relationship
between dissolved copper concentration
and phytoplankton biomass in the Usman
River (Voronezh Biosphere Reserve).
113
-------
Mainly due to the impact of these factors, variability of dissolved cop-
per concentration and phytoplankton biomass in the Usman water samples taken
simultaneously reaches 4-fold (December 29, 1980) and 41-fold (August 29,
1984), correspondingly, and it may be even more in the samples taken at in-.
tervals of several hours or days. Thus, averaging is levelling the natural
variability of the studied processes considerably. Averaging is conducted in
such a manner that the efficient noise suppression would be combined with
insignificant distortion of basic processes. Taking into account the above
said, it is clear why the correlation is found between alga biomass and con-
centration of chemicals in the environment and, in particular, nutrients
(Jones et al. 1984, Misztal et al. 1984).
Based on the data presented in the monograph of A.M. Nikanorov et al.
(1985), one may judge the accumulation ability of hydrobiota in relation to
chemical elements. Because of the methodological problems related to sampling
and analysis of ultrasmall samples of phytoplankton, it was impossible to study
specific peculiarities of the factors of chemical element accumulation by
phytoplankton in the Usman. There are no reasons, however, to assume that
accumulation ability of plankton micro-algae would be different from that of
freshwater invertebrates.
Kinetics of metal absorption usually consists of two stages: the light
one when fixation of ions on the surface of cells occurs with high pH during
photosynthesis and the dark one characterized by ion transport to the cells
with low pH due to increase of local carbonic acid concentration with respira-
tion (Morel 1985). In this case, the rate of metal absorption by phytoplank-
ton intensity of metal concentration decreases in water. Increase of metal
concentration in bottom sediments and alga biomass growth are interrelated
(Jackson 1987, Menarques and Lanza 1983, Hellmann 1983, Hamilton-Taylor et
al. 1984, Falkner et al. 1984, Morel 1985, Izrael et al. 1985), fluctuating
within 24 hours.
Metal concentration in phytoplankton cells depends not only on the con-
centration of metals in natural environment (Parker et al. 1982, Menargues and
Lanza 1983, Jones et al. 1984, Les and Walker 1984, Vymazal 1984), but on the
luminance, pH at the interface membrane-medium, balance constants, rate of
diffusion (accumulation) through cell membranes, physiological condition, arid
taxonomic affiliation of algae (Briand et al. 1978, Stary et alo 1983, Falkner
et al. 1984, Skowronski 1984, Flatau et al. 1984, Goreonova et al. 1984).
Laboratory investigation showed that the rate of phosphorus consumption by
phytoplankton depends on the size of the cells but not on the taxonomic affil-
iation (Smith and Kallf 1985). This affiliation, however, is disputed (Sommer
and Kilham 1985) and so needs additional investigation.
Taking into account the above, we understand daily variability, intensi-
ty of metal accumulation by phytoplankton and, hence, daily variability in
the concentration of dissolved metals in water. The last is exposed to seri-
ous influence of hydrophysical features of water bodies and streams, nonuni-
formity of microecological conditions within the water column (both vertical
and horizontal), errors due to sampling and analysis of dissolved forms of
metals, and other factors. All these parameters automatically determine
daily variability of phytoplankton biomass distribution as well, Summariz-
114
-------
ing the above, we may conclude that the content of chemical elements in
freshwater ecosystems is directly related to the general value of hydrobiota
biomass and its seasonal dynamics within these ecosystem (Figures 3 and 4).
The relationship between variation in biomass and season of the year for
phytoplankton is also true for zooplankton and periphyton (Dreu 1976, Pork
and Lokk 1979, Boynton et al. 1983, Izrael et al. 1985) and probably may be
considered as general for all hydrobiota.
Speaking of seasonal variability of hydrobiota biomass in freshwater eco-
systems and its role in water chemical composition formation in freshwater
ecosystems, we note that some representatives of hydrobiota, and, in particu-
lar, bivalve mollusks, possess similar geochemical function owing to consider-
ably varying filtration in different seasons. The most active filtration and
intensity of feeding is registered in summer months but in autumn, winter,
and early spring filtration rate decreases sharply up to full stop (Alimov
1981). Filtration intensity of bivalve mollusks increases with their weight
and size (Alimov 1981).
Discussing the problem of hydrobiota influence upon freshwater ecosystem
chemical composition, we note that the degree of this influence is related to
the intensity of metabolism. Hydrobiota metabolic rate is determined, however,
not only by seasonal variations in ambient temperature but by the peculiarities
of their life cycles and, in particular, the rate of physiological and repro-
duction activikty (Alimov 1981).
Speaking of hydrobiota metabolic rate, we must note that numerous inves-
tigations proved close relationships between the rate and intensity of ex-
change and the size of freshwater organisms. In most cases, the metabolic
rate in animals increases but its intensity decreases if their weight is in-
creasing. Summarizing investigations on this problem and quantification of
relationship between metabolic rate and size of the freshwater bivalve mol-
lusks were carried out by A.F. Alimov. The results of measurement of the
metabolic rate of bivalve mollusks indicate that these animals do not possess
either generic or specific character of exchange and metabolic rate is a func-
tion of their weight. We may assume that there are no significant differences
in metabolic rate (Alimov 1981) between the animals of the same size that
belong to different families Bivalvia. This phenomenon is not analyzed from
the point of view of deep understanding of hydrobiota role (in particular,
Bivalvia) in formation of freshwater ecosystem chemical composition. Taking
into account the role of animals in ecosystem functioning (in particular,
earth biogeocenoses), we may be more sure of the importance of the above
phenomenon for the processes of formation of water chemical composition in
water bodies and streams.
The investigation conducted is actually an attempt to directly estimate
the degree and character of hydrobiota influence on the chemical composition
of freshwater ecosystems, because investigations in this direction are of
special importance for development of buffer capacity of freshwater ecosystems
for heavy metals. While carrying out the above investigations, however, the
facts indirectly proving hydrobiota influence upon metal content in fresh-
water ecosystems were revealed. Only this phenomenon, anyway, can satisfac-
115
-------
torily explain disparity between historical dynamics of lead and mercury in
freshwater invertebrate tissue and the relatively stable level of metals in
bottom sediments of the Usman (Nikanorov and Zhulidov 1984, Nikanorov et al.
1982, 1983, 1985).
In characterizing hydrobiological parameters determining buffer capacity
of freshwater ecosystems for heavy metals, we note that, from the conceptual
point of view, the most important fact that should be taken into account is
that the action of the given group of parameters is directed mainly to partial
excretion of metals introduced to ecosystem from the biogeochemical cycle by
way of their burial in bottom sediments in the process of sedimenttogenesis
(Jackson 1978, Parker et al. 1982, De Master and Nittroner 1983, Hamilton-
Taylor et al. 1984). In this case, the measure of ecological efficiency of
metal excretion from the biogeochemical cycle by way of their burial in bot-
en
=L
o5
_^^
^
>
c
• — "
CD
§
O
"c
0)
I
CD
"o
o
"£
Jll*
f^
O
t
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0 ° Copper
0 A Lead
° A1'3 o A1'2 • Mercury
o °
oo o
0 A
0
o ,
o
A
o A
o
0 °
o
A A °
O o °
A°
A ° °
0 ° ' ° A
A A
A
° .A °
A ^
0 A o
A „ A °
- .
* ° A
• o A A
•* • • • *
. . • • . •.••:.._
200 400 600 800 1000
Biomass of Phytoplankton, Zooplankton, Periphyton
and Benthos, mg/m3 (Periphyton and Benthos
Converted into m2)
Figure 4. Chemical elements in freshwater ecosystems.
116
-------
torn sediments and the ecological consequences of this process for living
organisms will depend mainly on correlative rates of two processes—complexa-
tion of heavy metals introduced to ecosystem with dissolved organic matter
and transformation of metals into less toxic forms under the influence of
this process on one hand and accumulation of metals (and the most toxic forms
among them) by hydrobiota on the other.
At present, there is no unambiguous information on the kinetics of these
processes. It was provided, however, that in the process of cultivation of
blue-greens Chroococcus paris in the water with concentration of heavy metals
(Cd, Cu, Zn) from 2 ppm up to 90% of the total quantity of metals was sorbed
by the cells within 1 minute and fixation was practically complete in 10
minutes (Les and Walker 1984). In this case, absorptivity of the cell at pH
7 was 53, 120, and 65 mg/g dry weight for cadmium, copper, and zinc, respec-
tively (Les and Walker 1984).
REFERENCES
Barbault, M. 1985. Partage des ressources et organisation des peuplements.
Bull. Ecol. 19(l):63-68.
Berglind, R., G. Dave, and M.L. Sjobeck. 1985. The effects of lead on amino-
levelinic acid dehydratase activity, growth, hemoglobin content and repro-
duction in Daphnia magna. Ecotoxicol. and Environ. Safety. 9(2):216-229.
Blaylock, B.C., M.L. Frank, and J.F. McCarthy. 1985. Comparative toxicity
of copper and acridine to fish, Daphnia, and algae. Environ. Toxicol.
Chem. 4(1):63-71.
Borgmann, U. and C.C. Charlton. 1984. Copper complexation and toxicity to
Daphnia in natural waters. J. Great Lakes Res. 10(4):393-398.
Borgmann, U. and K.M. Ralph. 1983. Compiexation and toxicity of copper free
metal bioassay technique. Water Res. 17(11):1697-1703.
Boynton, W.R., C.A. Hill, P.C. Faikowski, C.W. Feefe, and W.M. Kemp. 1983.
Phytoplankton productivity in aquatic ecosystems. In: Physiol. Plant
Ecol. 5. Berlin, p. 305-327.
Breymeyer, A.I. 1979. Ecosystem homeostasis—search for a definition. Mem.
Zool. 32:3-11.
Briand, F., R. Trucco, and S. Ramamoorthy. 1978. Correlation between spe-
cific algae and heavy metal binding in lakes. J. Fish. Research Board
Can. 35(11):1482-1485.
Bronmark, C. 1985. Freshwater snail diversity: effects of pond area, habi-
tat heterogeneity and isolation. Oecologia. 67(1):127-131.
117
-------
Brown, J.H. 1984. On the relationship between abundance and distribution of
species. Amer. Natur. 123(2):255-279.
Brown, B.T. and B.M. Rattian. 1979. Toxicity of soluble copper and other
metals to Elodea canadensis. Environ. Pollut. 20(4):303-314.
Cairns, M.A., A.V. Nebeker, J.H. Gakstatter, and W.L. Griffis. 1984. Tox-
icity of copper-spiked sediments to freshwater invertebrates. Environ.
Toxicol. Chem. 3(3) -.435-445.
Davies, R.B. and B. Hutton. 1975. The effects of errors in the independent
variables in linear regression. Biometrica. 62:383-391.
DeMaster, D.J. and C.A. Nittrouer. 1983. Uptake, dissolution, and accumula-
tion of silica near the mouth of the Chanjiang River. In: Proc. Int.
Symp. Sediment Contin. Shelf Spec. Ref. East China Sea. Hangzhou. p.
215-219.
Falkner, G., P. Stresser, and D. Graffins. 1984. Phosphate uptake by blue-
green algae during an algal bloom. Verh. Int. Ver. theoret. and Angew.
Limnol. 22(1):195-199.
Flatau, G.N., R.T. Clement, and M.J. Ganthier. 1984. Fixation du cadmium par
une Pseudomonadacee marine vivante on tnee par lethanol ou le cyanure de
potassium. Chemosphere. 13(12):1397-1400.
Futuyma, D. 1973. Community structure and stability in constant environments.
Amer. Natur. 107(955):443-446.
Ghilarov, A.H. and A.G. Timonin. 1972. Relations between biomass and species
diversity in marine and freshwater zooplankton communities. Oikos. 23
(2):190-196.
Giesy J.P. 1983. Biological control of trace metal equilibria in surface
waters. In: Trace element speciation in surface waters and its ecologi-
cal implication. G.G. Leppard (ed.). Plenum Publishers Corp. p. 195-
210.
Giesy, J.P., A. Newill, and G.J. Leversee. 1983. Copper speciation in soft
acid, humic waters: effects on copper bioaccumulation by and toxicity
to Simocephalus serrulatus (Daphnidae). Sci. Total. Environ. 28:23-26.
Gurtz, M.E. and J.B. Wallace. 1984. Substrate-mediated response of stream
invertebrates to disturbance. Ecology. 65(5):156-1569.
Hamilton-Taylor, J., M. Willis, and C.S. Reynolds. 1984. Dispositional
fluxes of metals and phytoplankton in Winderrnere as measured by sedi-
ment traps. Limnol. Oceanogr. 29(4):695-710.
Hellmann, H. 1983. Zum begriff der anreicherung in der umweltschtz-diskus-
sion. Teil 11: Mechanismus and bewertung der anreicherung von spur-
ens toff en in gew ass em. Dtsch. gewassert Mitt. 27(5-6):146-153.
118
-------
Jackson, T.A. 1978. A biogeochemical study of heavy metals in lakes and
streams, and a proposed method for limiting heavy-metal pollution of nat-
ural waters. Verh. Int. Ver. theor. and angew. Limnol. 20(3):1945-1946.
Jones, J.R., M.M. Smart, and J.H. Burroughs. 1984. Factors related to algal
biomass in Missouri Ozark streams. Verh. Int. Ver. theor. und angew.
Limnol. 22(3):1867-1875.
Kimmerer, W.J. 1984. Diversity/stability: a criticism. Ecology. 65(6)-
1936-1938. y J'
Kindlmann, P. and J. Leps. 1985. What is stability? A mathematician's and
ecologist's point of view. Math. Res. 28:201-204.
Kolasa, J. and E. Bieseadka. 1984. Diversity concept in ecology. Acta
biother. 33(3):145-162.
Konemann, W.H. 1984. Ecotoxicology and environmental quality. Environ.
Protection: Standards, Compliance and Cost. Chichester. p. 94-103.
Les, A. and R.W. Walker. 1984. Toxicity and binding of copper, zinc, and
cadmium by the blue-green algae Chronococcus paris. Water, Air, Soil
Pollut. 23(2):129-139. '.
Matsunaga, K., K. Igarashi, S. Fukase, and H. Tsubota. 1984. Behavior of
organically bound iron in seawater of estuaries. Estuarine, Coastal and
Shelf Sci. 18(6):615-622.
Menargues, L. and E. Lanza. 1983. Speciation du cuivre par redissolution
anodique dans un milieu de culture de phytoplancton a differente stades
de son developpement. Vie Mar. 5:53-56.
Misztal, M., D. Krupa, and H. Small. 1984. The chemical composition of bot-
tom sediments and phytoplankton in the man-made Lake Zemborzyce near
Lublin. Acta Hydrobiol. 25-26:123-133.
Morel, F. 1985. Iron uptake and phytoplankton growth. In: Rev. Port. Quim,
Second Int. Conf. on Bioinorg. Chem. Abstr. p. 140-141.
Murdoch, W.W., M.A. Scott, and P. Ebworth. 1984. Effects of the general
predator Notonecta (Hemiptera) upon a freshwater community. J. Amin.
Ecol. 53(3):791-808.
Noon, B.R., D. Dawson, and J.P. Kelly. 1985. A search for stability gradi-
ents in North American breeding bird communities. Auk. 102(1): 64-81.
Nunney, L. 1985. Short time delays in population models: a role in enhanc-
ing stability. Ecology. 66(6):1849-1858.
Parker, J.I., K.A. Stanlaw, J.S. Marshall, and C.W. Kennedy. 1982. Sorption
and sedimentation of Zn and Cd by seston in southern Lake Michigan. J.
Great Lakes Res. 8(3):520-531.
119
-------
Peterfi, L.S., L. Momeu, and M. Veres. 1982. Efectul concentratiei ionice
asupra stabilitatii comunitatilor de diatomee in citeva izvoare minerale
din Romania. In: Evol. si adapt., Club-Napoca. p. 119-128.
Ravera, 0. 1983. Considerations on some ecological principles. In: Trends.
Ecol. Res. 1980. Proc. NATO ARW and INTECOL Workshop. Future and Use
Ecol. Decade Environ. Lonvain-la-Neuve. p. 145-162.
Reddy, N.M., and P.V. Rao. 1983. A possible mechanism of detoxification of
copper in the freshwater mollusc Lymnaea luteola. Indian J. Physiol.
and Pharmacol. 27(4):283-288.
Reice, S. 1985. Experimental disturbance and the maintenance of species
diversity in a stream community. Oecologia. 67(l):90-97.
Remmert, H. 1984. Okologie. Dritte, Neubearbeitete und Erweiterte Auflage.
Springer-Verlag, Berlin. 334 p.
Rygg B. 1985. Etfect of sediment copper on benthic fauna. Mar. Ecol.
'progr. Ser. 25(1):83-89.
Skowronski, T. Uptake of cadmium by Stichococcus bacillaris. Chemosphere.
13(2):1385-1389.
Smalls, I. and D. Cannon. 1983. Growth response of phytoplankton to environ-
mental factors. Austral. Water Resour. Counc. Conf. Ser. 7:58-0.
Smith, R.E.H. and J. Kallf. 1985. Phosphorus competition among phytoplankton:
a reply. Limnol. and Oceanogr. 40(2):440-444.
Sommer, U. and S.S. Kilham. 1985. Phytoplankton natural community competition
experiments: a reinterpretation. Limnol. and Oceanogr. 30(2):436-440.
Sposito, G. 1981. Trace metals in contaminated waters. Environ. Sci. Tech-
no!. 15(4):396-403.
Stary, J., B. Havlik, K. Kratzer, J. Pra'silova, and J. Hanusova. 1983. Cumu-
lation of zinc, cadmium and mercury on the alga Scenedesmus obliquus.
Acta Hydrochim. Hydrobiol. 11(4):401-409.
Stenseth N.C. 1985. The structure of food webs predicted from optimal food
selection models: an alternative to Pimm's stability hypothesis. Oikos.
44(2):361-364.
Svetlosanov, V.A. 1983. Model investigation of the problem "stability of
ecosystems." Ecologia. 2(3):255-261.
Thorp, J.H. and M.L. Cothron. Regulation of freshwater community structure at
multiple intensities of dragonfly predation. Ecology. 65(5):1546-1555.
Trojan, P. 1979. Units of homeostatic organization in terrestrial systems.
Mem. Zool. 32:13-23.
120
-------
Trojan, P. 1984. Ecosystem Homeostasis. W. Junk Publishers, The Nether-
lands. 140 p.
Ulrich, B. 1984. Stability and destabilization of Central European forest
ecosystem—a theoretical based approach. In: Trends Ecol» Res. Proc.
NATO ARW and INTECOL Workshop Future and Use Ecol., Decade Environ.,
Louvain-la-Neuve. p. 217-237.
Vymazel, J. 1984. Short-term uptake of heavy metals by periphyton algae.
Hydrobiologia. 119(3):171-179.
Winner, R.W, 1984. The tpxicity and bioaccumulation of cadmium an copper
as affected by hum! acid. Aquat. Toxicoi. 5(3):267-274.
Winner, R.W. 1985. Bioaccumulation and toxicity of copper as affected by
interactions between humic acid and water hardness. Water Res. 19(4):
449-455.
Wood, A.M. 1983. Available copper ligands and the apparent bioavailabiliy
of copper to natural phytoplankton assemblages. Sci. Total Environ.
28:51-64.
121
-------
COMMUNITY RESPONSE OF AQUATIC ORGANISMS TO PESTICIDE STRESS
by
S.J. Lozanol
INTRODUCTION
Problems in pollution studies often require the investigator to infer
species-environmental relationships from community composition data and habi-
tat measurements. Usually the data consist of abundance values for several
species and measurements of natural habitat. For pollution studies, habitat
measurements would include the water and/or sediment concentrations of a toxi-
cant. When sufficient differences exist between sites (e.g., lake, river
reach or experimental pond), it would be useful to use multivariate techniques
to demonstrate multi-species relationships between habitats. Multivariate
analysis techniques were used to describe the changes in the zooplankton
communities along a toxicant gradient in a literal enclosure study conducted
in a small, 5-ha pond in northern Minnesota.
The research program (Brazner et al. 1987) was designed to produce new
and improved procedures for determining primary and secondary ecological
effects of pesticides on non-target biota in natural aquatic systems. Chlor-
pyrifos was chosen as a model pesticide for the study because of prior ex-
perience in testing the compound in the laboratory and in outdoor systems at
the Environmental Research Laboratory-Duluth (Holcombe et al. 1982, Jarvinen
and Tanner 1982, Eaton et al. 1984) and because of the availability of
numerous laboratory and field studies using chlorpyrifos as a test chemical
(Marshall and Roberts 1978).
The research design was to use 12 (5m x 10m) enclosures constructed on
the north side of a 2-ha pond that has a diverse community of aquatic plants
and animals. The littoral areas of the pond are well developed with cat-
tails, pond grasses, and macrophytes growing in unconsolidated, highly
organic sediments. An inert polyolefin film formed three walls of each en-
closure with the fourth side (5m in length) being the shore of the pond.
Average depth of the enclosures at the deepest end was l.lm. Because the
enclosures included natural shoreline, littoral zone and sediments, all
ICenter for Lake Superior Environmental Studies, University of Wisconsin-
Superior, Superior, Wisconsin USA.
122
-------
components of the habitat and food supply for endemic pond organisms were
available. This allowed replication to be incorporated into the experimental
design without seriously compromising the ecological integrity of the results.
Responses of all major trophic levels and the physical and chemical en-
vironment to a single application of chlorpyrifos were measured during the
summer of 1986. Pre-treatment samples for characterizing biological, physical^
and chemical conditions were collected 10 days prior to pesticide application.
Post-treatment samples were collected periodically up to 405 days after appli-
cation when chlorpyrifos was no longer measurable in the water column. Chlor-^
pyrifos was applied to achieve the following average peak concentrations.
Treatment Group
Expected Concentrations
(ug/L)
Number of
Replicate Enclosures
Untreated
Low Concentration
Medium Concentration
High Concentration
0.0
0.5
5.0
20.0
2
2
4
4
The concentrations of chlorpyrifos were selected to encompass reported LC5Q
values for a wide variety of organisms present in the pond and the prescribed
application rates recommended for mosquito control.
EXPERIMENTAL
ENVIRONMENTAL CHEMISTRY
The water chemistry parameters of conductivity, pH, alkalinity, color,
turbidity, dissolved organic carbon (DOC), dissolved nitrate, and dissolved
phosphate were monitored weekly in each enclosure to reveal any dose-response
relationship between pesticide application and the environmental chemistry of
the enclosures. In addition, enclosures were monitored continuously for con-
ductivity, pH, temperature, and dissolved oxygen (DO). Chlorpyrifos [0.0-
diethyl 0-(3,5,6-trichloro-2-pyridyl) phosphorothioate] emulsifiable concen-
trate (22.4% active ingredient) was applied on June 16 at 10 a.m. A single
application was made to each of the 10 treatment enclosures. Enclosure
applications were spaced 15 min apart and the order of application was from
low to high concentration. A composite water sample was collected from each
enclosure at pre-determined times. Each enclosure was divided by volume into
four equal quadrants. Samples from each treatment enclosure were collected
at 1 day pre-treatment and at 1, 2, 4, 8 and 12 hours and 1, 2, 4, 8, 16, 32,
64, and 128 days after treatment.
123
-------
INVERTEBRATE STUDIES
Inverted funnel traps were used to sample zooplankton to determine
abundance and composition. This sampler, designed for use in littoral areas
(Whiteside and Williams 1975), is placed on the substrate where it collects
organisms as they move out of the vegetation at night. Samples are relatively
free of detritus and are easily counted.
The sampler consists of a plexiglass plate to which three glass funnels
are attached. The funnel mouths are placed on the substrate and the organ-
isms move up through the funnels into collection bottles centered around the
stem of each funnel. Samplers were placed at sunset in four selected loca-
tions in'the deeper zones of each enclosure, and retrieved early the follow-
ing morning. Samplers were preserved in 5% formalin. In the laboratory,
zooplankton were concentrated through a 80-U screen and identified as to
genus or species using a compound microscope.
MULTIVARIATE DATA ANALYSIS
For our analysis, we used the complete Iog10 zooplankton counts per
square meter from Day 4 and the preliminary count data from Day 4 to Day 404.
Stepwise discriminant analysis was employed to examine the effects of chlor-
pyrifos on the zooplankton community on Day 4. Discriminant analysis was done
through the computer programs available in Statistical Package for the Social
Sciences (Nie et al. 1970). Detrended correspondence analysis (Hill and
Gaugh 1980) was used to analyze the long term effects from Day 4 to Day 404.
Detrended correspondence analysis has been used in numerous field studies and
has the distinct advantage of producing axes that correspond to approximate
standard deviation units of zooplankton species counts. On the axes, 100
units are equivalent to one average standard deviation, which is close to one
half-change for many species replacements along an environmental gradient
(Hill and Gaugh 1980).
RESULTS AND DISCUSSION.
Analysis of pre-treatment samples revealed that the microinvertebrate
community in early June was dominated by cladocerans and copepods. The per-
centage of total abundance per square meter was 48 percent for cladocerans
and 28 percent for copepods. Rotifer populations, at 22 percent of total
abundance in pre-treatment samples, were also at their highest in June based
on preliminary results from pre-treatment to Day 128.
Pre-treatment percent similarity between reference versus low, medium
and high treatment populations was measured by Renkonen's Index (Renkonen
1938) for microinvertebrate funnel trap data. Percent similarities of spe-
cies abundance in the treatment enclosures relative to the control enclosures
was high, ranging from 81% for the low treatment enclosure to 88% for the
medium and high treatment enclosures. After pesticide application, there was
an overall decline in abundance ot species within the different treatment
enclosures.
124
-------
The microinvertebrate community was reduced at all three treatment
levels, with 30 of 35 taxa declining when compared to reference populations
(Table 1). Total abundance of the treated populations was reduced by at
least 60% as compared to reference populations. Cladocerans and ostraced
populations from treated enclosures (Table 1) were smaller (75 percent to 99
percent reduction) than the reference populations with almost complete mor-
tality at the medium and high treatment application rates. There was little
or no statistically significant impacts on copepod and rotifer populations.
Discriminant analysis was used to define those zooplankton species that
contributed to the differences between treatment levels in zooplankton com-
munities. The individual variables used in the analysis are listed along with
the relevant standardized function coefficients in Table 2. All three dis-
criminant functions were significant, although the first two functions ex-
plained 93 percent of the total variation. The effects of chloripyrifos on
zooplankton community responses were clearly demonstrated in the Day 4 samples
(Figure 1). There was no overlap between treatment groups classification.
It appears that the first discriminant function separates treated en-
closures from the control zooplankton communities. The second discriminant
function separates the three treatment groups (low to high). There appeared
to be no pattern to the order that a zooplankton taxa was selected for enter-
ing the discriminant analysis. Cyclocypris (ostracod), Mytillina (rotifer),
Copepodite (copepod) and Alona (cladoceran) were the first four species that
were included in the discriminant analysis. Most cladocerans and the one
species of ostaracods (sensitive species to chlorpyrifos) were correlated,
however, with the first discriminant function, whereas rotifers and copepods
were correlated with the second discriminant function (both groups were not
sensitive to chlorpyrifos). These results correspond to the analysis of
variance results (Table 1) discussed above.
Sayler et al. (1983) used a discriminant analysis to evaluate the sig-
nificance of synthetic oil on the functional activity of sediment microbial
communities. Effects of the synthetic oil on the microbial community re-
sponses were evident 1 month after treatment. In this study, the microin-
vertebrate community began recovery after 3 months.
To show the long term effects of chlorpyrifos in the enclosure studies,
a detrended correspondence analysis was employed (Hill and Gauch 1980). The
sample points, corresponding to control or treated enclosures, were used to
describe the community trajectories (vectors) in species space. Like all
ordination techniques, detrended correspondence analysis is a data reduction
technique, reducing -n—species space onto three or four axes. The goal of the
analysis is to arrange the individual species in a manner that discloses
their fundamental relationships, i.e., species abundance is used to reveal
the relationship between zooplankton communities and the habitat, in this
case the four treatment level groups (control, low, medium, high).
The results of the ordination are shown in Figures 2-4 for each treat-
ment comparison. By Day 64, the low treatment zooplankton community had
returned to the control cluster (Figure 2). For the medium and high treat-
ment groups, there was some evidence of recovery by Day 405, more than 1 year
125
-------
TABLE 1. ESTIMATED M1CROINVERTEBRATE ABUNDANCE GIVEN AS GEOMETRIC MEANS OF
ORGANISM NUMBERS PER SQUARE METER, TREATED AND CONTROL (UNTREATED)
ENCLOSURES, 1986 LITTORAL ENCLOSURE STUDY.
Taxa
Cladocerans
Chydorus
Pleuroxus
Simocephalus
Alona
Period aphnia
Copepoda
Acanthocyclops
Mesocyclops
Eucylcops
Nauplii
Copepodites
Harpacticoid
Unknown sp.
Ostracoda
Cyclocypris
Rotifera
Brachionus
Platyias
Monostyla
Polyarthra
Trichocerca
Keratella
Testudinella
Notholca
Lecane
Asplanchna
Euchalnis
Trichotria
Ploesoma
Mytillina
Unknown sp.
Other Invertebrates
Amoeba
Planaria
Nematode
Hydracarina
Control (+_ 2S.
762.0 (1783 -
35.2 ( 85 -
32.0 ( 77 -
39.1 ( 61 -
41.0 ( 57 -
137.7 ( 360 -
76.4 ( 203 -
32.7 ( 128 -
504.8 (1026 -
407.0 (1362 -
103.7 ( 189 -
0.8 ( 3 -
116.9 ( 191 -
37.8 ( 97 -
54.8 ( 146 -
67.8 ( 173 -
4.0 ( 10 -
101.5 ( 248 -
3.0 ( 11 -
31.8 ( 65 -
133.6 ( 392 -
4.3 ( 8 -
5.7 ( 10 -
73.5 ( 289 -
12.8 ( 27 -
4.2 ( 16 -
17.5 ( 39 -
29.4 ( 88 -
28.3 ( 43 -
140.8 ( 387 -
9.8 ( 13 -
9.4 ( 17 -
E.)
325)
14)
13)
25)
29)
52)
28)
8)
248)
121)
57)
0)
71)
14)
20)
26)
1)
41)
0)
15)
45)
2)
3)
18)
6)
1)
8)
y)
18)
51)
7)
5)
Low
5.9 *
0.3 *
1.5 *
8.6 *
0.1 *
19.8
10.3
14.4
103.4
54.1
12.6
0.3
28.7 *
15.2
1.9
6.3 *
1.7
28.4
3.7
6.5
45.1
0.5
0.0
8.2
8.8
3.5
7.1
32.7
95.8
4.4
0.8
19.0
Med ium
1.3 *
0.2 *
0.1 *
0.2 *
0.0 *
4.5 *
30.7
12.9
185.9
75.0
90.4
1.6
16.7 *
38.1
11.5
19.1
18.2
69.6
17.1
25.8
42.2
1.1
1.4
23.2
6.2
3.2
8.0
35.4
52.5
138.9
5.4
17.3
High
0.8 *
0.0 *
0.0 *
0.0 *
0.0 *
1.5 *
18.7
11.8
140.1
51.3 *
77.8
1.0
3.8 *
25.4
14.0
9.1 *
8.7
25.7
10.2
19.7
50.3
2.0
1.9
15.7
10.4
2.0
2.8
16.8
74.2
63.5
1.5 *
8.7
*Tukey's Studentized Range Test, p <^ 0.05.
126
-------
aJ ti-r pesticide application (Figures 3 arid 4). The, recovery rates lor the
low treatment was comparable to the microblal communities reported by Sayler
et al. (1983). The enclosure walls were not removed until 45U days after
treatment, thereby eliminating the possibility of zo.oplankton colonization
from untreated to treated littoral zones. The slower recovery rates in the
medium and high treatment groups for the zooplankton community may have been
more rapid if the enclosure walls had been removed. As stated earlier, the
species that were most important in the ordination .analysis were sensitive
TABLE 2. ZOOPLANKTON VARIABLES UTILIZED BY STEPWISE DISCRIMINANT ANALYSIS
TO DISCRIMINATE BETWEEN EXPERIMENTAL ENCLOSURES AFTER CHLOPYRIFOS
TREATMENT.
Discriminating
variable
Eigenvalues
Percent Variation
Explained' -
Wilk's Lambda
Discriminant function coefficient
Function 1 Function 2 Function 3
Cladocerans
Chydorus sphaericus
Pleuroxus
Sida crystallina
Alona costata
Ceriodaphnia
Copepoda
Acanthocy clops
Nauplii
Copepodites
Os traced a
Cyclocypris
Rotif era
Platyias
Monostyla
Notholca
Lecane
Trichotria
Ploesoma
Mytillina
Other Invertebrates
Amoeba
Planaria
0.60
-0.55
0.64
-0.09
1.06
-0.84
-0.52
0.81
0.11
-0.61-
-0.08 ?"
1.12
0.44
0.44
0.55
-0.05
-1.24
0.48
-0.16
-0.66
0.02
-0.14
0.83
-0.64
-2.22
1.7.2
-1.28
....
1.03
-0.43
1 .56
,0.68
0.68
1.26
-0.49
-0.52
0.33
0.19
-0.11
-0.36
1.08
-0.09
-0.26
-1.25
0.34
-0.33
-0.68
-0.68
0.89
0.65
0.65
-0.01
0.23
0.72
-0.53
28 .
64
0.001*
13
29
0.018*
7
0.245*
<_' 0.05, Chi Squared Test
127
-------
species including cladocerans, amphipods and ostracods. Species that were
important for determining the second axis were rotiters, copepods and two
species of cladocerans. Under the constraints of our experimental design,
we have found that recovery for zooplankton species from a littoral zone of
a pond is negatively correlated with the concentration of chlorpyrifos.
In summary, the two methods of multivariate analysis have succesfully
demonstrated the relationship between zooplankton community and habitat.
Recovery can be demonstrated for different pesticide treatment levels and a
correspondence between zooplankton populations and habitat was established.
o
20
oJ CL
u_
-------
DCA2
140
120
100
80
60
40
20 -
0 -
O Control
D Low Treatment
August
July
^ June
0
20
40
60
80
100
DCA 1
Figure 2. Detrended correspondence analysis—zooplankton. Distribution of
samples on first and second detrended correspondence ordianational axes.
Codes within circles (control enclosures) and squares (treatment enclosures)
are days relative to chlorpyrifos treatment. Low enclosures are boxed to
emphasize the temporal succession of the treated zooplankton community.
REFERENCES
Brazner, J.C., S.J. Lozano, M.L. Knuth, L.J. Heinis, D.A. Jensen, K.W. Sar-
gent, S.L. O'Halloran, S.L. Bertelsen, O.K. Tanner, E.R. Kline & R.E.
Siefert. 1987. The effects of chlorpyritos on a natural aquatic system:
A research design for littoral enclosure studies and preliminary data
report. Progress Report, U.S. EPA, ERL-Duluth, Duluth, MM.
Brock, D.A. 1977. Comparison of community similarity indices. J. Water
Pollut. Control Fed. 49:2488-2494.
129
-------
Eaton, J.G., J. Arthur, R. Hermanutz, R. Kieter, L. Mueller, R. Anderson,
R. Erickson, B. Nordling, J. Rogers & H. Prtichard. 1984. Biological
effects ot Uursban on an outdoor experimental stream ecosystem. Project
Report, U.S. EPA, ERL-Duluth, Duluth, MN.
Hill, M.O. and H.G. Gaugh. 1980.
proved ordination technique.
Detrended correspondence analysis, an im-
Vegetatio 42:47-58.
Holcombe, G.W., G.L. Phipps & D.K. Tanner. 1982. The acute toxicity of
Kelthane, Dursbane, bisulfoton, Pydrin, and Permethrin to fathead
minnows Pimephales promelas and rainbow trout Salino gairdnen. Environ.
PoLlut. (Series A) 29:167-178.
DCA2
100
80
60
40
20
1987
GO
Q
S
lie] \
July
August
O Control
D Medium Treatment
October
0
20
40
60
DCA 1
80
100
120
Figure 3. Detrended correspondence analysis-zooplankton. Distribution of
samples on first and second detrended correspondence ordxnatxonal axes.
Codes within circles (control enclosures) and squares (treatment enclosures)
art days relative to chlorpyrifos treatment. Medium enclosures --boxed to
emphasize the temporal succession of the treated zooplankton community.
130
-------
DCA2
100
80
60
40
20
0
1987
(4)
a
El
June
ra B
July
„/' August
October
O Control
Q High Treatment
20
40
60
-------
Sayler, G.S., R.E. Perkins, T.W. Sherriil, B.K. perkins, M.C. Reid, M.S.
Shields, H.L. Long, and J.W. Davis. iy83. Microcosm and experimental
pond evaluation of microbial community response to synthetic oil con-
tamination to freshwater sediments. Applied, and Environmental Micro-
biology. 46:211-219.
Whiteside, M.C. and J.B. Williams. 1975. A new sampling technique for aquatic
ecologists. Verh. Int. Verein. Limnol. 19:1534-1539.
132
-------
THEORETICAL AND METHODOLOGICAL ASPECTS
OF MODELING LACUSTRINE ECOSYSTEMS
by
A.A. Matveyev1, A.M. Nikanorov1, Yu. A. Dombrovskiy2, and V.V. Selytin2
ABSTRACT
Selection of a particular mathematical model for simulating an environ-
mental ecosystem is largely a subjective process. This paper examines some
means of reducing this subjectivity by identifying methods that facilitate
an a priori evaluation of the required area, the dimensions, and the degree
of detail desired.
INTRODUCTION
A broad range of ecosystem water models is utilized at the present time,
from the simplest inlet-outlet models that include empirically selected re-
lationships to cumbersome models that have dozens and even hundreds of varia-
ble simulation systems. Models also are differentiated according to degree
of localization, from those that describe individual bays and shore areas to
models depicting the entire water body.
In selecting a particular mathematical method for describing a system
and obtaining the desired degree of model resolution, the researcher is
guided by the purpose of the modeling effort, the available data, and the
technical capabilities of each model. To a large part, such a choice is
largely subjective. The purpose of this work is to examine some of the means
of reducing this subjectivity by identifying methods that facilitate an a
priori evaluation of the required area, the dimensions, and degree of detail.
Also discussed are some standardized quantitative analysis processes for
ecosystems such as Lake Baikal and Lake Sevan.
PRIMARY WATER ECOSYSTEM MODEL AND MODIFICATIONS
The primary orientation in a mathematically modeled ecosystem for water
bodies utilizes the following model
^Hydrochemical Institute, Rostov-on-Don, USSR.
2Mechanics and Applied Mathematics Research Institute, Rostov State Univer-
sity, Rostov-on-Don, USSR.
133
-------
— = div Dgrady - vgrady - K(y) ,
Dt
divV = 0
y(0) = y0
vy - Dgrady|r = q,
(1)
(2)
(3)
(4)
where:
yfc-RJj} - ecosystem condition vector
D - dispersion matrix
v - transport velocity vector field
K(y) - kinetic operator
y0 - initial condition vector
q - substance flow over boundary surface G
As a result of the commonality of physical principles that lie at the
system's basis, most of the existing models of spatially distributed systems
may be directly derived from Equations 1 through 4. The construction of the
models is generally differentiated by:
o space-and-time aggregation, and consequently by mathematical formal-
ism (partially derived equations, ordinary differential equations,
and difference equations)
o dimensions and degree of non-linear concrete understanding of the
kinetic member K(y)
o ability to define the vector field V (calculation with the aid of a
special hydrodynamic model or formation on the basis of full-scale
survey data)
This model analysis primarily relies on two modifications of the origi-
nal model (Equations 1 through 4), which are shown in abbreviated form below.
CHAMBER (COMPARTMENTAL) MODEL
If a water body is divided into M fixed spaces (chambers) and a new con-
dition vector is introduced
134
-------
where Yr is the reserve or concentration of the i-component in the K-chamber,
and
r-1
K = entier( ) + 1,
N
(5)
i = r-N(K-l)
(6)
Then Equations 1 through 4 may be shown as
r =
j£N(t)
(7)
where:
M(k) - the number of chambers that border with K-chamber
P.. - substance flow from j to i component of the ecosystem
during the course of the biotic cycle
N(i) - number of components interacting with i-component
Here Pr depends on the dimensions of the chambers and parameters D and V, the
overall advective and diffusion overflow of substance from chamber C into
chamber K in a unit of time.
We should note that the first total (Eq. 7) also includes interchange
flows with external mediums Poe and Peo, i.e., it takes into account the
boundary condition (Eq. 4); whereas the second total, which represents the
kinetic member, was derived on the basis of the additive principle of biotic
flows. Equation 7 may be written in a matrix form:
y = (F-F0)y + u, (8)
P = Fy, (9)
where:
F(frs), F = diag (fP) - non-negating matrix dimensions (NxM) x (NxM)
i ••
P - production vector (Pr = ^Pre+ Pri)
e
u - intake vector
The apparent correlation
(10)
-------
ensures that solutions 8 and 9 are positive.
In stationary conditions
(ii)
Knowing matrix W makes it possible to calculate the contributions of any in-
flow U collected in K-chamber:
(12)
Calculation of the relative discharge ^s of the three rivers that flow
into Lake Baikal: Selenga (S = 15), Barguzin (S = 26), and Upper Angara
(S - 42) was conducted with the aid of a chamber water exchange model (M -
and showed that the most significant contribution to the levels of concentra-
tion in all areas of the lake comes from Selenga. The values of relative
depositions cxk)!5 range from 0.69-0.80 (hydrocarbons), 0.28-0.42 (phenols),
0.46-0.65 (zinc, lead). The values of relative depositions ak,42 are suf-
ficiently low and consist of 0.11-020 for hydrocarbons, 0.10-0.31 for phenols,
and 0.08-0.21 for zinc and lead.
DISCRETE MODEL (DYNAMIC BALANCE METHOD)
Given a certain characteristic time interval T and going from deriva-
tives to differences, one may get the discrete analysis of Equations 8 and 9.
The production-balance modification of the difference model has the clearest
physical meaning, which may be formally constructed by implicit approxima-
tion as
t+T
Pydt « TFy
(13)
Then, it can be easily shown that Equations 8 and 9 become
- B(yt + PT + Uc),
(14)
PT = A(yt + P
B
A = TFB,
(15)
136
-------
I = (6r), Ut
t+t
System 14 and 15 characterize the dynamic balance method. Let's note
that the stationary condition matrix W is now expressed through A, B:
W = B[I-(A+B)]-1
(16)
OVERALL SUBSTANCE DISTRIBUTION FEATURES
FOR THE ENTIRE LAKE BODY
Let us represent the water body in the form of a unidimensional system
that takes in a flow of non-conservative impurities. Then the distribution
of y concentration along a certain characteristic x axis is described by
Equation 17
(17)
whose stationary solution has the following appearance
y =
(18)
where
2D
and
C2 are determined by boundary conditions.
Let L represent the typical water body dimension, one determined by the
distance between inlet and outlet. Assuming that *x , <0, ^2 > ° and
taking into account that function y(x) must be a monotonous reduction, we
get the apparent disparity
(19)
Let us find a condition in which distribution (Eq. 18) is close to being
equal, i.e., y(0) and y(L) are sufficiently close. Keeping in mind (Eq. 19)
and A2 > | AjJ it ls not difficult to establish that the criterion we seek
can be shown by the expression
(20)
137
-------
Let us introduce the following parameters:
T.=I/K - time constant for "removing" impurities
T =L/v=V/AV - time of conditional water change (V - lake volume,
V V - flow volume)
TD=L2/D - diffusion time
TV h -
a- — , b -
*k M T K
Then Eq. 20 is written:
1
2
TkT b
« 1
(21)
Apparently what follows in the evaluation is
, , N
< v» 1 shows that turbulent exchange has a
The correlation - -
T T a
predominant role in the substance dispersion mechanism, and when - < 1
the migration process is dominant. It should be noted that use of all three
Let us apply the given schematic for analyzing the distribution of some
substances in Baikal. Let us evaluate tv , TD , TK
V, 23000 km3
D _
AVb 60 km3/202
138
400 yrs.
-------
The width of Lake Baikal (L = 40 km) may be taken as the characteristic dimen-
sions because the principal migration occurs along the axis of the Selenga
River Delta/Angara source. For such an area, the turbulent diffusion coeffi-
cient is on the order of
cm*
D = 10
Then
D
50 years.
The calculated rates for removing substances (Table 1) are on the order of 5
x 10 L/year or TK = 200 years. Thus, a = 2, B = 0.5 and ye (0.4, 0.5)..
TABLE 1. EXTERNAL BALANCE OF CERTAIN SUBSTANCES IN LAKE BAIKAL
Balance Influx of U, Run-off fr
element tons/year Angara,
— Lous per
substance year
Phenols 310 79
Hydrocarbons 19306 7551
Zinc 513 138
Lead 44 24
om AU Overall amounts Estimates
tons per in lake, Y of removal
year thousands rates, 1C,
per year L/year
231 80.5 2.9 x
11755 2576 4.6 x
375 52.9 7.1 x
20 9.2 2.2 x
10-3
10-3
10-3
10-3
From this we can conclude that the determining mechanism for dispersing
quasiconservative substances in Baikal is the turbulent exchange, ana the
distribution of these substances has a certain inequality
(y(L)/y(0)
0.6
Let us note that for the more labile impurities, although they are removed
slowly from the ecosystem (i.e., Tk = 2 year), a = 200, B = 5, M = 5,
yB*s/yo ~ e " °-007> which explains the localized effect of the affecting
zones of some of the sources, having a relatively high level of concentra-
tion. (Another cause of localization may be dilution by purer sources.)
COMPARING COMPLETE DISPERSION AND MIGRATION MODELS
When time correlation TD } TV> Tfc is such that - »1 and B «1,
b
then the water body may be considered to be completely intermixed and instead
of Eq. 17, a chamber model is used.
139
-------
1 1
(y°-y)
(23)
where y° is the substance concentration upon intake.
o
When - «1, then the diffusion is relatively low and Eq. 17 becomes a migra-
b
tion equation.
1
-yT y,
Tk
(24)
where: T =x/v, Tt[D, TV]
Equations 23 and 24, which describe the finite conditions, are much
simpler than the original (Eq. 17). These equations considerably facilitate
the modeling of the hydrochemical process. In addition there is an absence
of diffusion parameter TD, the assessment of which is always problematic.
Should the diffusion model (Eq. 17) be replaced by one of the approxi-
mate schematics (Eq. 23 or 24), it is necessary to know the magnitude of
error. Let us mark y± for solution 23 and y£ for solution 24 (Figure 1)
where
l+Tv/Tk
(25)
y2 = y°exp(-T/Tk),
(26)
Apparently the maximum error reading of Delta y
Ay < max { y-yj.
(27a)
may be large when replacement of model 17 is inadequate, causing big, relative
errors with the increase of
a - TV/T k
In solving practical problems, it is often sufficient to evaluate only
the average concentration for a given volume. The error rate of such an eval-
uation will be found when Eq. 17 is replace by Eq. 23 or Eq. 24, i.e., con-
forming to indeterminant conditions of parameter TD.
140
-------
Y,(T)
A
"•u
0.3
-e-o.2
0.1
yi, then to get the relative accuracy of
-------
_ V2~vl l-exp(-a)-aexp(-a)
-------
«" / ^ / ^
32 x a x a^
We can easily see in Figure 1 that, as m increases, (32, a^) grows rather
quickly 3nd the quality of the evaluation worsens. In particular, when m
the relative error reading for determining K may get to 80%:
m-l-en m
= 3,
(32)
en m
The high sensitivity of the parameter that has been identified (the solution
of an inverse problem) to the type of model is formally explained by the
weak sensitivity of the direct problem solution (average concentration y)
to parameter a.
Non-stationary Error
Beginning from a certain moment in time tQ, let the concentration of the
substance in the water body increase at the inlet point, for example, accord-
ing to the demonstrative law:
i
y0
r(t-t)
(33)
Then the evaluation of K according to formulas 30 and 31 will result in the
appearance of additional dynamic errors. Let us examine the transfer equa-
tion Eq. 24 first. The nonstationary solution at the outlet point has the
form
y°(t-Tb)=y°-(k"r)Tb+r(t-to),
(34)
from which we derive
K2, t 0 the evaluation of the
kinetic parameter is too low and too high when it is negative. It may be
shown that the evaluation of the complete intermixing model (Eq. 23) has an
identical characteristic:
143
-------
Ki,ttc
where v(t)=expl-(— + K,+r)(t-t0)] When t
STRUCTURE OF THE BIOTIC CYCLE
The design of the ecosystem's kinetic operator, describing the interac-
tion and transformation of its individual components during the biotic cycle,
is an informal process. Here we should take into account both general mech-
anisms that are endemic to lacustrine bodies, as well as their specific
features that at times play a decisive role in the model's synthesis.
The most important moment is the selection of the vector of the ecosys-
tem's state. Naturally, this selection, to a large degree, is determined by
the purpose of the study and the area of its implementation, and is unavoid-
ably tied to aggregation of the elements of the ecosystem. Also taken into
account is any information that is necessary for identification and verifi-
cation.
A quantitative method for resolving the question of the aggregation of
the variables and the selection of an optimal (meaning, the least redundant)
condition vector is closely tied, however, to the analysis of the substance
flow structure and energy in the ecosystem. A partial understanding of these
ilows is derived from the biotic types of the lacustrine ecosystems, for
which a methodology and a means of plotting them was developed by G.G. Vin-
berg. Moving from energy to real aspects and adding the organic and mineral
components, as well as the exchange flows occuring with the environment, we
get the complete picture that characterizes the biotic cycle of biogenic ele-
ments. All these aspects serve as the foundation for developing the mathe-
matical models of ecosystems. The nitrogen cycle dynamic model in the pela-
gic layer of Lake Baikal, one which relied on an expanded biotic picture, was
plotted and studied earlier.
Analysis of the flow structure will be conducted on the basis of an
aggregated functional diagram of the pelagic layer (Figure 2a), one that
relects the principal stages of substance and energy transformation. This
diagram includes three links in the biotic chain:
X - initial producers, autotrophs (phytoplankton on the top)
144
-------
x - Phytoplankton
yi - Labile organic substance
V2 - Aqueous humus
b - Bacteriaplankton
Z - Zooplankton
S - Biogenes
LI, 12 - Water runoff
A
Primary cycle
Slow cycle
Secondary cycle
External cycle
Internal
cycle
\
Rapid cycle
X
Initial cycle
B
Figure 2. Structure of the lacustrine biotic cycle—(A)
aggregated functional schematic, (B) decomposition
diagram for small parameters, and (C) macrostructure.
145
-------
B - heterotrophs, reducers (bacterioplankton, simple forms)
L - heterotrophs, consumers (zooplankton, fish, as well as hydrochemical
components)
S - mineral forms of biogenes (assimilated by phytoplankton)
y^ - dissolved labile organic substance
y2 - aqueous humus
Notation of the two organic substance fractions is due to their different
roles in the hydrochemical and hydrobiologic activity because aqueous humus
is significantly different from labile fractions both in composition and
speed of mineralization.
Marking W = (x, yls y2, B, L> S) as the vector of the condition of the
ecosystem, P - (PJH) the matrix of substance flows between individual com-
ponents, l/i, U2 the vectors of inlet and outlet flows, the following dynamic
form is written
W = (P-PT)
(37)
Averaging Eq. 37 over a particular characteristic period of time (for example,
one year), and omitting members on the right side of the equation that provide
a significantly smaller contribution to the algebraic totals in comparison
with other members of the same sign, we come to the reduced model
W
(P-P1) efD1-U2
(38)
After examining the biotic composition of the lakes, which was derived
during the International Biologic Program, as well as relying on the general-
izing works of G.G. Vinberg, l.B. Ivanova, A.F. Alimova, and B.A. Skopintsev,
it is possible to conclude that there is regular transfer from Eq. 37 to Eq.
38. Therefore, the origianl system (Eq. 37) is divided into three subsystems:
initial cycle (x, yj_, B, S) , slow cycle (y2), and s secondary cycle (L) , which
together with the external cycle that is created by flows U^ and U2 forms the
macrostrueture of the lacustrine cycler (Figure 2c).
The same result may be achieved by a somewhat different means, by compar-
ing the internal and external flows:
(39)
146
-------
where A is the primary products (assimilation), R is the overall destruction.
Splitting the internal cycle on the basis of the speed of transformation into
fast and slow cycles, we have
C»K2,
(40)
where:
C = min (A, R)/q the speed of the rapid cycle
q =x+yi+B+L+S the amount of substance in the rapid cycle
K2 = rate of mineralization of the aqueous humus y2; thus stressing
the basic (initial subsystem in the rapid cycle, having a mass
of qj = x + yi + b + S, and a secondary subsystem (L) that
includes the higher trophic levels:
(41)
This procedure is shown in diagram form on Figure 2b.
Conditions 39 through 41 have a non-rigid form and are done for a broad
range of lacustrine body types. The specifics of the ecosystem, as applicable
to this analysis scheme, are reflected in concrete values of flows A, R, Ui,
U2 (as the degree of flowage in the lake rises, the roles of U]_, U2 is
increased); coefficient values K2 and C (C is reduced from 10-20 L/year for
shallow well intermixing and well heated lakes to 0.1 to 1 L/year for deep
oligotrophic lakes); correlation of labile (q) and hard-to-acidify (y2) sub-
stances (in distrophic water bodies the share of y2 is high and can reach 50%
higher); composition and concrete proportions o£ qL and L (in deep water
bodies due to the relative thinness of the production layer, the basic part
of qx, consists of mineral substances. For Baikal, the variable (and conse-
quently the initial cycle) should include a cymatoa, "rachok epishchura," a
member of the microzooplankton family, because of its significant contribution
to the overall destruction picture (25% R), which compares with breathing of
bacteria (62% R).
In addition to the theoretical aspects, the cross-linking of the biotic
cycle is of interest as a base for constructing reduced models, which are
at the base of theoretical and analytical studies.
REDUCED MODELS OF SUBSTANCE CYCLES IN THE ECOSYSTEMS
OF LAKE BAIKAL AND LAKE SEVAN
The existence of the macros trueture of biotic flows, which was examined
in the previous section, makes it possible to carry out reductions of the
original system. To describe the space-and-time dynamics of phytoplankton
and biogenic elements, it is possible to utilize the initial cycle model.
147
-------
where:
x = ps(x)-Ex,
y = Ex-Ky, (42)
s - Ky-p(s)x,
y - y]L + B, the aggregate of detritus bacteria
p(S) - the rate of initial assimilation
K - the rate of mineralization
E - the die-off rate of the phytoplankton
Below we will examine another reduced model, one which makes it possible to
examine some changing trends in the hydrochemical activity and the primary
products as a result of the changes in the volume and composition of the
substances that enter Lake Baikal.
Let us mark q^ = x + y^ + B + S as the amount of limiting biogenesis tak-
ing part in the initial cycle. Apparently, the variable qj belongs in the
slow category. The slow cycle model has the form
ql = (l-n)Q°b-vq1+K2y2-§A-pA,
y2 = iiQ°b-vy-§A-K2y2,
(43)
where:
A -
QO -
*
v —
p -
II -
§ -
cqj, the primary products
the total concentration of the limiting biogenesis in the inflow
flow velocity
the amount of primary products that is removed from the cycle as
a result of sedimentation and burial
amount of aqueous humus in Q°
amount of aqueous humus in A
rate of aqueous humus mineralization y2
Let us find the stationary solution of system (43);
148
-------
Q°
l-n+§
(44)
y2
K2
where: 6 = —, A = pc/v, y =§ c/v.
(45)
The average content of biogenetic material in the water body is:
Q-qi+y2 = Q<
1+6+A+6M+U
(46)
Formulas 44 through 46 show that, like separate components qit q2, the overall
amount of Q depends not only on the inflow of concentrations Q°, but also on
the composition of the inflow.
Since the primary product is proportional to q,, then based on Eq. 44
it follows that '
6
A!ri=i: A|ri=0 = (47)
1+6
Thus, with sufficiently low correlation of 6 = K2/v, the reduction of the
amount of allochthonous organic substances (in the form of aqueous humus) in
the inflow leads to a proportional increase in the primary products of the
water body. Let us also note that in accordance to the concept shown in model
43, when predicting the content of aqueous humus and the productivity of the
water body, differentiation on the basis of biologic (labile and stable)
aspects rather than the chemical (mineral and organic substances is more
important.
For forecasts using model 43, it is necessary to know the dlmensionless
parameters n, A, 6, y, as well as value Q°.
As a rule, only the most superficial a priori evaluations of the model
coefficients are available, and these are adjusted also by modeling. Let
us show another sufficiently general evaluation method. Let WeW be the
vector of indicators that are directly measured with some accuracy, where
i't.R+- is the area, whose dimensions are determined with a potential accu-
racy W, and Ot 0Q - the vector of unknown paramters, where 8 CR% - the loca-
tion of initial evaluations. If F(0) = W - the equation system that ties in
the parameters and observations, as for example Eq. 44 and Eq. 45, then the
narrowing of initial mass t)Q is given by the expression
= oce
(48)
149
-------
where
= F(u0).
The location i2 is usually represented by an in-dimensional parallelepiped, and,
it reliable evaluations are absent, 00 should quite naturally be taken as
positive.
Based on the available published material by K.K. Votintsev, I.V. Glaz-
unova, L.A. Vykristyuk, Ye. N. Tarasova, V.T. Bogdanova, and A.I. Mesheryakova
of the USSR Academy of Sciences Limnological Institute, as well as on evalua-
tions of the composition and rate of destruction of the organic material by
B.A. Skopintsev, we took the following areas of ^ and 6O for nitrogen and
phosphorus:
350
MI'N
M3~
MFP
< Q° < 450 — ,
MFP
30 < Q° < 40 ,
M3 P M3
MFN MFN
160 < QN < 190 ,
o *•' O
M3 M3
MFP MFP
14 < QD < 18 —- ,
M3
90
10
56
M3
MFP
M3~
MFN
MFP
M3~
87
MFN
M3~
MFP
Ml'N
MFP
20 < v2N/y2P < 28 ,
8 < ^in/^lp < 12 »
0.1 < Q/QO < 0.4 ,
(49)
150
-------
0.3 <
0.2 <
0 <
0 <
TI
X
Y
y
< 0.8 ,
< 2 ,
< 10 ,
< 8 ,
(50)
Evaluation of y2n> Y2P was carried out for Nopn, Popn at large depths; Q^,
Op were calculated on the basis of the total influx of nitrogen and phos-
phorus with the surface runoff and from the atmosphere. To evaluate X, cal-
culations done by L.A. Vykristyuk and K.K. Votintsev for the Copn flow that
ends up in the bottom deposits were utilized. In addition, it was assumed
that the utilized hydrochemical information pertains to the quasi-stationary
activity of Baikal. In accordance with algorithm 48, a congruent central
evaluation with dimensionless parameters was found
n = 0.5
x - i.-5
= 6
= 2,
= 77 mg/m
which corresponds to the values
Qg = 395 mg/m3, QN = 179 mg/m3, qiN = 102 mg/m3, y2
Q° = 36 mg/m3, Qp = 15.3 mg/m3, qlp = 11 mg/m3, y2p = 4,3 mg/m3
*[Note: m-gamma in formula above should be read as mg].
The correlation q^n/q^p = 9.2 indicates adequate balancing of that part of
nitrogen and phosphorous that actively participates in the cycle.
In comparison with other large lakes, the retention coefficient is
relatively low
R =
= 1 - Q/QO « 0.6
(51)
which attests to the need for stricter limitations on the influx of biogenic
materials from the runoff.
Thus, since parameter 6 is sufficiently large, then changes of 11 do not
seriously affect the primary products. The basic factor in its increase is
played by Q°.
A similar mode (Eq. 43) for slow variables was utilized to reconstruct
and predict the trophic state of Lake Sevan, whose natural activity has been
disrupted by dropping water levels between the 1930s and 1970s. This result-
ed in a substantial increase of primary products and worsening of oxygen
151
-------
conditions in the hypolimnion. Because biogenic exchange with the bottom bed
plays a rather significant role in Lake Sevan, whose average depth after the
drawdown is only 31 meters, variables that include the amounts of nitrogen
and phosphorus in the active bottom layer were added to the model.
Figure 3 shows the model reconstruction results of the period when draw-
down occurred, as well as predicted estimates of primary products with reduced
otuflows from the lake and influx from river runoffs. An important result of
the model analysis was the conclusion that one of the main reasons for the in-
crease in primary products (PP) is the too rapid drawdown of the water level
(up to 1 meter per year), which the ecosystem was unable to handle. At the
present time, it has been noted that there has been a downward trend in pri-
mary products, one that is minimized by an increased flow of nitrogen and
phosphorus from the runoff. Taking into account this factor, the model
demonstrates the stabilization of the primary products of the lake at a level
that is more than two times in excess of allowable levels (Figure 3).
E
W
X
50
40
30
20
10
0
500
^400
o
CD
-cvj' 300
CJ200
Q.
Q_
100
Average depth (H)
Primary products under
different management activities
_i i_
1930 1940 1950 1960 1970 1980 1990 2000 2020 2030 2040
Year
Figure 3. Primary products dynamics of Lake Sevan under various external
effects.
152
-------
ASSESSMENT OF RISKS OF TOXIC POLLUTANTS TO AQUATIC ORGANISMS
AND ECOSYSTEMS USING A SEQUENTIAL MODELING APPROACH
by
R.A. Park1, J.J. Anderson^, G.L. Swartzman^,
R. Morison3, and J.M. EmlerA
ABSTRACT
A sequential modeling strategy is presented in which increasingly complex
fate and eftects simulation models of aquatic ecosystems are used to assess
the environmental risks of chemical exposure. The strategy is based on the
assumption that complex models better represent the natural systems, thereby
decreasing analysis uncertainty.
INTRODUCTION
A major challenge for environmental protection is the assessment of risks
to the quality of terrestrial and aquatic ecosystems from exposure to pesti-
cides and industrial chemicals. The premanutacture chemical review program of
the U.S. Toxic Substances Control Act, for example, specifically calls for the
predictive assessment of environmental risk (USEPA 1979); yet that program
allows only 90 days for the analysis. Mount (1979) has stated that risk
assessment should relate to site-specific properties of ecosystems and to dif-
ferences in organisms, water chemistry, and sediments. Mathematical models
have long been recognized as one means of quickly translating theory, laboratory
findings,'and mesocosm results to field conditions, thus providing a rationale
for risk assessment.
Many models have been proposed to aid in risk assessment. In Figure 1
representative models are arrayed along three axes on the basis of their
spatial, biological, and chemical complexity. To the left of the diagram
iRolcomb Research Institute, Butler University, Indianapolis, IN, USA;
^Center for Quantitative Studies, University of Washington, Seattle, VIA, USA;
30ffice of Toxic Substances, U.S. Environmental Protection Agency, Washington,
DC, USA;
^Environmental Research Laboratory, U.S. Environmental Protection Agency,
Corvallis, OR, USA (Current Address: National Fisheries Research Center,
U.S. Fish and Wildlife Service, Seattle, WA).
153
-------
n
Ol
O
O
O
uu
O
O)
=)
CO
T3
CO O'
> CD
CO Q.
CO CO
MEXAMS
Borgmanni
FOAM
• PART
GETS
/
TOX-SCR
QUAL2E
s
s
s
SWACOM HSPF
t
s
/
1
CEQUALi
ALWAS
AQUATO
PEST
CMRA
S y
/ /
WASTOX
<} A
. 1 2
SPATIAL COMPLEXITY
3-D
Figure 1. Representative models of varying levels
of complexity useful for risk assessment.
are the traditional screening models, exhibiting low spatial complexity and
varying levels of chemical and biological complexity. To the right of the
diagram are models of greater spatial complexity, exhibiting the level of
detail appropriate for application to specific sites. Generally there is a
tradeoff between increasing chemical complexity and increasing spatial com-
plexity. The NOAA and WASTOX models are the most demanding of data and
computational power, they have both spatial and biological complexity and
owe their existence to the intensive field efforts arid large datasets that
have been amassed for the Great Lakes. Such complex models are not appropri-
ate for routine risk assessment.
Recognizing that the U.S. Environmental Protection Agency must evaluate
large numbers of chemicals each year based on minimal amounts of data, our
goal has been to develop a fast and judicious procedure capable of separating
chemicals that can either be accepted or rejected in the initial phase of the
analysis from those that need additional study with more sophisticated tools.
The resulting sequential modeling strategy, based on existing and new models
and designed to assist in decision making, is an important contribution to
environmental risk assessment.
OVERVIEW
Based on personal experience and the deliberations of two workshops, we
have developed a multi-tiered strategy for applying simulation models to
154
-------
assess the probable tate and effects of chemicals. The strategy recognizes
that models ot increasing complexity require more effort to apply, but pro-
vide superior representations of ecosystemic buffering of potentially harmful
chemicals, thereby decreasing uncertainty in the risk assessment.
As shown in Figure 2, fate and effects models of increasing complexity
are used for those chemicals that can neither be accepted nor rejected on
the basis of respectively conservative and liberal assumptions used in multi-
ple executions of simpler models.
MODELING STEADY-STATE, SHORT-TERM EXPOSURE
In its simplest form, exposure can be represented as steady-state par-
tioning of a chemical among environmental phases. Partitioning models can
consider transfers among air, water, and soil compartments; in fact, these
elegant models account for simultaneous transfers, preserve mass balance, and
yet are parsimonious. For this reason, partitioning models have enjoyed a
great deal of popularity as evaluative tools in risk assessment. In their
simplest form, they utilize partition coefficients to predict the distribution
of a chemical.
Efficient computation of both equilibrium and nonequilibrium partitioning
is facilitated using the chemical engineering concept of fugacity, an approach
advocated by Mackay and colleagues (Mackay 1979, Mackay and Paterson 1981,
Mackay et al. 1983, 198'6, Neely and Mackay 1982). Fugacity is the "escaping
tendency" of a chemical from a particular phase; if equilibrium is assumed,
a single fugacity is computed for all phases.
Equilibrium fugacity is represented in the PART model (Park, in prep.),
implemented in the Lotus 1-2-3 spreadsheet program. Missing parameters are
estimated using established quantitative structural activity relationships.
The model has been verified by comparison with Mackay's results. Field
validation was performed using data on parathion, which was applied to give
0.05 ppm initial concentration in Israeli fish ponds (Perry and Gasith 1978);
the predictions seem to be consistent with observations (Figure 3).
The simulation represents steady-state conditions 7.32 days after ini-
tial dosing, the time computed tor fish to reach equilibrium. We propose to
use this model to compute the environmental exposures for both the first- and
second-tier effects models (Figure 2).
MODELING SINGLE-SPECIES COHORT EFFECTS
Dose-response toxicity models provide the basis for all ecosystemic risk
assessments. The challenge is to incorporate dose response into a methodology
that includes natural fluctuations in viability and mortality. The newly de-
veloped HEALTH model (Anderson, in review) assumes that, at the beginning of
a lite stage, an organism has an initial level of health that can increase or
decrease over time as the organism encounters beneficial and deleterious
155
-------
STEADY-STATEESHOHT-TEflM
EXPOSUHE FUSACITY
SURVlVOflSHIP
PART
GNOTOaiOTIC ECOSYSTEM
HEALTH
UWCEBIAJME
ACCI
REJECT
MICROCOSM
KIOKffl
UNCEI
ITAlHDt!
ACCEPT
REJECT
DYNAMIC LONG-TERM
EXPOSURE
NATURAL ECOSYSTEM
STREAM OR RESERVOIR
g
AOUATOX
ACCEPT
REJECT
Figure 2. A sequential strategy for using models and evaluating
uncertainty in risk assessment of chemicals potentially
harmful to the environment.
156
-------
events. The two parameters of the model, initial health and rate of health
loss, can be estimated using as few as three points on a survivorship curve
(Figure 4). Health changes include damages to biochemical systems when they
are "hit" by toxic substances (ct. Clayson et al. 1985).
The dynamics of health are, modeled as a continuous form of the "gambler's
ruin" process, using a random walk with drift in which the organism eventually
loses when health becomes zero and mortality occurs (Figure 5). Because a
broad range of environmental conditions- is implicit in the survivorship curves
that form the basis for the model, and because parameter estimation incor-
porates large sources of error, this model generates a large region of un-
certainty that may require further analysis.
EVALUATING MODEL UNCERTAINTY
Many problems are associated with the communication of risk assessments
(Slovic et al. 1982), especially because of the complexities and uncertainties
inherent in risk data. Fundamental differences in approaching uncertainty by
scientists and regulators result in underevaluation of scientific uncertainty
with regard to risk (Gawiak and Byrd 19W7).
PART Screen 1
A1: [W8] 'PART Ver 2.1 - Fugacity Model (R.A. Park, Holcomb Research Inst.) MENU
Notes Compute Modify Save Quit PPMGraph Graph% Output
Use arrow keys and [Enter] to select choice, [Alt] A to return to menu
ABCDE f GH
PART Ver 2.1 - Fugacity Model
-------
T i me
Figure 4. Change in survivorship curves with
organism age as initial health, Y, and rate
of change of health, R, are varied (Anderson,
in review).
A graphical representation of model uncertainty should help break down
communications barriers between scientists and regulators. A particular model
is run with at least two and possibly many combinations of assumptions, in-
cluding the most conservative and least conservative options, which are rep-
resented by two curves in the graph (Figure 6). The area between the curves
is the region of uncertainty due to model accuracy; biological, environmental,
and chemical variability; and lack of adequate data on the chemical of inter-
est. It the likely environmental concentration for a chemical falls within
the region of uncertainty, further analysis is indicated, either with or
without additional data on the chemical.
MODELING REPLICABLE ECOSYSTEM1C EFFECTS
If a more detailed analysis is indicated, MICMOD (Swartzman and Rose 1984)
may be used. It represents a gnotobiotic ecosystem containing nitrogen, phos-
phorus, eight phytoplankton groups, and five zooplankton groups (Taub and Crow
1980); it has the advantage of having been well validated against the results
of replicable microcosm experiments. It can be used to evaluate both direct
158
-------
and indirect effects of a toxic chemical on zooplankton; the direct effects are
through toxicity, and the indirect effects are through changes in phytoplankton
biomass that serves as food for the zooplankton. The model, however, does not
represent the complex degradation pathways and buffering capacity of a natural
ecosystem, so that the region of uncertainty may still be too great (Figure 2).
MODELING NATURAL ECOSYSTEMIC FATE AND EFFECTS
No truly appropriate ecosystem effects model exists for aquatic systems
at this time (cf. Bartell et al. 1982, 1983, Barnthouse et al. 1982, O'Neill
et al. 1983, Borgmann 1985, Barnthouse and Suter 1986, Bartell 1987). The
AQUATOX model, currently under development by the senior author, is one
response to an obvious need for toxic effects models. The model uses eco-
system bioenergetic algorithms from the CLEANER (Park et al. 1980, 1986) and
LAKETRACE (Park 1985) models, transport and degradation algorithms from the
PART and PEST (Park et al. 1980, 1981, 1982) models, and toxicokinetic
algorithms from the FGETS model (Barber et al. 1986, Suarez et al. 1987).
Dose-response formulations are still being developed.
X
_c
•p
^«
ft*
0)
IE
0
1
T i me , • t
Figure 5. Two random walks of health over time;
mortality occurs when health goes to zero.
Differences in paths are due to random fluc-
tuations (Anderson, in review).
159
-------
0 10 20
Cone. >t g/1
Figure 6. Tripartite decision graph showing median popu-
lation reduction dose-response curves generated by
the most and least conservative implementations of a
model (Morison and Anderson 1987)..
The model (Figure 7) represents an ecosystem with nitrogen (nitrate and
ammonia), phosphate, oxygen, two groups of algae, generalized zoobenthos-
zooplankton, two groups of fish, and detritus in a stream or reservoir. It
is designed to evaluate both direct and indirect toxicological effects on
typical aquatic ecosystems at risk from intentional and unintentional appli-
cations of chemicals; default site conditions are provided. The model is
implemented on a microcomputer and takes full advantage of that user-friendly
computing environment. Pull-down menus (Figure 8) are used to pick options,
including full-screen editing of chemical and biological characteristics
(Figure 9) and site conditions. Simulation results are summarized in tabular
and graphical form (Figure 10).
SUMMARY
In summary, a sequential modeling strategy has been developed in which
increasingly complex fate and effects simulation models of aquatic ecosystems
are used to assess the environmental risks of a chemical. The strategy is
based on the assumption that the more complex models better represent the
natural buffering of biotic systems against deleterious chemicals, thereby
decreasing the uncertainty in the analysis. A graphical representation of
160
-------
Figure 7. The AQUATOX model for simulating the fate and both direct and
indirect effects of toxic chemicals on aquatic ecosystems.
Figure 8. An example of a pull-down menu in AQUATOX.
uncertainty is used to convey the results of the analyses to decision makers,
indicating that, at a particular stage in the sequence, a chemical may be
accepted for licensing, rejected, or subjected to the next stage of analysis.
ACKNOWLEDGMENTS
Research supported in part by the U.S. Fish and Wildlife Service (Coop-
erative Agreement 14-16-0009-87-954) and by the U.S. Environmental Protection
Agency (Contract 7B1037NAEX).
161
-------
LAKE NOCKAMIXON, PA
10 1
1 -;
0.1 -=
0.01 -:
0.001
-13
+PO4 e
nNITRO e
DATE
183
LAKE NOCKAMIXON, PA
100 -|
10 -^
1 -i
0.1 -
0.01
-13
+OTHER e
nZOOPL e
DATE
183
Figure 10. AQUATOX plots of annual patterns of
concentration of key ecosystem components
for default conditions without toxicological
impacts.
REFERENCES
Anderson, James J. In review. Mortality and survivorship based on a sto-
chastic model of organism health.
Barber, M.C., L.A. Suarez, and R.R. Lassiter. 1986. Kinetic exchange of
nonpolar organic pollutants by tish. Environmental Research Laboratory,
Athens, GA (in press).
162
-------
Barnthouse, L.W., D.L. DeAngelis, R.H. Gardner, R.V. O'Neill, C.D. Powers,
G.W. Suter, II, and D.S. Vaughan. 1982. Methodology for Environmental
Risk Analysis. Oak Ridge National Laboratory, Oak Ridge, TN. ORNL/
TM-8167. 67 p.
Barnthouse, L.W., and G.W. Suter, II (eds.). 1986. User's Manual for Eco-
logical Risk Assessment. Oak Ridge National Laboratory, Oak Ridge TN.
ORNL-6251. 215 p. '
Bartell, S.M. 1987. User Manual for Ecosystem Uncertainty Analysis Demon-
stration Program. Office of Toxic Substances, U.S. Environmental
Protection Agency, Washington, DC. 26 p.
Bartell, S.M., R.H. Gardner, R.V. O'Neill, and J.M. Giddings. 1983. Error
analysis of predicted fate of anthracene in a simulated pond. Environ.
Toxicol. and Chem. 2:19-28.
Bartell, S.M., R.V. O'Neill, and R.H. Gardner. 1982. Aquatic ecosystem
models for risk assessment. In: Analysis of Ecological Systems: State-
of-the-Art in Ecological Modelling. Lauenroth, W.K. , G.V. Skogerboe,
and M. Plug (eds.) pp. 123-127. Elsevier Scientific Publishing Company.
borgmann, U. 1985. Predicting the effect of toxic substances on pelagic eco-
systems. In: The Science of the Total Environment 44:111-121.
Clayson, D.B., D. Krewski, and I. Munro. 1985. Toxicological Risk Assessment:
Volume 1 Biological and Statistical Criteria. CRC Press, Boca Raton, FL.
Gawak, W.M. , and D.M. Byrd. 1987. Divergent approach to uncertainty in
risk assessment: mathematical expression compared to circumstantial
evidence. In: Uncertainty in Risk Management and Decision Making. V.T.
Covello, L.B. Lave, A. Moghissi, and V.R.R. Uppuluri (eds.). Plenum
Press, New York.
Mackay, D. 1979. Finding fugacity feasible. Environ. Sci. and Tech. 13-
1218-1223.
Mackay, D. and S. Paterson. 1981. Calculating fugacity. Environ. Sci. and
Tech. 15:1006-1014.
Mackay, D., S. Paterson, and M. Joy. 1983. Application of fugacity models to
the estimation of chemical distribution and persistence in the environment.
In: Fate of Chemicals in the Environment. R.L. Swann, and A. Eschenroeder
(eds.). American Chemical Society, Washington, D.C. pp. 175-196.
Mackay, D., S. Paterson, and W.H. Schroeder. 1986. Model describing the
rates of transfer processes of organic chemicals between atmosphere and
water. Environ. Sci. and Tech. 20(8):810-816.
Morison, R., and J.J. Anderson. 1987. Graphical representation of model un-
certainty for risk assessment. Presented at Ecodynamics Workshop on
Theoretical Ecology, Julich, Federal Republic of Germany, October.
163
-------
Mount, D.I. 1979.
environments.
Factors affecting acceptable margins of safety for aquatic
Fourth ASTM Symposium on Aquatic Toxicology, Chicago, IL.
Neely, W.B. and D. Mackay. 1982. Evaluative model for estimating environ-
mental tate. Modeling the Fate of Chemicals in the Aquatic Environment.
K.L. Dickson, A.W. Maki, and J. Cairns, Jr. (eds.). Ann Arbor Science
Publishers, pp. 127-143.
O'Neill, R.V., S.M. Bartell, and R.H. Gardner. 1983. Patterns of toxicologi-
cal effects in ecosystems: A modeling study. Environ. Toxicol. and
Chem. 2:451-461.
Park, R.A. 1985. LAKETRACE, a menu-driven aquatic ecosystem model for micro-
computers. Presentation at International Society of Ecological Modelling
meeting at Gainesville, FL, August.
Park, R.A., C.D. Collins, C.I. Connolly, J.R. Albanese, and B.B. MacLeod.
1980. Documentation of the Aquatic Ecosystem Model MS.Cleaner. U.S.
Environmental Protection Agency, Athens, GA. (Unpublished report).
Park, R.A., C.I. Connolly, J.R. Albanese, L.S. Ciesceri, G.W. Heitzman, H.H.
Herbrandson, B.H. Indyke, J.R. Loehe, S. Ross, U.D. Sharma, and W.W.
Shust»r. 1980. Modeling Transport and Behavior of Pesticides and Other
Toxic Organic Materials in Aquatic Environments. Rensselaer Polytechnic
Institute, Troy, NY. Center tor Ecological Modeling Report No. 7. 163 pp.
Park, R.A., C.I. Connolly, J.R. Albanese, L.S. Ciesceri, G.W. Heitzman, H.H.
Herbrandson, B.H. Indyke, J.R. Loehe, S. Ross, D.D. Sharma, and W.W.
Shuster. 1982. Modeling the Fate of Toxic Organic Materials in Aquatic
Environments. U.S. Environmental Protection Agency, Athens, GA. EPA/
600/3-82/028.
Park, R.A., B.H. Indyke, and G.W. Heitzman. 1981. Predicting the fate of
coal-derived pollutants in aquatic environments. Paper presented at
Energy and Ecological Modelling Symposium, Louisville, KY, April 20-23,
1981.
Park, R.A. B.B. MacLeod, B. Indyke, C.D. Collins, J.R. Albanese, and D. Merchant,
1986. Documentation of the Aquatic Ecosystem Model MINI.CLEANER. U.S.
Environmental Protection Agency. Athens, GA. (Unpublished report).
Perry, A.S., and Gasith, A. 1978. An Integrated Study of Substrate-Biological
Species Interaction in an Aquatic Ecosystem. I. Fate of Parathion in
Fish Pond Ecosystem and Its Impact on Food-Chain Organisms. Working
Paper 16. Research Contract No. 1724/RB, International Atomic Energy
Commission, Vienna.
Slovic, P., B. Fischhoff, and S. Lichtenstein. 1982. Facts and fears: under-
standing perceived risk. In: Judgement Under Uncertainty: Heuristics
and Biases. D. Kahneman, P. Slovic, and A. Tversky (eds.). Cambridge
Univ. Press.
164
-------
Suarez, L.A. , M.C. Barber, and R.R. Lassiter. 1986. GETS, A Simulation Model
for Dynamic Bioaccumulation of Nonpolar Organics by Gill Exchange: A
User's Guide. U.S. Environmental Protection Agency. Athens, GA. EPA/
600/3-86/057.
Swartzman, G.L., and K.A. Rose. 1984. Simulating the biological effects of
toxicants in aquatic microcosm systems. Ecol. Model., 22:123-134.
Taub, F.B., and M.E. Crow. 1980. Synthesizing aquatic microcosms. In:
Microcosms in Ecological Research, vol. 52. J.P. Geisy (ed.). U.S.
Dept. Energy, Washington, DC.
U.S. Environmental Protection Agency. 1979. Guidance for premanufacture
testing: discussion of policy issues, alternative approaches, and test
methods. Federal Register 44(53):16240-16292.
165
-------
REMOTE MONITORING OF ECOLOGICAL CONDITION OF AQUATIC ECOSYSTEMS
by
A.A. Gittelsonl and A.M. Nikanorov1
ABSTRACT
This work examines the development of scientific grounds for remote sens-
ing of the factors determining buffer capacity of aquatic ecosystems for heavy
metal pollution. These methods may be used in short-term biomonitoring of
heavy metals in surface waters.
INTRODUCTION
One of the basic objectives of water chemistry and aquatic biology is to
develop scientific grounds and methods of monitoring the ecological condition
of aquatic ecosystems. Buffer capacity of a water body is probably an import-
ant factor for decision-making in this kind of monitoring. Ecosystrem buffer
capacity for heavy metals is determined by three basic parameters (Izrael et
al. 1985, Nikanorov et al. 1985): sorption-accumulation capacity of hydro-
biota, complexation capacity of dissolved organic matter (DOM), and deioniza-
tion capacity of bottom sediments. Sorjstion of heavy metals by mineral sus-
pension also is important. Ecological efficiency of removal of metals from
the biogeochemical cycle is determined by metal complexation with DOM (i.e.,
their transformation into less toxic forms) (Prokofiev 1981, 1983), metal
accumulation by hydrobiota, and burial in bottom sediments in the process of
sedimentation.
Metal toxicity is proportional to the rate of a compound's accumulation
in living cells (Izrael et al. 1985). Ionic forms of heavy metals possess
maximum accumulation ability. The ability of DOM to form complexes with
heavy metals is determined by many parameters of the environment, including
mineralization (Lapin and Krasyukov 1986). Ambient temperature and species
composition are important factors of metal uptake by phytopiankton cells.
Thus, we may name the factors mainly determining buffer capacity of
aquatic ecosystems for heavy metal pollution. They are hydrobiota biomass
and its condition determined by chlorophyll-a, suspended minerals and dis-
solved organics concentrations, temperature, and water salinity. Measurement
iRydrochemical Institute, Goskomhydromet, USSR.
"l66
-------
of these parameters with quick-screening methods would allow monitoring of
the aquatic ecosystem's ecological condition at a higher level and decision-
making in the actual time scale.
This work examines the development of scientific grounds for remote
sensing.of the factors determining buffer capacity of aquatic ecosystems for
heavy metal pollution. These methods may be used as a basis for short-term
biomonitoring of heavy metals in surface waters.
REMOTE SENSING OF CHLOROPHYLL-a IN PHYTOPLANKTON
We carried out a whole set of investigations of radiation characteristics
of water bodies in the visual spectrum band and examined their hydrochemical
and hydrobiological parameters. Radiation parameters, invariant in relation
to remote sensing instruments, survey conditions, seasonal variations in
species composition of phytoplankton, etc., are determined on the basis of
this set. Radiation models of mesotrophic and eutrophic water bodies relating
their spectral luminance factors to hydrobiological and hydrochemical parame-
ters are developed.
The following set of instruments was developed.
1. Spectrometer with high spectral resolution for measuring luminances
of upwelling radiation (B+), sky in zenith (Bjj), and irradiance (Eo) in
the mode of continuous recording in the wave band from 430 to 750 nm (Gittel-
son et al. 1986a).
2. Quick-screening spectrometer measuring Bf,BH, and Eo values in nine
spectral channels, 10 to 15 nm wide, in the band from 430 to 750 nm in less
than a second.
3. Portable spectrophotometer for measuring of water attenuation factor
in nine spectral channels, 15-nm wide (Gaivoronskii et al. 1983)
4. Nephelometer for measuring water scattering factors at 90°.
5. Spectrofluorimeter for measuring water fluorescence intensity (Git-
tleson et al. 1986a).
6. Instrument for registration of the data of measurement recording in-
formation for subsequent "unloading" to the computer.
Figure 1 presents spectral luminance factors typical of the studied
water bodies—the rivers Don, Seversky Donetz, and Kuban, and the reservoirs
Tsymlyanskoye, Kuibyshevskoye, Rybinskoye, Sea of Azov, Lake Balaton. Maxi-
mum luminance at the waves from 540 to 580 nm is mainly determined by light
scattering by suspended solids. At the high concentrations of phytoplankton
in the band 630 to 635 nm, we register minimum P(^) and maximum is registered
at the waves 650 to 655 nm. At all CXA concentrations, P(^) minimum is regis-
tered at 670 nm wave and maximum varies from 685 nm at low CYA values to more
than 700 nm is registered at C > 30 mg m
-3
XA
XA
All local P(^) dependency
167
-------
1.0
0.8
3 0.6
*- 0.4
0.2
0
1 - Don River
2 - Seversky Donetz River
3 - Taganrog Bay, Sea of Azov
450 500 550 600 650 700 750
X, nm
Figure 1. Typical relationships between spectral luminance
factor (SLF) and wave length for different water bodies.
extremes are mainly related to spectral motion of the factors of light absorp-
tion by phytoplankton pigments, 3C(A) minima of P(A) at 630 and 670 rm cor-
respond to 5^(^) maxima and maximum at 650 nm corresponds to minimum ata?(A).
P(A) maximum at the wave length more than 685 nm may be determined by lumi-
niscence of phytoplankton pigments (Gordon 1979) and variation in the ratio
between water absorption factors and phytoplankton pigments at high concentra-
tions (Vasilov and Kopelevich 1982). Not excluding possible contribution of
luminescence signal in water radiation at 680 nm, we must point out that the
shift of maximum location to long wave band with C^ increase proves the
validity of the model (Vasilov and Kopelevich 1982).
Decoding features of optically active ingredients in spectrometric in-
formation must be the function of spectral luminance factor (SLF) at differ-
ent waves such as color index, where one of P( A ^ values is as variable
as possible with variation of CK concentration and another one depends on CK
to minimum extent. However, both this value and P(A) should reflect the in-
fluence of survey conditions and apparatus function of the sensor as accur-
ately as possible. Spectral motion of primary hydrooptical parameters is the
most Important factor in selection of decoding features. p(A) varies
with % mainly due to variation ina?(A). In the short wave band of the
spectrum, X(A) is related togp^A), XPQ^(^~) and *B(A) and in the long wave
band UPQB " °) ls related only to,a?T(A) and:3C*B(A), '3e(670) increasing with
Cj^ growth and *(700) decreases. So, p(675) may be the components of decoding
p(700) may be also.
Availability of the above P(A), ;e(A) and o18Q(A) allowed us to try the
following decoding features (Gittelson et al. 1986e, Kondratyev et al. 1987b).
P(700)/P(560); 'P(550)/P(670);
168
-------
P(700)/P(670); P(700)//*f£ P
-------
where the number of stations is n = 68, correlation coefficient is r - 0.962,
factor is F = r2(n-l)/(l-r2) = 843, standard error of 6CXA evaluation is
not more than 2.43 mg m3. Suggested features were validated at Lake Balaton
in a wider range of CXA concentrations. If CXA is 4.5 to 100 mg m~ (Figure 3)
= 122.85[P(700)/P(560)]2'3, mg m
-3
(2)
when n - 103, r = 0.954, F = 2046, <$CXA = 2.8 mg m
tures suggested are as efficient as the above ones.
-3
The rest of the fea-
Stability of models (1) and (2) was tried many times. Using equations
(1) and (2), CXA were estimated on the basis of remote surveys and values
were compared with C^ determined analytically. For the rivers Don and
Seversky Donetz and the sea of Azov, standard deviation of the initial and
measured values does not exceed 3.2 mg m~3 and it does not exceed 2.3 mg m
(Figure 4) for Lake Balaton.
O
0 0.2 0.4 0.6 0.8 1.0
p (700) //>(560)
Figure 3. Relationships between chlorophyll-a
concentration in phytbplankton and p (700)/ p
(560) ratio for Lake Balaton.
170
-------
ro
I
o
'-o
o
I
120
100
80
60
Q.
O
J5
JZ
O
40
20 40 60 80 100 120
Chlorophyll-a (measured), mg rrf3
Figure 4. Comparison of chlorophyll-a concentra-
tion values predicted by radiation model of the
(1) - CnP and analytically determined
REMOTE SENSING OF SUSPENDED SOLID CONCENTRATION
Optical features of suspended solids are mainly determined according to
scattering factors. The most serious variation of P( *•) with C«5« ' variation
is observed in the band 540 to 580 nm; sftp.is minimum in this band. So, the
selection of one of SLF making up the decoding feature is evident—P(560).
The results of factor analysis of spectrometric data (Kondratyev et al. 1987a)
show that in the wave band from 580 to 660 nm and from 500 to 520 nm spectral
variables do not contribute greatly to the dispersion of three marked factors.
This allows us to assume that the influence of SLF is unimportant in these
wave bands to the final factor solution. This allows us to apply SLF in the
above wave bands for valuation of C«ie in decoding features on the basis of
spectrometric information. We evaluated a number of features and the follow-
ing ones turned out to be the most efficient:
Regression equation of C«« constraint with radiation parameters for the
rivers Don and Seversky Donetz is the following
CW = 60.981[P(560)-P(520)]/[P(560)+P(520)]}°-49
where n = 61, r = 0.93, F = 373, bC^ = 3.2 g m~3.
(3)
171
-------
There are two reasons for rather serious errors of Cva estimation. One
oi° them is the variation of 6i80^) and p(X) ln "referent" areas with CW -
variation, the other one is the variation in granulometric composition of par-
ticulate matter, which affects SLF seriously. In the process of remote .sens-
ing of surface water quality, we should probably use an approach for Cat de-
termination and a parameter functionally bound with C«« by the average radius
of the particles and by the law of their distribution according to their size.
This integral parameter determining environment turbidity is rather closely
bound to the factor determining upwelling radiation and possessing maximum
loads in the wave band from 540 to 660 nm (Kondratyev et al. 1987a).
With considerable variation in the concentrations of particulate matter
at the water surface in the process of chlorophyll-a concentration^ determina-
tion of p(620) and p(520), which are in minimum dependency on Cess
is rather efficient.
and C
XA»
ro
tn
o
CD
•o
-------
Measurement of DOM concentration is carried out according to its lumines-
cence in natural waters. Spectre-luminescent features of humic and fulvic
acids are studied to understand the demands for the systems of induction of
the signal of DOM fluorescence, its registration, and determination of decod-
ing features of ingredients on the basis of the information obtained.
When observing at the wave A. = 470 run, the maximum at the excitation
spectrum of fulvic acids is marked at the waves 340-350 ran. When we have
humic acids with A = 550 nm, it is marked at the waves 460-470 nm. In
the wave band from 400 to 450 nm, there is a plateau in the excitation spec-
trum; excitation intensity at this plateau is 0.7-0.8 of maximum (Figure 6).
Spectral features of fluorescence excitation when it is observed at the
waves 470 and 550 nm allows one to estimate concentrations of fulvic and humic
acids in solutions. With excitation at XB = 350 nm, fulvic acid (FA) fluor-
escence is maximum when XH = 470 nm, and humic acid (HA) fluorescence is
- Fulvic acids
- Fulvic acids
- Humic acids
- Humic acids
450 500 550 600 650 700 750
X, nm
Figure 6. Fluorescence spectra of fulvic acids (1,2) and
humic (3,4) with excitation at waves A =350 nm (1,3) and
A =470 nm (2,4). Insertion presents fluorescence excita-
tion spectra: 1 - of fulvic acids With X =470 nm; 2 -
of humic acids with A=560 nm.
173
-------
maximum when X^ = 500 ran (Its fluorescence intensity being much lower than
that for FA). With excitation at Xg = 470 nm, HA fluorescence intensity is
4 to 5 times higher than that for FA (Figure 6). Thus, with fluorescence ex-
citation at the wave Xg = 350 nm and observation at X^ = 470 nm, it is pos-
sible to estimate FA concentration and at XB = 470 nm and Xg = 560 nm, we
may estimate HA concentration. It is reasonable to use iteration (taking into
account contribution of HA fluorescence at XH = 470 nm and FA fluorescence at
Xjj - 560 nm) to estimate their concentrations in mixed solutions.
On the basis of spectral features identified with a lidar-spectroanalyz-
er, possibilities for remote sensing of HA and FA concentrations were studied.
Investigations were carried out in a laboratory at the distances of sensing
from 2 to 5 meters. Solutions were put into quartz glass tubes that were 0.15
diameter and 0.3 m long. Power of laser exciting radiation was from 100 to
300 kW at X = 347 nm when generating impulse at 30 to 35 nm. FA fluores-
cence maximum was registered with a lidar receiver in the wave band 447+_ 10
nm and 540+10 nra for HA (Figure 7). Fluorescence spectral features identi-
fied with laser excitation allow one to estimate concentrations of DOM with
one excitation - 347 nm. Intensity of reverse radiation at the waves 447 and
540 nm influences concentrations of HA and FA in two-component solutions.
Relationships of fluorescent parameter (ratio between fluorescence signal at
certain wave lengths to the signal of water Raman scattering) indicate that
this parameter is a measure of HA and FA concentration in a solution (Figure
7).
We estimated fluorescence section of both components of organic matter
attributed to mean mass molecule. Calculation of fluorescence section ]c=( 1.5+JD.5) * 10"
strad. For such complex multicomponent substances as HA and FA, the fluores-
cence section may vary considerably depending on the relation between mole-
cules with different mass in a solution because efficiency of FA and HA
fluorescence is mainly determined by this factor.
The sensitivity threshold of the method as well as its error depends on
the conditions of measurement and in particular, probing depth, hydrooptical
174
-------
features of the environment, etc. In our experiments, the sensitivity
threshold of the method was less than 5 ppb at the probing depth of 0.25 m.
It is evident that, at greater depths of probing, the sensitivity threshold
of the method will be much lower.
Application of the method of fluorescence parameter at inland surface
waters is often complicated due to rather high values of water attenuation
factor and problems connected to the evaluation of fluorescence contribution
of different organic substances to the signal at the frequency of water Raman
scattering. So, surface waters need a completely different method of evalua-
tion of hydrooptical parameters of the environment and of evaluation of fluor-
escent signal in remote laser probing. It may be used on the SLF measurement
with lidar-spectroanalyzer in the intervals between laser impulses. SLF
equals the ratio between luminescences of radiations upwelling from water and
downwelling upon it. SLF provides information on absorbing and scattering
features of aquatic environment. Using decoding features of optically active
ingredients defined for inland surface waters (Gittelson et al.!986e, Kondrat-
12
10
8
1 - Humic acid
2 - Fulvic acid
3 - Humic/fulvic
solution
1 - Humic acid
2 - Fulvic acid
123456
Cone., mg kg'1
430 460 490 520 550 580
X, nm
Figure 7. Fluorescence spectra obtained on lidar—
spectroanalyzer for humic acid, fulvic acid, and
a combined solution. Insertion presents rela-
tionship between (f>o
175
-------
yev et al. I987b) on spectrometric information and evaluating fluorescence
signal registered by lidar on SLF functional at different wave lengths, it is
possible to take into account the influence of absorbing and scattering fea-
tures of the environment upon the intensity of fluorescent signal registered
by lidar (Gittelson et al. 1986c). This allows one to increase the field of
application of laser spectroscopy considerably and to apply it for remote
sensing of organic matter concentration in surface waters. Testing of the
method suggested on a number of water bodies proved its validity.
REMOTE SENSING OF WATER SALINITY AND TEMPERATURE
The relationship between SHF radiation of water bodies and their physico-
chemical parameters (PCP) is the basis for the SHF radiation method of water
salinity and temperature estimation. Luminance temperature is the measure of
self-radiation. The relationship between the luminance temperature of water
and its PCP is studied under laboratory and field conditions. Laboratory-
measuring the SHF-radiometric complex included radiometers with wave lengths
2.25, 3.0, 7.77, 18, and 30 c and fluctuation sensitivity at least 0.3 K.
All the field measurements were conducted with the help of the airplane AN-2.
Antenna systems of this airborne complex provided relations between airplane
flight altitude, H, and the side of resoluton element at the area D=0.7 H.
Some results of the experiments and calculations are presented in Figures 8
and 9. Field experiments cover the measurements of PCP of water bodies with
salinity from 0 to 400 g kg~l.
The investigations provided stable functional relationship between self-
radiation field intensity and water salinity, which allows remote sensing
(Gittelson et al. 1986b), Gittelson et al. 1986d, Gittelson et al 1987). Im-
portant spectral features in radiation-salinity and radiation temperature re-
lationshps are detected that are the basis for actual solution of an incorrect
inverse problem—a determination of PCP of water bodies according to their SHF
self-radiation and development of the procedure for remote measurement of PCP.
According to modeling options and experimental information, radiation
features of water bodies in the SHF band depend on water salinity and temp-
erature and on1 the condition of the water surface as well. The degree of in-
fluence of different parameters upon radiation features however,, is quite
different according to the spectrum. This peculiarity determines possibilities
for remote sensing of hydrophysical parameters and evaluation of their origin
and intensity on the basis of the data of SHF radiation intensity measurement
(luminance temperature TA) in certain fragments of the spectrum solving a
system of radiation-hydrophysical equations (Gittelson et al. 1986d).
(5)
' FAi(0A"">Qj V =
i = 1,2,... ,m m>n
Because the relationship among Tj? and water temperature and salinity is
non-linear (Figures 8 and 9), Equation 5 is non-linear in the general case.
If the absolute values of temperature, salinity, and other parameters Qj are
176
-------
120
Calculation
Laboratory experiment
Reid experiment
30
60
90
120
150
180
Salinity, g/kg
Figure 8. Relationship between water luminance tempera-
ture in SHF band and its salinity with different
thermodynamic termperature values.
not the features identified but their AQj variations linearized and pre-
sented as:
n
i—1,2,3,...,m
(6)
where ATj?i is variation in luminance temperature at the fragments of spectrum
\±, related to variations in PCP of a water body. a.y^AT.si/AQj are coeffi-
cients describing radiation field sensitivity at the wave length X; to varia-
tions of Qj parameter (within the limits of linear relationship i^T.si(AQj). In
centimetric and decimetric wave bands, the rate 3TW3TO strongly depends on
water surface temperature. The area of wave lengths from 5 to 8 c where emis-
sivity does not depend seriously on water PCP and is about 0.5 K/°C is the
exception. The influence of salinity of water bodies is especially strong in
decimetric area of the spectrum (Figure 8) where the rate of radiation-salin-
ity relationship varies from 0.5 to 1 K/°/ .
177
-------
Possibilities of remote sensing of physico-chemical parameters in water
bodies according to their SHF-self-radiation are tested in laboratory and
field conditions in a wide spectrum of variations in water salinity (from 0
to 400 g kg"1) and temperature (from 5° to 27°C).
Determination of the exact characteristics of the method in the field is
rather a serious problem. Due to high variability of PCP of water bodies and
pilot plots in time and space, joint experiments with simultaneous contact and
noncontact measurements of PCP are presented.
Joint simultaneous measurements of geophysical parameters with quick-
screening contact and remote methods allowed us to estimate exact characteris-
tics of the method in a wide range of PCP variation.
In the result of validation, testing, and mmetrological evaluation of the
method, Its metrological and technological-economic parameters were defined:
CD
120
110
100
90
CD
CL
£ 80
CD
i
70
60
50
40
X=3sm, S = 180%o
Calculation
Laboratory experiment
Reid experiment
X=30sm, S=180%o
10 20
Surface Temperature
30
Figure 9. Spectral relationship between sensitivity of
SHF radiation intensity and variations in thermodynamic
temperature of water surface: S=0°/oo ; S=200/oo; S=40°/
1-T0=30°C; 2-T0=20°C; 3-To=10°C; 1-TQ=00C.
oo;
178
-------
- Response threshold according to salinity from 0.2 g kg l for HC1 to
2 g kg"1 for NaCl;
- Salinity estimation range—from the value corresponding to response
threshold to 400 g kg"1.
- Resolution according to salinity with radiometer time constant
T = 1 s is 1 g kg"1;
- Determination error of area-averaged salinity with radiometer time
constantT = 1 s with confidence 0.95 is 2 g kg"1;
- Resolution according to water surface temperature with radiometer time
constant 1 s is 0.6°C.
f
- Determination error of area-averaged water surface temperature is not
higher than 2°C with confidence level 0.95 in temperature variation
range from 0 to 35°C.
Spatial resolution is 0.7 of the flight altitude. Capacity of hydrochem-
ical survey of the pilot plot with 1 km distance between the routes is 162
km2/hour. The results of hydrochemical survey are presented in 1 hour as the
charts of salinity and water temperature variations.
Figure 10 presents comparison of water salinity data measured with con-
tact methods and remote sensing using SHF-radiometric complex. Figure 11
presents comparison of contact and remote sensing data on thermodynamic
temperature of water bodies.
CONCLUSION
The investigations created the necessary prerequisites for development
and application of noncontact methods of chlorophyl-a concentration estima-
tion in phytoplankton with standard error not more than 2.3 mg m~3, suspended
soils concentration with standard error less than 3.5 ppb, concentrations of
humic and fulvic acids with the error less than 5 ppb, water slainity with
the error about 0.5 g kg"1 and temperature—less than 0.6°C. Validation of
these methods with short-term control of aquatic ecosystems allows us to
conclude that their application further improves monitoring of water bodies,
provides short-term determination of buffer capacity, and allows decision
making in real-time.
REFERENCES
Gaivoronskii, Y.F., A.A. Gittelson, G.A. Dubovitskii, G.P. Keidan, and L.L.
Lopatchenko. 1983. Spectrophotometer for remote measuring of hydro-
optical characteristics. Pages 231-233 In: Technology for the State
System of Observation and Control of the Environment. Central Design
Office, Obninsk USSR.
179
-------
• Loke Monytch-Gudilo
A Lake Manytch-Gudilo
o Black Sea
A Ochotsk Sea
a Dnieper-Bug Estuary
• Syvash Bay
16 18 20
Salinity (contact), g kg'1
Figure 10. Comparison of results of water salinity
measurements with contact (Mcont) and SHF radiation
(Mdist) methods.
Gordon, H.R. 1979. Diffucive reflectance, of the ocean: the theory of its
augmentation by chlorophyll-a fluorescence at 685 nm. Appl. Opt.18(8):
1161-1166.
Gittelson, A.A., N.A. Dubovitskii, G.P. Keidan, and L.L. Lopatchemko. 1986a.
Remote Determination of Hydrochemical and Hydrobiological Characteristics
of Ecosystems by Means of Their Radiation in the Visible Spectrum Range.
Hydrochemical Institute, Rostov-on-Don, USSR. Publication no. 8010-V-86.
50 p.
Gittelson, A.A., A.G. Grankov, Y.F. Gavoronskii, et al. 1986b. Results of
meteorological attestation of a superhigh-frequency radiometric complex
for non-contact remote determination of mineralization (salinity) and
temperature of aqueous objects. Pages 33-36 In: Non-contact Methods and
Means of Measuring Oceanographic Parameters. Gidrometioizdat, Moscow
USSR.
Gittelson, A.A., A.B. Gretsov, G.A. Dubovitskii, and V.A. Feigelman. 1986c.
The feasibility of the remote laser method of determining organic subs-
tances in inland surface waters. Gidrokhimicheskie materialy. XCIV:3-10.
180
-------
Gittelson, A.A., L.L. LOpatchenko, B.M. Liberman, and F.A. Mkrtchian. 1986d.
Remote det'erminiation of the physical and chemical parameters of aqueous
objects by their irradiation at superhigh-frequency wavelengths. Pages
328-355 In: Comprehensive Global Monitoring of the Status of the Bio-
sphere, Third International Symposium. Gidrometeoizdat, Moscow USSR.
Gittelson, A.A., A.M. Nikanorov, G. Szabo, and F. Szilagyi. 1986e. Monitor-
ing of detected changes in surface waters. Pages 111-121 In: Proceedings
of Budapest Symposium in Water Quality, July 1986. JAHS Publication.
Izrael, Y.A., A.M. Nikanorov, I.A. Lapin, A.V. Zhulidov, and N.A. Dubova. 1985.
Evaluation of buffer vessel of small water channels for heavy metals.
Reports of the Academy of Sciences. 283(3):703-706.
Kondratyev, K.Y., G.P. Garbuzov, and A.A. Gittelson. 1987a. A new approach to
the problem of monitoring the state of aqueous ecosystems by diffusly
scattered radiation spectrum. Reports of the Academy of Sciences. 265(4).
Kondratyev, K.Y., A.A. Gittleson, and G.P. Keidan. 1987b. A method for remote
determination of hydrochemical and hydrobiological characteristics of
aqueous objects. Reports of the Academy of Sciences. 295(2).
Lapin, I.A. and V.N. Krasyukov. 1986. The role of humus substances in complex
formation and migration processes of heavy metals in surface water. Vod-
niye resursy. (1):134-145.
Nikanorov, A.M., A.V. Zhulidov, and A.D. Pokarzhevskii. 1985. Biomonitoring
of Heavy Metals in Freshwater Ecosystems. Gidrometeoizdat, Leningrad
USSR. 144 p.
O
o
CD
"o
CD
CD
E
CD
Q.
I
28
26
24
22
20
18
0
• Lake Manytch-Gudilo
A Lake Manytch-Gudilo
• Syvash Bay
o Caspian Sea
A Caspian Sea
a Caspian Sea
0 Caspian Sea
18 20 22 24 26 28
Temperature (contact), °C
Figure 11. Comparison of results of thermo-
dynamic water temperature measurements with
contact (Tcotlt) and SHF-radiometric (Tdls(;)
methods.
181
-------
Prokofiev, A.K. 1981. Chemical forms of mercury, cadmium and zinc in natural
aqueous media. Uspekhi khimii. (l):54-84.
Prokofii-v, A.K. 1983. Determining physical and chemical forms of trace ele-
ments in natural waters. Uspekhi khimi. (3):483-498.
Vasilkov, A.P. and O.V. Kopelevich. 1982. The reason for the approximate
700 nm in the spectrum far radiation from the width of the sea. Okeanol-
ogiya. XXII(6):945-950.
182
-------
-------
S- 03
8
01
K)
O>
00
CTl
CO
CO
.? 3
3- 0 0
sjs-i-
2 o i
522
s °
§•§
I
U)
'•&-'<;
------- |